PBCore 2.0 Schema Development Tasks

Document Sample
PBCore 2.0 Schema Development Tasks Powered By Docstoc
					ID   Related Request       Change Request
     IDs (see attached
     WGBH compiled
     requests)

R1   5,8a,9,46,72          For creator, contributor, and publisher
                           provide support for management of (a)
                           affiliation, (b) character portrayal
                           (contributor only?), and (c) additional
                           roles.




R2   10,19b,45,65,66,84,89 Guidance on PBCore Profile Development
R3   56             Allow management of dates and types of
                    dates that pertain to the intellectual
                    content of the asset (beyond
                    coverage_spatial).




R4   68             Provide recommendations to
                    identify/describe asset types that do not
                    conform to the typical TV-Guide-style
                    genre lists (such as pre-production
                    classifications of content such as
                    interview, broll, stock footage, sot,
                    outtakes, etc)
R5   12b            Adjust language regex to allow for empty
                    <language> elements.




R6   24             Allow for multiple expressions of
                    formatGenerations inside an instantiation.

R7   19a,20,30,32   For technical fields (such as frameSize,
                    bitDepth, frameRate), position the
                    controlled vocabularies as
                    recommendations of expression and not
                    the definitive list.
R8    6,73              Provide guidance on fulfillment of
                        PBCore's technical metadata.




R9    11,17,21b,23,57   Provide a method to support the
                        technical description of instantiations
                        made up of multiple parts within a single
                        instantiation structure.




R10   16,20,26,28,73    Clarification of formatStandard and
                        formatEncoding
R11   14,40,91      Allow support for multiple expressions of
                    time-based descriptive metadata within a
                    single asset records.




R12   19a,20        Add a 'unit_of_measurement' attribute
                    on appropriate technical metadata fields.




R13   18,41,51,87   Allow support for inline XML (xsd:any)
                    within extension, rights, and (maybe)
                    description.
R14   19b,21a,42,71   Descriptive Vocabulary Review and
                      Additions, Make the namespace, title, and
                      version of recommended vocabularies
                      more clear (to facilitate better integration
                      in source and version attributes)




R15   7a, 7d,29,42    Technical Vocabulary Review and
                      Additions, Make the namespace, title, and
                      version of recommended vocabularies
                      more clear (to facilitate better integration
                      in source and version attributes)
R16   63       Updates to extension and annotation
               (and essenceTrack annotation) structure.




R17   67       Remove semantically-meaningless
               ordering requirements for PBCore
               elements. Additional adjust minOccurs to
               prevent semantically unclear expressions.




R18   2,3,69   Guide support for non-unique identifiers
               such as episode numbers, chapter
               numbers, series numbers, series
               identifiers, and other identifiers.
R19   4,36,44,59,74,88   Allow support for Linked Data




R20   12a,22a            Recommend ISO 639-3


R21   25                 Support subjectTypes (similar to MODS,
                         MARC, and others)
R22   43,78              Allow or provide guidance on validatable
                         expression of multiple PBCore assets
                         within a single XML document/feed.


R23   47                 Allow complex coverage statements (as in
                         PBCore 1.0)
R24   49,80        Relation Element




R25   87           Allow classification of rightsSummary
                   statements




R26   31?,58,81?   Provide guidance on how to or provide
                   specific means to document information
                   of the relationship being instantiations of
                   an asset.




R27   90           API documentation
R28   Provide guidance and clarification on use
      of source and version attributes. //
      Resolve overlap between source attribute
      and
      subjectAuthorityUsed/genreAuthorityUse
      d
Use Case                                                Related         Is   Is   Is
                                                        PBCore         Sche Voca Docu
                                                        element(s)      ma bular ment
                                                                              y ation

a. Allow association of creators, contributor, or        creator        x         x
publishers with affiliated organizations (stations,      contributor
companies, etc).                                         publisher
b. PBCore does not specifically address character
portrayal in the contributor elements. For instance for
a Sesame Street segment, PBCore can state
<pbcoreContributor><contributor>Henson,
Jim</contributor><contributorRole>Actor</contributo
rRole></pbcoreContributor> but does not provide a
specific means to clarify whether Jim Henson is
portraying Ernie or Kermit.
c. PBCore limits expression of roles to one per creator,
contributor, or publisher, although frequently one
creator, contributor, or publisher performs more than
one role to avoid entering a contributor multiple times
to state all of their roles.
Many PBCore users work in specialized or controlled                               x
collections where much of PBCore's field structure and
controlled vocabularies are not relevant. Provide
guidance on constraining and profiling PBCore (via
controlled vocabularies, suppression of elements,
authorities, and extensions) to streamline application
of PBCore within controlled environments. See Jack's
comments on ensuring lossless interoperability
between PBCore and COVE via profiles.
For example, the film “Gone With the Wind” was             date?              x       x
released in 1936. If cataloging a VHS instantiation of
the film then dateCreated refers to the VHS tape's
release. PBCore should provide a specific method to
store dates affiliated with the intellectual content (i.e.
to say that GWTW was release in 1936).
Additionally documentation on dateCreated and
dateIssued should be clarified. Often users will store
dateCreated and dateIssued values that do not pertain
to the instantiation or even occur before the technical
format listed in formatDigital or formatPhysical even
existed (due to using the instantiation-specific date
fields to manage asset-specific date information).


More attention is needed within controlled                 genre?                 x   x
vocabularies and documentation in regards to non-
broadcast media material in order to increase
relevance of PBCore to earlier stages of media
workflow.


With PBCore 1.3 an empty language element, such as         language           x
<language/> will product a schema validation error:

element language: Schemas validity error : Element
'{http://www.pbcore.org/PBCore/PBCoreNamespace.h
tml}language': [facet 'pattern'] The value '' is not
accepted by the pattern '[a-z]{3}((;[a-z]{3})?)*'


An instantiation may serve more than one role within       format             x
a collection, such being both the 'Original' and the       generations
'Preservation Master'.
Many requests were to supplement existing controlled       frame size, bit        x   x
vocabularies with additional values for bitDepths,         depth, frame
frameSizes and other technical specifications. The         rate, file size,
expression of these fields should be based on the          data rate,
nature of the instantiation rather than the contents of    others?
the picklist. (i.e. Saying "20 fps" (not on the list) is
certainly better than saying "Other" which is on the
list.)
In most cases, technical cataloging of file based                              x
content can be performed more effectively and
accurately by software rather than manual work.
Provide guidance on tools that can aid in technical
cataloging, such as file utility to assess the mime_type
(formatDigital value) of an instantiation.

A film copy made up of multiple reels.                     instantiation   x   x
Elementary streams (i.e. Omneon generated files from
PBS: aiff, m2v, and Quicktime reference file).
Multiple files that must be concatenated in
presentation (i.e. DVD VOB structure).
An asset contains one of more instantiations
comprised of more than one part that all belong to A
single intellectual object.
Up to PBCore 1.1.1 formatStandard performed a dual- formatStandard?        x   x
duty of commenting on the video signal type or the
container of architecture of the digital file. Given that
formatStandard was depreciated in 1.2, there is now
essenceTrackStandard to manage video signal type but
nothing to track the architecture. For instance,
knowing whether a digital file is BWF format or OP1A
MXF is too specific to be managed within formatDigital
(mime type only) but often essential for workflow. This
could make formatStandard (or a renamed version of
it) more equivalent to formatPhysical were there is a
structure way to express generic and specific
information about the architecture of the file. Adjust
picklist and recommendations for the use of the field
distinct from essenceTrackStandard (which focuses on
video signal, file architectural info isn't as relevant at
the track level). Additionally as formatStandard,
formatEncoding, and formatDigital are often confused
documentation and examples should be clarification
to make their meaning more clear and disctinct.
(Additionally the picklist for formatEncoding contains
values not directly related to encoding such as MXF.
There is also more need for attention to analog
"Segmentation is used for a variety of purposes. For                    x   x
example, when logging a file, a cataloguer may want
to note than a celebrity was featured from minute 2 to
minute 4. For rights management, it might be
necessary to understand that a still photograph was
featured, for which there are no rights to distribute
online. From the perspective of navigation,
organizations might need to insert chapter points into
a video.", "It is important to be able to describe
segments of an asset and relate those to the whole,
rather than have to create new PBCore records for
each segment (and have flimsy relationships between
segments and assets)."

Management of 'fps', 'ips' and 'rpm' per frame rate,                    x   x
'kilobits/sec', 'megabits/sec' for data rate, etc.
Allowing more technical fields to focus on machine-
readable expression of numbers or equations rather
than extra text expressions. // (Consider audio
catalog resist using a field called framerate for ips and
rpm data, consider renaming or moving.)

Allow for possibility for PBCore to be transmitted with description     x   x
external standards (METSRights, ODRL, PREMIS,           rightsSummary
custom data, etc) expressed inline.                     annotation
                                                        extension
genreAuthority                                            genre             x
  add references to id3 as a genreAuthority               contributorRole
  genre recommendations from Andy Lanset
     Unknown
     Station Identification
     Sound Effects
     Public Hearing
     Oral History
     Mood Music
     Lecture
     Soundtrack (film or filmstrip)
     Field Recording
     Concert Rehearsal
     Advertisment or Commercial (which would refer
to conventional commercials as well as political or
campaign ads)

contributorRole
   Announcer
   Senior Broadcast Engineer
   Senior Producer
formatGenerations                                         generations,      x
   remove prepended media_type from                       physical
format_generations controlled vocabulary
   from Andy Lanset:
     add 'Listening Master' (audio equivalent of
'Viewing Master')
     add 'Mix - Minus' (original content without
announcer) and 'Mix - Plus" (announcer and
intermission production)
   from Mary Miller:
      dupe sound negative
      dupe picture negative
      duplicate preservation print

formatPhysical
   HDCam SR
   Audio Cassette
   For the existing formatPhysical list, where specific
terms are supported, add support for generic terms
(i.e. '1 inch videotape: PI-3V' is supported by not '1
inch videotape').
Extension currently contains three discrete values          extension,annota   x
(authority, type and value) within two fields. Also the     tion,essenceTrac
extension can only contain key-value pairs and not          kAnnotation
hierarchical structures. Add third field extension and
allow support for inline XML. Also since annotation is
the extension of the instantiation, follow the same
structure here as well.

"The schema should not require a particular sequence                           x
of element tags. This is unusual in my experience for
schema, and I believe it unnecessarily increases the
complexity of creating valid PBCore XML. I also don't
see any upside to people who are writing PBCore
parsers to have the tags in a particular order. I believe
most parsers simply read the XML and create a data
object which may no longer preserve the order." //
PBCore currently allows confusing structural
expressions such as
<pbcoreContributor><contributorRole>Producers</con
tributorRole></pbcoreContributor> since many of the
primary elements of the PBCore containers are not
themslves required within the container (additionally
the schema's comments and minOccurs statements
frequently contradict each other on whether an
element is required or not).

"Provide a formal structure for series of episodes -                           ?   x
insuring that they remain linked together and not
bound by series name." / "Provide a formal structure
for chapters within an episode - vital for web use. Each
chapter must have separate meta-data, while the
episode also can have the same meta-data structure."
"The idea is you could have a Contributor named                    x       x
'Hughes, Langston', or 'Smith, John' where the
common name may or may not be ambiguous or
unique. But if you add an actual UID to the name, it
would be unambiguous and machine-readable. The
Linked Data way of doing this is to include a URI for
the person. This makes it possible to traverse the
universe of objects containing this URI, and it links our
PBCore object to that universe. We could optionally
add URIs to place names, organizations, subjects, and
relationships in the PBCore record. The way to allow
this would be to add optional attributes to the PBCore
schema, so you could include UIDs inside elements."


Include ISO 639-3 (new since PBCore language            language           x
documentation was written) as an addtiional
recommendation for language authority.
Allow distinction of the type of subject (whether       subject    x       x
geographical, topical, personal, etc).
Exchange of PBCore data between systems is                         ?       x
complicated by lack of recommended methodology or
means to package multiple assets into a single
document or package (see MARCXML's collection
element, modsCollection, atom, etc)
Allow expression of coverage to link expression of                 ?   x   x
coverage, such as an expression of geospatial data
with a human-readible geographical name, and a date
that represents the coverage of that geographical area.
The relationIdentifier is the only PBCore identifier that relation   x   x   x
does not express the asscoiated source of the
identifier. Within systems that provide globally unique
identifiers, this is fine; however systems that assign
identifiers that are only unique within the system can
not properly resolve. Additionally even if the
relationIdentifier value is globally unique, it may not
provide enough information to determine under what
authority or system that identifier can be resolved
against. -- Additionally relationTypes only express
parent-child relationships and not sibling relationships
(such as sequential ordering and positioning, see also
R18)
The utility of rightsSummary could be greatly             rights     x   x   x
expanded by providing a rightsType field and including
the source and version attribute on the rights
element. Potential rightsType could include:
Permission Statement, Constraints, Obligations on
Use, Access Permission, Indemnification Statement.

Within archival and production environments, assets relation         ?       x
will be represented by multiple instantiations to serve
different purposes (DVD Master, broadcast copy, web
copy, backup, etc). Typically all copies are derived
from one another. Provide a means to track
associations of formatRelatorType, formatRelatorID,
and formatRelatorIDSource to allow representation of
relationships such as 'is dub of', 'is copy of', 'is digitized
from', etc. [Note DD also expressed need to express
instantiation-specific rights statements]

Provide sample documentation for interacting with                            x
PBCore records such as using OAI to harvest PBCore.
Better documentation to associate the Element          source/version   ?   x
Version and Namespace attributes from the PBCore
data dictionary with their implementation in the
source and version attributes. (Allow a version
attribute at PBCoreDescriptionDocument as it is
currently tough to figure out what version of PBCore a
document is). (Consider using SKOS or publication
dates to version PBCore's recommended picklists to
encourage more consistent referenece with PBCore
records).
 Is  Is
Use Profil
Case e




 x     x
x




    x
x
x




x
 Chg     New      v1.                  WGBH PBCore Team
                      v2.0 v3.0 n/a                                  Name
Req. #   Req. #    3                     Evaluated Type



  1       R15     x    x              Controlled vocab with source   Bruce Jacobs
                                      reference. This could be a
                                      PBCore picklist.



  2       R18          x              Similar to 49. Address in      Bruce Jacobs
                                      documentation and training
  3       R18          x              similar to #2                  Bruce Jacobs


  4       R19          x              similar to 44. Use 44          Jack Brighton



  5       R1           x              same as 2nd half of 46, and Mary Miller
                                      same as 72. Schema change.




  6       R8           x              Already included in 1.3.       Mary Miller
                                      Documentation update
                                      needed. Need to review
                                      controlled vocabulary.
7a   R15   x       Controlled Vocabulary review Mary Miller
                   - formatGenerations




7b             x   Already able to do this         Mary Miller




7c             x   Already able to do this - each Mary Miller
                   file could be a record




7d   R15   x       Controlled Vocabulary review Mary Miller
                   - formatGenerations


8a   R1    x       Schema Change - may add         Mary Miller
                   optional annotation attribute
                   to string elements




8b             x   1.3 allows you to source your Mary Miller
                   Controlled Vocabulary... don't
                   have to use PBCore picklist
9    R1   x   Similar to 46a. Possible       Mary Miller
              schema change, parsing
              challenge




10   R2   x   use minimal required fields    Rich Rarey
              for PBCore, add extensions
              for domain specific data.
              Could also use a style sheet
              to transform PBCore
              metadata. Use case and
              training issue.
11    R9    x   x       Can make an annotation on       Mary Miller
                        instantiations now. Addressed
                        also in 21, 37 --- This is
                        really 2 Questions: 1. how to
                        record 1 program that spans
                        multiple tapes (part 1, part 2,
                        etc.), 2. need clarification -
                        does she mean multiple types
                        of assets for one work or
                        multiple parts for one work?
                        Whichever one, we need
                        better documentation and use
                        case because this is now
                        possible with 1.3. Same as
                        #23.




12a   R20       x       Update documentation,            Dave Rice
                        include ISO 639-3 as a
                        language recommendation.
                        Need clarification about
                        empty tag issue.



12b   R5        x       Need clarification about         Dave Rice
13a                 x   empty tag issue. May be
                        Recommend stripped down          Kara Van
                        PBcore (minimal fields) or       Malssen
                        another instantiation data set
                        (eg MPEG7).

13b         x           Solved by 1.3. Instantiation     Kara Van
                        no longer required.              Malssen




14    R11       x       Solved by 1.3. assetType         Kara Van
                        allows description of            Malssen
                        segments. Also format
                        timestart and duration
                        describe the segment. Need
15                 x       Add radio genres to           Stefan Wray
                           vocabulary for genre




16     R10         x       Similar to #1 and #6. Need to Yvonne Ng
                           review these cvs.
                           (formatStandard is PBCore
                           1.1)
17     R9      x   x       Same as 11. Need better       Yvonne Ng
                           documentation and use cases


18     R13         x       Will give examples for rights - Yvonne Ng
                           1) allowing for a link to
                           external document and 2)
                           American Archive use case
                           example spelling out use of
                           rights schema as extension
19a   R7,R12       x       See                             Andy Lanset
                           http://pbcore.org/PBCore/form
                           atFrameRate.html. Will
                           consider a field name change.



19b   R2,R14       x       Review genre controlled       Andy Lanset
                           vocabulary




19c                    x   Can be done with                Andy Lanset
                           formatPhysical. Documents
                           can be instantiations. (include
                           as example in documentation)
20    R7,R10,R       x   Review controlled               Andy Lanset
         12              vocabularies for sound.




21a     R14          x   CV review                       Andy Lanset




21b     R9       x   x   Same as 11. Need better     Andy Lanset
                         documentation and use cases




22a     R20          x   same as 12a. Update             dave rice
                         documentation




22b              x       solved by authority attribute in dave rice
                         1.3. Document.
23   R9     x   Same as 11. Need better     dave rice
                documentation and use cases




24   R6     x   Schema change                   dave rice



25   R21    x   Schema change and               Leslie Bourgeois
                documentation. Thinking of
                adding 3 attributes for 2.0 -
                source, version, and
                annotation. Need to structure
                annotations, or add type
                attribute.

26   R10    x   Done but review controlled      Jabran Soubeih
                vocabularies. Update
                documentation




27          x   Done but review controlled      Jabran Soubeih
                vocabularies. Update
                documentation



28   R10    x   Review frame rate CV            Jabran Soubeih


29   R15    x   Review formatPhysical CV        Jabran Soubeih

30   R7     x   Review framesize CV             Jabran Soubeih




31   R26?   x   Could use annotation field. Or Jabran Soubeih
                create specific relationship
                type. Either solve or create
                example of how to do this in
                documentation.




32   R7     x   Review formatbitdepth cv        Jabran Soubeih
33             x   Need clarification. Start time Jabran Soubeih
                   plus duration = end. Frame
                   specific metadata better dealt
                   with by another schema. Or
                   create a pbcore record per
                   frame.
34             x   Done. Express with use of      Jabran Soubeih
                   colons and semi-colons. See
                   http://pbcore.org/PBCore/form
35             x   Use pbcoreRelation and         Jabran Soubeih
                   assetType




36   R19   x       will review for controlled     Barrett Golding
                   vocabulary input.




37             x   multiple records can point to Stephanie
                   the same formatIdentifier,    Sapienza
                   each instantiation could have
                   a different start time on the
                   dvd

38             x   PBCore is not intended for  Walter Allasia
                   preservation metadata.
                   Recommend implementing
                   extension and using
                   PREMIS. Or create a METS
                   wrapper and include PBCore.
                   (Could show example in
                   documentation)



39         x       Update documentation - use     Walter Allasia
                   essenceTrackType.




40   R11   x       Similar to #2 - segmentation   Walter Allasia
                   issue. Use assetType from
                   1.3. Documentation needed.
41    R13      x   Same as #18                 Walter Allasia




42   R14,R42   x   Controlled Vocabulary review Andy Lanset
                   - formatGenerations and
                   contributor




43    R22      x   similar to 78. Schema change Daniel Jacobson
                   or a documentation/use case
                   solution
44   R19   x   similar to #4.                  Jack Brighton




45   R2    x   creating profiles - applications Jack Brighton
               of PBcore for different
               purposes (eg COVE, SIP for
               AA) - Question of ability to
               deliver within existing
               resources/timeline.Also,
               potential overlap with
               Broadway Video work about
               SIPs, AIPs and DIPs..




46   R1    x   Similar to 5.                   Hannah
                                               Sommers
47   R23   x       Need clarification of linkage     Hannah
                   from Dave Rice                    Sommers




48             x   controlled vocab review           Hannah
                                                     Sommers




49   R24       x   similar to #2. relation defines   Hannah
                   relationship but need to          Sommers
                   research sequencing
                   capability.
50         x       Need clarification. Seems like Hannah
                   Hannah wants to create a        Sommers
                   record for this unique sound
                   and its source. 1.3 addresses
                   this by allowing instantiation-
                   free records.




51   R13   x   x   Transcripts can be a             Hannah
                   description type. Possible to Sommers
                   insert html into a pbcore
                   string using escape
                   characters. We should
                   provide documentation on
                   how to do this. You can also
                   create a pbcore record for
                   your transcript and relate it to
                   the media asset. WGBH does
                   with TEI and SMIL to
                   synchronize transcript to
                   media.
52             x   Same as 18 and 41                Hannah
                                                    Sommers




53         x   x   Similar to 78. 1.3 provides the WGBH -known
                   mechanism to solve this.         needs
                   Need to define a collection
                   profile and/or create
                   documentation. Review
                   discussions on how to create
                   collections and feeds. May
                   use METS wrapper or Jack
                   Brighton's collection structure.

54             x   This is addressed in 1.2.     WGBH -known
                   Need to document              needs
                   essencetrack and how to
                   repeat.
55         x       x   In 1.2                          WGBH -known
                                                       needs



56   R3        x       Review date functions in        WGBH -known
                       PBCore                          needs




57   R9        x       Similar to 11, 21, 37.          WGBH -known
                       Documentation issue             needs


58   R26           x   use different unique ids        WGBH -known
                                                       needs
59   R19       x       Similar to 78 and 53. 1.2       WGBH -known
                       does some of this.              needs
                       StructureMap idea is like
                       METS and the collection
                       change requests.




60             x       Kara did some work on this      WGBH -known
                       for NDIIPP. A profile thing.    needs
                       see notes. Create profiles to
                       handle PREMIS.

61                 x   implementation issue, not a     WGBH -known
                       PBCore issue                    needs

62                 x   out of scope                    WGBH -known
                                                       needs

63   R16       x       rework extension structure.     WGBH (Dave M)
                       Need either more fields in
                       extension or attributes.

64                 x   As far as we can tell an    WGBH
                       application issue, not a
                       schema issue. Maybe part of
                       the DD atom feeds issue
                       number 78.
65    R2?     x       Profile or use case issue with WGBH
                      additional fields for
                      acessibility as extensions.
                      Create use case or user
                      profile? NB: Change request
                      has indicated to us that
                      people don't know how to use
                      PBCore - we need to figure
                      out how to deal with this and
                      prioritize list for use cases.




66    R2      x       controlled vocabulary issue - WGBH
                      stripped down subset of
                      vocabuary, limited. Also a
                      use case - how to develop a
                      controlled vocabulary for your
                      use of PBCore.
67    R17     x       Schema issue DM and PP         WGBH - Peter P
                      need to work out. Add
                      structure or not add
                      structure.....


68    R4      x       controlled vocabulary issue.    WGBH-MLA
                      Either an asset type or in
                      another element. Could be
                      annotation or attribute issue
69    R18     x       controlled vocabulary issue.    WGBH -
                      Either an asset type or in      Inventory
                      another element.
70                x   can already be done,            WGBH -
                      misunderstanding                Inventory


71    R14     x       controlled vocabulary           WGBH -
                                                      Inventory
72    R1      x       similar to 5 and 46 Could       WGBH -
                      add field for affiliation.      Inventory




73   R8,R10   x       Same as 26, 29. Controlled      WGBH -
                      vocabulary review and           Inventory
                      documentation
74   R19       x   Review MODS for         WGBH - Chris B
                   authorityURI protocol




75         x       done, need better
                   documentation




76         x       done, need better
                   documentation
77         x       done, need better
                   documentation




78   R22       x   Similar to 43. Schema
                   change or a
                   documentation/use case
                   solution




79             x   Similar to 53. Documentation.
                   Now that we have assetType
                   (1.3) we can combine with
                   existing relationtypes. Also
                   relates to collection needs.



80   R24   x       done, need better
                   documentation
81   R26?       x   Rights per instantiation.




82          x




83          x       done, need better
                    documentation
84   R2   x   done, need better
              documentation




85        x   done, need better
              documentation




86        x   done, need better
              documentation
87   R13,R25   x   Same as 18, 41, and 52




88    R19      x   Documentation and review
                   common practices.




89     R2      x   Documentation.
90   R11   x   Will add example of how to
               do this to documentation.
               Show how to use OAI to
               harvest PBCore metadata.




91   R27   x   Documentation. Same as 2,
               3, and 49.
Title                  Change Request


                                    Public 2.0 Change Requests
Chief Technologist     Much more tightly constrain the media taxonomy, so there is hope for
                       machines to be able to read the data.

                       For example, within PBCore currently, I could use the terms "HD", "High
                       Definition" or "High Def", "1080i" or "720p" interchangeably. A human
                       could read these. A machine could not.
Chief Technologist     Provide a formal structure for series of episodes - insuring that they
                       remain linked together and not bound by series name.
Chief Technologist     Provide a formal structure for chapters within an episode - vital for web
                       use. Each chapter must have separate meta-data, while the episode also
                       can have the same meta-data structure.
Attributes             The addition of attributes to elements such as subject, genre, and
                       contributor to allow inclusion of UIDs for the values they contain. This
                       would enable PBCore to be very useful in the emerging linked data
                       universe.
Peabody Awards         Currently, in the containers for creator and contributor, you can't indicate
Collection Archivist   the contributor's affiliation.

                       Here's an example of a type of credit I work with on a regular basis:

                       Producer(s): Ric Burns (Executive producer)
                       Margaret Drain (Executive producer, American Experience)
                       Judy Crichton (Executive producer, WGBH)
                       Lisa Ades
                       Ric Burns
                       Steve Rivo

                       PBCore does not provide a way to document the person's affiliation.
                       One solution:

                       <pbcoreCreator>
                       <creator>Burns, Ric</creator>
                       <creatorRole>Executive producer</creatorRole>
                       </pbcoreCreator>

                       <pbcoreCreator>
                       <creator>Drain, Margaret</creator>
                       <creatorRole>Executive producer</creatorRole>
                       <creatorAffiliation>American Experience</creatorAffiliation>
Peabody Awards         Maybe we just aren't speaking the same language, but we cant' find a
Collection Archivist   formatDigital value for the streaming flash files we are creating, unless
                       you are using video/vnd.sealed.swf to cover flash.

                       We currently create .flv files using an on2 codec and f4V files using an
                       H.264 codec.

                       Maybe you are using video/H264 for f-4v files but what about .flv?

                       If the values I need are there, maybe what you need is a dictionary rather
                       than just a term list. If not, please add the terms. thanks
Peabody Awards         Term lists need updates. Now that I am trying to actually use PBCore to
Collection Archivist   catalog our collection, I am stumped repeatedly. I have, for one particular
                       title (originally on 16mm) a dupe sound negative, a dupe picture negative,
                       and a duplicate preservation print. I can't find anything in
                       format/generations to accommodate this.

Peabody Awards         Working with more recent items, I have a Website that is an html file
Collection Archivist   saved on a DVD. There is no option for format/physical for something
                       like this.


Peabody Awards         I have a CD with logos in .jpg format and credit sheets and biographies in
Collection Archivist   .pdf format. Again, there is no format/physical for a disc containing
                       anything other than audio or video, and no way to address multiple file
                       types with format/digital.


Peabody Awards         As compared to video, audio options on format generations are severely
Collection Archivist   limited. There is no plain "audio master" or "listening master" or
                       "preservation master", for example.
                       Relevant URLs:
Peabody Awards         Here is an excerpt from a bibliographic record for "The Autobiography of
Collection Archivist   Miss Jane Pittman."

                       511 Cast: Cicely Tyson (Jane Pittman), Odetta (Big Laura), Josephine
                       Premice (Madame Gautier), Michael Murphy (Quentin Lerner).

                       PBCore does not provide a way to include the role played by each cast
                       member; this is not irrelevant information. Here is one solution:

                       -<pbcoreContributor>
                       <contributor>Tyson, Cicely<contributor/>
                       <contributorRole>Actor<contributorRole/>
                       <contributorRoleElaboration>Miss Jane
                       Pittman</contributorRoleElaboration>
                       </pbcoreContributor>

                       -<pbcoreContributor>
                       <contributor>Odetta<contributor/>
                       <contributorRole>Cast<contributorRole/>
                       <contributorRoleElaboration>Big Laura</contributorRoleElaboration>
                       </pbcoreContributor>

                       etc. etc.

Peabody Awards         Please ALSO note that "Actor" is not included in the drop-down menu for
Collection Archivist   creators. Without Cicely Tyson, this production would NOT exist as we
                       know it. Should be possible to include her as creator, not contributor.

                       While my solution may not be the best one, this IS a problem that needs
                       solving.
Peabody Awards         Currently, in the containers for creator and contributor, you can't repeat
Collection Archivist   the creator/contributor role element. You can repeat the container, but
                       this can be messy.

                       From a recent record:
                       executive producer, producer, writer, director, Eric V. Tait, Jr.
                       leads to this:

                       <pbcoreCreator>
                       <creator>Tait, Eric V.</creator>
                       <creatorRole>Executive producer</creatorRole>
                       </pbcoreCreator>

                       <pbcoreCreator>
                       <creator>Tait, Eric V.</creator>
                       <creatorRole>Producer</creatorRole>
                       </pbcoreCreator>

                       <pbcoreCreator>
                       <creator>Tait, Eric V.</creator>
                       <creatorRole>Writer</creatorRole>
                       </pbcoreCreator>

                       <pbcoreCreator>
                       <creator>Tait, Eric V.</creator>
                       <creatorRole>Director</creatorRole>
                       </pbcoreCreator>

                       Why not this:

                       <pbcoreCreator>
                       <creator>Tait, Eric V.</creator>
Manager, Strategic     <creatorRole>Executive
                       PBCore is many fine things, but terse it is not. In developing our PAIS
Technology             project (HD Radio tagging for visually impaired listeners), I looked at
Applications,          using the PBCore structure for the basis of the tagging and determined
NPRLabs                that it was too verbose, and useful tags could not be used in the time-
                       sensitive, limited bandwidth available in the HD Radio Program
                       Associated Data (PSD) channel.
                       My request is this:
                       Create a terse subset of PBCore that is specifically designed for
                       transmission in bandwidth challenged environments, such as HD Radio
                       PSD channels, and FM RDBS streams (even more bandwidth limited).
                       This subset would allow me to get the data to the receiver quickly, and
                       could have hooks to allow the listener to access the richer PBCore trove
                       for that data item using other [internet-based] means.
Peabody Awards         Currently, PBCore is inadequate for dealing with a work whose duration
Collection Archivist   is such that it is split between multiple carriers. An example would be a 2-
                       hour made for TV movie from the 1970s that is split between two 3/4 inch
                       U-Matic tapes. Analog and digital copies of the original may also be
                       divided into two parts, or may have been edited together for storage on a
                       single analog carrier or as a single digital file.

                       PBCore is currently structured upon the presupposition of a 1-1
                       correspondence between intellectual content/property metadata and
                       instantiation metadata. There is nothing to distinguish an instantiation
                       that holds part of the content of a work to one that holds the entire
                       content. In the event of several instantiations that each old part of the
                       work and together hold the entire content, there is no way to show the
                       relationship between the parts.

                       It would seem that this could easily be resolved by nesting multi-part
                       instantiations in an additional element container.

                       I know that I am not the only would-be PBCore user who has struggled
                       with this problem. Your solving it would remove a significant obstacle to
                       more widespread PBCore implementation.
Senior Consultant      yay the change request form!
                       This is mainly to test this out, but I think updating the data dictionary to
                       support a recommendation for ISO 639-3 in the language elements would
                       be helpful for some more multi-lingual applications of pbcore. Also the
                       data dictionary's recommendation for handling multiple languages
                       (eng;fre) is not valid within the schema.


Senior Consultant      re: language (see 12a) Additionally in the schema, I think it may make
Metadata Specialist    Modularity of PBCore if an empty language tag was supported by the
                       PBCore easier to use Content Classes. There are often circumstances
                       where I have needed to use only the Instantiation class, but not the
                       Intellectual Content or Intellectual Property (these are served by other
                       standards such as Dublin Core or MODS).

Metadata Specialist    (see 13a) On the reverse, there are situations where I'd like to only use
                       the Intellectual Content and Intellectual Property elements to describe
                       content at a level where there aren't physical/digital items. An example
                       would be the description of a Series, which isn't embodied Instantiations
                       (while Episodes of that Series are). This could be achieved by changing
                       the requirement for elements such as formatLocation from Mandatory to
                       Mandatory if Applicable (if using Instantiation, for example), and adding
                       the ability to attribute identifiers to content classes.

Metadata Specialist    Support for segment-level description. Currently all descriptive metadata
                       is limited to the asset level. It is important to be able to describe
                       segments of an asset and relate those to the whole, rather than have to
                       create new PBCore records for each segment (and have flimsy
                       relationships between segments and assets).
Communications           Request adding the following vocabularies for genres:
Director
                         Alternative
                         Blues
                         Children's
                         Classical
                         Country
                         Eclectic
                         Electronica
                         Experimental
                         Folk
                         International
                         Jazz
                         Latin
                         R&B and Soul
                         Rap and Hip Hop
                         Rock

                         This list was derived from looking at the common denominator terms from
Archivist                Improve controlled vocabulary for formatStandard and formatDigital, with
                         better distinction between what goes in each element. Also, there are
                         some values in formatDigital that are confused with values that should be
                         in formatEncoding.
Archivist                A way to handle instantiations that have more than one part (not essence
                         tracks). We have many instantiations like this due to the way our media
                         is captured (i.e. software creates multiple files in succession as it
                         captures content. All belong to same instantiation of a single intellectual
Archivist                More granular and machine-readable rights elements. E.g. an element
                         similar to METSRIGHTS Rights Category would be helpful for flagging
                         content.



Director of Archives - The field for 'Frame Rate' could simply be a generic term: 'speed' or you
New York Public        could have a specific field for 'Analog Audio Speed' that would include
Radio (WNYC-           speeds for both disc and tape formats. Discs would have the option list
WQXR)                  of being RPM (revolutions per minute) at 16 2/3, 331/3, 45, 78, and
                       Other. Tape would have the option list of being IPS (inches per second)
                       at 15/16, 1 7/8, 3 3/4, 7 1/2, 15, and 30.
Director of Archives - Additionally - the GENRE drop down list is bloated with specific sports -
New York Public        not generally a public broadcasting focus. I would eliminate the specific
Radio (WNYC-           sports (i.e. baseball, figure skating, etc..) and just have "Sports" .
WQXR)                  Additionally it would be helpful to add the following to the GENRE list:
                       Unknown; Station Identification; Sound Effects; Public Hearing; Oral
                       History; Mood Music; Lecture; Soundtrack (film or filmstrip); Field
                       Recording; Concert Rehearsal; and Advertisment or Commercial which
                       would refer to conventional commercials as well as political or campaign
                       ads.
Director of Archives - It would also be useful to have "Broadcast Script" or "Document" referring
New York Public        to a media item that may no longer exist but the original paper record or
Radio (WNYC-           script of the program remains.
WQXR)
Director of Archives - Forgive me if I'm repeating myself here. I would really suggest that you
New York Public        folks may want to review the audio side of the IASA cataloging rules
Radio (WNYC-           manual. I don't know if it has been updated, I have a 1999 edition.
WQXR)
                       If I have not already mentioned it, BWF or Broadcast Wave File needs to
                       be added to the list of formats digital -audio (5.04).

                         If there is no separate listing for analog audio speed, under frame rate
                         (25.18) I would add 15/16ths ips to the audio speed list. While this is not
                         great sound, there are a fair number broadcast logger tapes around that
                         were recorded at this speed by stations for legal purposes and FCC
                         requirements.

                         Under "Format Generations" (25.07) it would be good to have
                         "Audio/Preservation Master" and "Audio/Preservation Copy" and
                         "Compilation Copy"

                       Under "Format Standard" (25.08) it would be good to have some listing
                       for analog audio that indicates "lateral" and "vertical" for the disc cutting.
                       Both types of cutting were used for broadcast transcription discs. Both
Director of Archives - 1)    Contributor Role controlled vocabulary should include: Senior
New York Public        Broadcast Engineer; Senior Producer.
Radio (WNYC-
WQXR)                  2)    The recommended formatGenerations controlled vocabulary
                       applies separate vocabularies for each mediaType, so even though
                       ‘Moving Image/Preservation Master’ is a recommended value there is no
                       equivalent ‘Audio/Preservation Master’.

Director of Archives -   3)      PBCore’s instantiations appear only to support rendtions that are
New York Public          made up of only one tape file. Most broadcast collections have items that
Radio (WNYC-             span multiple tapes or files or parts or segments. For example WNYC's
WQXR)                    Leonard Lopate and Brian Lehrer shows are each two hours long. The
                         first reel or CD or BWF of the program is typically part A or 1 of 2 and the
                         second reel or CD or BWF is part B or 2 of 2. Jonathan Schwartz’s
                         program is four hours long. Each hour is a part or segment.

update language iso PBCore recommends ISO 639-2. Since PBCore made this
639                 recommendation there is ISO 639-3. Can we update the recommendation
recommendations     to add this much more complete standard. The 639-3 codes are already
                    valid against the regex in the schema.

update language iso Or for extra credit add an authority attribute like MODS has, so that the
639                 authority could be stated and to possibly allow use of other (or no)
recommendations     language standard.

                         This was a use case at Link TV where their music and culture department
                         has some fairly significant language requirements. IOS 639-2 didn't cut it.
                         Here is where we also need to use the semicolon delimited list for codes
                         that the data dictionary recommended but was invalid in the 1.2.1
                         schema.
handling multi -    I probably already posted this, but on nearly every PBCore project I've
component           worked on this has been a major stumbling block. Most of the reasoning
instantiations      on this issue is spelled out here.
                    http://www.pbcoreresources.org/article/instantiations_components_and_e
                    ssence_tracks/
                    Relevant URLs:
                    http://www.pbcoreresources.org/article/instantiations_components_and_e
                    ssence_tracks/
allow more than one sometimes an instantiation has more than one role, it would be helpful to
express of          state more than one formatGenerations. At WNET their prior database
formatGenerations   allowed multiple formatGenerations per instantiation, which make
                    mapping to PBCore challenging.
Archivist           There needs to be a way to break up or label (possibly through
                    attributes?) the pbcoreSubject field to indicate the type of subject being
                    referenced. For instance in MARC 21 Bibliographic, the subject added
                    entry fields in the 600s are different for personal names, corporate
                    names, chronological terms, topical terms, and geographic names, etc.
                    Also, PBS DLL v. 1.2 has something similar within the Coverage fields,
                    where there is a separate field for Coverage Event, Coverage
                    Geographical, Coverage Organization, Coverage People, and Coverage
Exec Dir of         25.08 formatStandard
Engineering         Remove MPEG - MPEG is a compression not a format as NTSC, PAL or
                    SECAM. A video could be MPEG and NTSC. If that was the case the it
                    should have MPEG under fromatEncoding and NTSC unser
                    formatStandard. A user might pick MPEG in both and then we wouldn't
                    know if it is PAL or NTSC. Remove NTSC Progressive
                    I don't think there is anything such format and it would just lead to
                    incorrect data entry.
Exec Dir of         25.08 formatStandard
Engineering         There is no choice for Dolby E under Audio. I'd also suggest that many
                    of the choices are actually compression not format. I'd suggest you move
                    compression of audio to 25.09 formatEncoding and put mono, audio, 5.1,
                    etc in 25.08 formatStandard unless that is dealt with somewhere else
                    then maybe just remove audio from 25.08
Exec Dir of         25.18 formatFrameRate
Engineering         There isn't a choice for 50 fields/sec (25fps) this is used and we convert
                    to it for international delivery.
Exec Dir of         25.03 formatPhysical
Engineering         There isn't a choice for HDCAM SR. I think that should be added.
Exec Dir of         25.16 formatFrameSize
Engineering         There isn't a choice for 1440x1080 which is a common sub-sampled
                    frame size for 1920x1080 that's been compressed. A tape like HDCAM
                    which is compressed at 1440x1080 would still be 1920x1080 but a file
                    that's compressed at 1440x1080 may need to be indicated as such?

Exec Dir of          25.00 pbcoreInstantiation
Engineering          Is there a way to indicate that an asset has been converted from its
                     native format to the current format. We often have two different
                     instances of the same asset one in native NTSC and the other in HD
                     which was upconverted from NTSC. If someone wanted a different
                     format than the current HD format it might be better to go back to the
                     original NTSC to do a "better" upconversion. Just knowing that this HD
                     asset was not native HD but was upconverted from something else would
                     be helpful. The next bit that would be nice is what was the native or what
Exec Dir of          is the GUID of the "parent" native asset so it could be referenced the
                     25.14 formatBitDepth
Engineering          There's no choice for 12 bit. 4:4:4 can be 12 bit.
Exec Dir of          25.11 formatTimeStart
Engineering          I would recommend the name of this change to formatSOM and another
                     descriptor formatEOM be added. Assuming that the EOM will be the
                     same as the duration is inaccurate when describing a scene. Having
                     EOM would also allow for scene or even frame specific metadata which I
                     feel has been missing from pbcore.
Exec Dir of          add field for formatTimecode
Engineering          It may be useful to have a field to describe the format of the timecode:
                     smpte non-drop, smpte-drop, etc.
Exec Dir of          identifierParent
Engineering          identifierSourceParent
                     Consider adding fields to identify the parent of a media asset. There may
                     be a video program that has video and audio. An organization may take
                     the audio track out and now has an asset that's audio only that came
                     from a parent that someone may want to identify.
HearingVoices.com    Hoping PBCore will incorporate the journalistic metadata innovations of
                     the Value Added News project:
                     http://valueaddednews.org/

                     (part of the Media Standards Trust: Transparency Initiative
                     http://www.mediastandardstrust.org/projects/transparency.aspx
                     fronted by Tim Berners-Lee)
                     I'm preparing a major field mapping for the transition to PBCore for a
                     collection of avant-garde and animation titles. The vast majority of these
Managing Director,   titles are very short (from 3-10 minutes), and so our library and study
iotaCenter           center DVD reference collection has units that contain several titles. This
                     creates a problem as I'm fairly sure PBCore does not allow several works
                     to be associated with one instantiation.
R&D Manager          Change Request: Preservation metadata. Currently content descriptive
                     metadata are not dealing with preservation metadata. Usually they are
                     left to the archives and very small steps have been done towards a
                     standardization. Concerning specifically the contents coming from public
                     television environment, I strongly suggest to define a minimal set of
                     metadata for managing the preservation of the content. For example,
                     near "INTELLECTUAL PROPERTY" where "creator, publisher,
                     contributor,..." are defined, we can have a "PRESERVATION
                     DESCRIPTION" where a "preservation holder" can be defined (that could
                     be a broadcaster already preserving the content) or a "master or copy" in
R&D Manager          order to understand what shouldSpeeches. In the case a video content
                     Change Request: Subtitles and be done.
                     has subtitles (as well as in the case an automatic content analyzer
                     creates the STT for a specific audio track in the video), how PBCore can
                     support the text description? There are no elements for this. It should be
                     supported the temporal description of the Text. That will enable the
                     categorization and indexing of the shots inside the video.
R&D Manager          Change Request: Temporal Segmentation

                     The metadata describing the content must support temporal
                     segmentation, in order to express different temporal decompositions.
                     We can have a temporal decomposition describing a specific behavior of
                     the content such as chapters and shots.
                     These decompositions are available in MPEG-7
R&D Manager            Change Request: Rights description

                       Current situation:
                       "PBCore does not allow for storage of an external rights xml standard
                       inline within PBCore.
                       Potentially external rights standards could be used inline within the Rights
                       element if the schema allowed this flexibility."
                       Also the Element "pbcoreRightsSummary" is a flat text element,
                       considering only the ODRL Permissions, Constraints, and Obligations.

                       I suggest to enable have external Rights Metadata, expressed in a
                       whatever format. ODRL could be useful for some simple example of
                       BusinessToConsumer licenses.
                       Also MPEG-21 REL may be bounded to the B2C. In order to support a
                       complex description of B2B (the common behavior of a broadcaster) a
                       more detailed and complex schema is required. We are currently working
                       on MVCO (MediaValueChainOntology) trying to describe "contracts"
                       between broadcasters.
                       There are no "ultimate" standards, but I think that PBCore should enable
                       further extensions and new rights schemas.
                       This could be done allowing an external format/schema linked into a
                       specific element in the "Intellectual Property" class.


Director of Archives - Change Request: I can't recall if this is a repeat, so forgive me if it is.
New York Public
Radio (WNYC-           Format generations - It would be good to have "Access Copy" or "User
WQXR)                  Copy" - We have a large number of duplicate CDs for our future
                       'listening room' for producers and others doing research where they can
                       come and listen to CDs on their own.

                       Format generations - It would be good to have "Compilation Copy" . To
                       save shelf space we used to compile up to ten news features on a single
                       shelf reference/master CD while also maintaining the original story CD.

                       Format generations - Concert production here generally results in to
                       types of tapes. "Mix-Minus" which is the original concert without the
                       broadcast announcer or inserted production interviews and "Mix -Plus"
                       which is what is produced for broadcast with the announcer and
                       intermission production(s).

Director of App.       I would like to see PBCore evolve into a feed-based structure, allowing
Dev., NPR Digital      natively more than one document to be distributed in a single request
Media                  (without the need for other protocols or standards). I have previously
                       suggested a wrapper XML element called that will allow multiple
                       elements to be bundled inside of it.
Director of New      On the topic of Linked Data, the idea is you could have a Contributor
Media & Innovation   named “Hughes, Langston,” or “Smith, John” where the common name
                     may or may not be ambiguous or unique. But if you add an actual UID
                     to the name, it would be unambiguous and machine-readable. The
                     Linked Data way of doing this is to include a URI for the person. This
                     makes it possible to traverse the universe of objects containing this
                     URI, and it links our PBCore object to that universe. We could
                     optionally add URIs to place names, organizations, subjects, and
                     relationships in the PBCore record. The way to allow this would be to
                     add optional attributes to the PBCore schema, so you could include
                     UIDs inside elements. I wouldn’t make it mandatory though, because
                     many PBCore users might not be ready to go to the effort of adding
Director of New      The recommendation of adding Profiles seems very important to me.
Media & Innovation   For example, a PBCore record conforming to a COVE profile would be
                     able to validate against COVE metadata requirements. The COVE
                     metadata requirements include a character-count limit on Title and
                     Description fields, and a specific COVE taxonomy. If I could feed COVE
                     a PBCore record that COVE knows is COVE-compliant, I wouldn’t have
                     to enter metadata into COVE by hand which is the current workflow.
                     You could specify a Profile by adding a namespace attribute in the
                     PBCore document root, and attributes inside the relevant elements. So
                     COVE would need a schema to validate the document, but that’s easy.
                     I’m guessing Public Media Publisher or whatever is coming down the
                     pike could function in the same way. PBCore documents based on
                     Profiles would remain valid and parse-able by applications that don’t
                     care about Profiles. All we’d be doing is adding functionality for
                     applications that do know about them and make it easier for people
                     like me at stations that need to feed the same content to different
                     systems. Attributes would have to be allowed by the PBCore schema to
                     make any of this possible. Then you could unambiguously identify a
                     wide number of values, include URIs for elements like subject,
                     contributor, and identifierSource, ease PBCore into the Linked Data
                     universe, and enable the construction of usable (and machine-
                     readable) profiles.

                     NPR 2.0 Requirements (NPR updated 7/27/10)
Librarian            Contributor (Element 16) / Named entity w/ role NPR’s librarians
                     often cite the name/role metadata as the most valuable we collect.
                      The metadata gives us the ability to locate tape quickly on deadline
                     because we capture the name of the newsmaker with a code for the
                     context (direct interview, public statement, singing, reading, acting,
                     about, etc.) This metadata supports a very specific use case
                     though, that we don’t expect to be replicated elsewhere. But would
                     this metadata be useful to the American Archive? I think the answer
                     is yes. Now the question is how to harmonize the metadata with the
Librarian   Coverage (Element 7) / Geography
            PBCore’s Coverage element allows metadata to be collected that
            pertain to either spatial or temporal characteristics. I believe this
            approach may limit the American Archive’s ability to make full use
            of the power of both geographic and temporal metadata. It’s easy to
            imagine users of the American Archive wanting to locate all content
            about a specific location, or to use this metadata to generate
            visuals, such as maps, etc. Or, users might desire a cross section
            of stories about the 1950s, or 30s. It would be theoretically possible
            to locate relevant content in response to these queries using the
            current schema, but I think separating the two characteristics into
            separate elements will make it easier to collect cleaner metadata,
            and give greater flexibility in how the metadata is leveraged.
            [HS update: Request to treat Geographic and Temporal metadata
            separately stands. Allowing a linkage of the two would give even
            greater power to researchers, as Dave Rice reports was supported
            in previous releases. ]
Librarian   Genre (Element 5) / Format
            Thanks for referring me to the documentation on this one. The
            definition for PBCore’s Genre element states that it reflects the
            “structure of the presentation, as well as the topical nature of the
            content in a generalized form.” This is another place where I think
            overall searchability and metadata collection could benefit from a
            separation of two characteristics – and it seems that the creation of
            two separate controlled lists would also be fairly straightforward. I
            would be interested to hear the arguments though, for keeping them
            together, as I’m sure that there were use cases that lead to this
            reality.
            NPR has a format field that I wasn’t sure would appropriately map to
            the Genre element. We’re capturing metadata about the structural
            presentation in that field.
            [HS update: Our discussion of 7/14 provided helpful explanation of
            the current state of the Genre picklist. Seems the American Archive
            will want to define a specific set of useful genres for its purposes.
            NPR expresses structure of presentation metadata, both at the
Librarian   Parent & Child metadata (in sequence)
            I’m not sure how to use PBCore to indicate the hierarchical and
            sibling relationships between stories and with the overarching
            series. The RelationType element looks like the closest match, but
            I’m still not sure where in the schema the notion of sequencing of
            story parts can be rendered. In thinking about the American Archive
            use case, I can imagine a user locating a story, and then wanting to
            see a listing of other stories from the same program (in sequence).
            Does the Relation element capture enough to be able to satisfy this
            use case?
            [HS update: Request stands: relates to Digital Dawn REQs in S02.
            In a newsmagazine program, sequence of stories is important. We
            would like to be able to leverage the story segment number as
            numerical metadata, so that the number retains its intelligence and
Librarian   Unique Sound
            Historically, NPR hasn’t been able to retain copies of the raw
            materials that go into production. So, the assets we’re describing
            are the finished product. Our unique sound field is used to describe
            sound that comes from an external source and that could be re-
            used in the future. I consider it to be our substitute for cataloging
            raw elements. This is an important searchable field for us, and I
            think it would be even more important to be able to search this
            metadata in the American Archive because of the descriptive nature
            of the metadata in this field. Not sure how to map it though.
            [HS update: Relates to Digital Dawn REQ S09. We could use an
            Extension element here except that Dave Rice advises the extension
            doesn’t support paired data elements. Unless there is better
            support for segmentation, we will lose some data by using the
            extension element, but at least descriptions of clips within
Librarian   Transcripts
            NPR transcripts are full-text representations of our news stories,
            and can be useful to locate exact quotations. I don’t think I’ve ever
            seen a standard schema that has a field for “transcript,” but since
            the American Archive isn’t fully designed yet, I thought I should ask
            the question of how transcripts might be handled. Is there a vision
            for transcripts?
            [HS update: I now understand that transcripts are intended to map
            to the Description element. Good. Dave Rice advises that html and
            other format tags are not supported, so we would lose some textual
            intelligence, but we could probably write an ingest script to format
            the transcript correctly on PBCore ingest.
            Digital Dawn’s REQ S05.02 should make even clearer the intended
Librarian   Rights
            We don’t have one specific way to collect information about rights
            at this time, but were we to add a field for it, it would need to collect
            metadata at the item or story level. Also, it would need to be
            searchable according to controlled vocabulary, if at all possible, in
            order to define what users with certain permissions may access.
            [HS update: Nothing to add here. NPR practice is still evolving.]

                        WGBH 2.0 Change Requests
            The way content collections are handled, which is critical to lay the
            groundwork for a central archive of public broadcasting content.




            The current format elements need additional work: they currently
            apply to both video and audio without distinction. When creating
            metadata for an asset that has both video and audio there is no way
            to repeat the format elements within an instantiation. A restructure
            of the format elements in the XSD would allow for the repeat of
            some elements with the parent container of audio or video or others
May need to add an overall formatdatarate for the complete asset
(one that includes a video and and audio track) as we will have
separate elements for video formatdatarate and audio
formatdatarate.
Need a place to indicate the date the “intellectual content” was
made: for example, “Gone With the Wind” filmed in 1936. Takes
place in 1865. So coverage.temporal = 1865. Our instantiation is a
VHS version, and the VHS was made in 1990. then dateCreated =
1990. But where do you indicate that the “intellectual content” was
made in 1936?
Need a good way to represent instantiations consisting of more than
one physical item. For example, we have video of an event on
campus, shot in Betamax. The master spans 2 Betamax cassettes.
However, there is nowhere to indicate that the Betamax/master
How to represent instantiations that are identical copies?

Version 2.0 should take a much more structural approach. Two big
issues to address: Recursive description for components of an asset –
the “structure map” issue. The need to support recursive, linked
descriptions for subsets of a production is an issue you are already
considering. The cleanest way to approach it would be to create a
second schema, “pbcoreStructureMap” and change the current
schema to pbcoreDescriptionDocument”. The current Schema 1.0
would become the pbcoreDescriptionDocument and Schema 2.0
would reference both the pbcoreStructureMap and the
pbcoreDescriptionDocument schemas Schema 2.0 may include
structureMap and descriptionDocument subschemas and define
how they relate to each other, a la the METS document schema, or
it may create the structureMap within the 2,0 schema by
PREMIS. Another next step would be to map to the PREMIS schema
and probably extend the Rights and Instantiation elements to insure
that required elements are supported. There aren’t that many
required elements, but there are some elements specific to
preservation thatAare currently lacking in PBCore. is dynamically
(Cataloging tool) data entry list for "Subject" that
contingent on the value of "subject authority".

(Cataloging tool) A comprehensive guide to understanding the built-
in report utility. While it's largely an optional feature, a FileMaker
developer could customize its output.
(Dave MacCarn): PBCore needs to have extension structure analyzed.
Concerned that we not turn PBCore into MPEG 7, or SMPTE metadata
dictionary – way too many fields. Could poll for controlled vocabulary
(many requests are for CV changes, not elements).
(Steve Snyder, Catherine Murphy/The World): Metadata is simple and
manual. Tag in iTunes and Public Interactive's software. Metadata not
searchable in web search. Use Audio Vault (similar to Dalet), and AP's
ENPS for description. Ideal: to have automatically generated RSS feed
(currently manually updated).
(Madeleine Rothberg/NCAM)- Accessibility/ISO Access for All
considerations (based on XML, moving away from container to flat
based structure). Metadata about modalities of content access (screen
sizes, font colors, how to control, etc.) Teachers Domain.




(Eli Ingraham/Forum Network): Needed culture and identity, region of
world but wasn't in PBCore. Reduced high level topic categories (from
PBCore & 10 others incl. TV Guide) from 20 to 10. Created
intermediate layer of a preset tag layer from PBCore/PBS/NPR
metadata to ensure reliability.
The schema should not require a particular sequence of element tags.
This is unusual in my experience for schema, and I believe it
unnecessarily increases the complexity of creating valid PBCore XML. I
also don't see any upside to people who are writing PBCore parsers to
have the tags in a particular order. I believe most parsers simply read
Ability to describe a data object which may no longer preserve
the XML and create elements: interview, b-roll, stock footage the




Episode # - where to put this? Is it a title type? Can it be an
identifier? (they include it in EBUObjectTypeCode but as subset?)
 I want to repeat Identifier using NOLA code and any other id I
want to use. FormatIdentifer is more important/stable to me... Why
can’t I repeat identifier w/in an instantiation?
Enhance role vocabularies with audio roles – see EBUCore and
DAM
Creator/contributor/publisher and roles – how to differentiate between
person and organization? Add contributor type? What if you want to
express “Elizabeth Deane of WGBH as Producer”? How to do this?




FormatDigital and FormatEncoding – streamline these vocabs –
confusion and overlap between file types, codecs and mimetypes
LC released MODS version 3.4, and added a couple things we might
want to parallel in PBCore 2.0

Functional changes
Add the "displayLabel" attribute to all top-level MODS elements that now
lack it. Rationale: makes MODS overall more consistent, and promotes
easier use of MODS as a back-end format for customized cataloging and
discovery systems. (March 11, 2009)

RDA-related changes
Add authorityURI and valueURI attributes to elements that have authority
attribute (October 14, 2009)
Add authorityURI and valueURI attributes to subject subelements
(October 14,
2009)

    Digital Dawn/AVPS/WGBH 1.3/2.0 Grid
01.01 Definition of an “Asset”
Allow a means to define what the “asset” is intended to depict.
PBCore 2.0 should begin by offering a more granular definition of
what is meant by an “asset”. The “Asset” definition should not be
viewed as a passive definition driven by current content practices.
Rather, “asset” definition should be a conscious decision to define
the digital building blocks that will meet Public Media’s business
objectives.
Public Media generally talks in terms of “episodes” as its
fundamental asset, however, Public Media’s business objectives
clearly call for a much m ore granular asset categorization that
allows episodes to be de-assembled into audio tracks, closed
S01.02 Varied Content Types
Support a wider group of Asset content types
PBCore 2.0 should also support a wider understanding and
definition set for content type, specifically content types should:
• Be flexible enough to support methods for defining new “content
type profiles”
Include some of the common content types that are used in Public
Media, including but not limited to:
o TV: series, episode, clip,
o Radio: show, story
Type: Schema Change
S01.03 Abstract Assets
Support the definition and creation of “abstract” assets – assets
which do not include a corresponding instantiation or are only
represented as temporal portions of another asset’s instantiations
without their own discrete instantiations, such as segments or clips
or episodes.
Support for abstract assets is important for re-purposing segments
of existing content into new products. Definition of Public Media’s
foundational asset types is an important step in formulating how the
Public Media information architecture will address abstract assets
that have no instantiation (for example “series”). Once the
foundational asset types have been defined different approaches
can be evaluated for packaging abstract assets.
One approach for creating abstract assets is through the use of
XML-based wrappers like ATOM (see below).
S1.04 Feed Creation
Gain the ability to package PB Core into a feed or XML wrapper
Users of PB Core should be able to package metadata into a feed or
XML wrapper to facilitate transfer between systems and
organizations. One approach could be through the use of wrappers
like ATOM.
ATOM provides a supported standard that can be used to package
and distribute Public Media assets. The Atom Syndication Format is
an XML language used for web feeds, while the Atom Publish
Protocol (AtomPub or APP) is a simple HTTP-based protocol for
creating and updating web resources. The advantage of using this
approach for Public Media is in sustainability (a large community
shares the cost of maintenance of standards) and compliance with
best practices (Public Media avoids the risk of developing a custom
standard that is not reflective or compliant with broader industry
S02.01 Asset Relationships
Enable Relationship of Assets Outside of Field Entries
Improve relationship handling and expression between assets. This
should occur in a way that allows for and encourages support for
inheritance from one asset to others.
Please see note on XML-based wrappers in S01.03, as one potential
solution.
Type: Schema Change
S02.02 Relationship Definitions
Define Relationships for Content Types
Develop a mechanism to define the relationships that a content type
might have (ex. a Series will have Episodes, and Episode may have
Clips).
Type: Schema Change
S03.01 Applying containers to an instantiation
An Instantiation may have specific rights, relation or title
information that applies only to it and not to the entire asset.
Currently containers cannot be applied to Instantiations. A structure
is needed in order to relate Instantiation-specific metadata.
This issue should be addressed through the expansion of content
types (S02.01) and relationship definitions (S02.02), which will foster
structures that can be used to relate containers to instantiations.
Type: Schema Change




S04.01 Source ID
Ensure that Source ID fields (identifierSource, subjectIdentiferUsed,
etc.) are retained through implementation and Source ID values
retained through workflow
Source IDs are essential to understanding identifiers, descriptors,
and identifying related assets. PBCore records that rely on multiple
sources for their values must have an accompanying field for
Source ID that is filled and carried between systems.
Type: Implementation Workflow
S04.02 Multi-Source Identifiers
Source Identifiers from Multiple Sources and Terms from Multiple
Controlled Vocabularies
Ensuring that repeating data sets can be applied to single assets
(S03.01) will support structurally the requirement for expressing
both a term/unique identifier with an accompanying field defining
the source of the term/ID. User training may be required to ensure
consistent application of both elements in a record.
Type: Implementation Recommendation
S04.03 Multiple Controlled Vocabularies
Support terms from multiple controlled vocabularies.
PBCore 2.0 should support multiple controlled vocabularies in a
structured manner. Specifically, it should allow users to define a
“profile” or “extension”, which relates its metadata sets to a
specific controlled vocabulary and mappings to PBCore.
Provide a means for the community to nominate new terms for
PBCore’s recommended controlled vocabulary. Provide additional
guidance for the selection or refinement of PBCore’s existing
controlled vocabularies to meet specific use cases. Often the full
PBCore vocabularies are too extensive for specific use, divided
existing lists by use cases may allow users to apply lists more
relevant to their application.
S05.01 Genre Element Set
Revise the Genre controlled vocabulary to become more user
friendly and expand documentation and examples.
The purpose of Genre element set (per PBCore User Guide) requires
review and revision, specifically:
• Clarify the use of Genre in the schema and the PB User Guide.
Currently the User guide explains that the element set can be used
for both topical descriptors and editorial format descriptors. This
leads to confusion.
• Remove outdated descriptors from picklist
• Add new terms added as needed by users and by the use cases
defined above.
S05.02 Description Element Set
Revise the Description element set to be more user friendly.
The description element set requires review in terms of its original
design and the current practices of users in filling it. The number of
acceptable descriptions should be carefully analyzed with an eye to
evaluating whether the description types available to users should
be limited, whether Description should be divided into separate
fields for different types of Descriptions, or whether to maintain the
current Description element set and apply more effective user
training.
S06.01 Rights Management
Define an approach for rights management.
PBC 2.0 should include a more standard approach to rights
management, either through PBC itself (similar to the discussion of
“packaging” above) through and open standard for rights
management. A few specific requirements are important to note:
• Rights metadata should apply to assets at all levels of granularity,
from Clip to Series. This becomes especially important when
repurposing content.
• Once an abstract asset has been defined for a particular
distribution purpose (for example, assembling music and clips from
previous broadcasts to create a new asset) the rights should be able
to be defined for the new asset – as opposed to attempting to define
rights within a series of individual PBCore records.
• Apply a standard vocabulary to rights so that it can become a sort
able field to facilitate accessibility and decision-making.
As a note, one candidate approach for modeling digital rights is the
Open Digital Rights Language (ODRL) for standardizing the
expression of rights information over content. ODRL provides
flexible and interoperable mechanisms to support use of digital
resources in publishing, distributing and consuming of electronic
publications, digital images, audio and movies, learning objects,
computer software and other digital creations.
Type: Schema Change




S07.01 Linked Data
Supported Linked Data to supplement metadata from structured and
unstructured fields.
PBC 2.0 should make some accommodation for supporting linked
data to harvest information from digital sources. This topic requires
additional exploration and definition, as well as more specific
requirements.
Type: Schema
S08.01 ProfilesChange
Support the use of profiles as a mechanism for extending PBC to
customized scenarios.
PBC2.0 should allow extension of the schema to custom scenarios
through the use of profiles (“applications”). These should:
• Be used in the case of content providers that ingest metadata into
the American Archive and distribution partners which receive
content from the American Archive
• Allow certain metadata elements to be defined on a per-entity
basis, including but not limited to: Description, element sets,
controlled vocabularies, content types and required fields
Type: Implementation Recommendation
S08.02 Basic APIs
Publish a set of basic APIs for retrieving information from the
American Archive.
Since the American Archive may make its content available to
entities which do not require complex metadata and which do not
have trained metadata experts on staff, the American Archive
should create a basic set of APIs that support a few basic queries in
native PBC format, including but not limited to:
• Retrieving lists of available content with genre, content type
• Retrieving specific assets for display or playout
Type: Implementation Recommendation
S09.01 Segmentation
Define the method for accomplishing segmentation in common use
cases.
Segmentation should be addressed structurally in prior
requirements. However, there are still a variety of options about how
segmentation could be handled. PBC 2.0 should explicitly define the
way to do this in a way that satisfies common use cases, including
but not limited to:
• Creating chapter points within media files
• Creating clips from media files
• Supporting cataloguing of time based metadata, such as bottom
thirds
• Supporting time code based rights management
Type: Schema Change
Relevant URLs//Notes          Urgency                     Category



                              This is a road block for me Vocabulary or Definition




                              This is a road block for me XSD or Schema or Structure

                              This is a road block for me XSD or Schema or Structure


http://www.pbcoreresources.org This would be nice to have Tag or Element,
/article/pbcore_and_the_seman                                  XSD or Schema or
tic_web/                                                  Structure

                              This would be nice to have XSD or Schema or Structure




                              This is a stumbling block   Tag or Element,
                              for me                            Vocabulary or Definition
This is a stumbling block   Vocabulary or Definition
for me




This is a stumbling block   Vocabulary or Definition
for me



This is a stumbling block   Vocabulary or Definition
for me




This is a stumbling block   Vocabulary or Definition
for me


This is a stumbling block   Tag or Element
for me




This is a stumbling block   Tag or Element
for me
                              This is a stumbling block   XSD or Schema or Structure
                              for me




see: The Technical Basis of   This is a road block for me Other
PAIS, (downloadable .pdf)
 http://www.nprlabs.org
This is a road block for me Tag or Element,
                                 XSD or Schema or
                            Structure




This would be nice to have Vocabulary or Definition




This would be nice to have Vocabulary or Definition
This is a stumbling block Tag or Element
for me




This is a stumbling block   Tag or Element
for me




This is a road block for me XSD or Schema or Structure
allmusic.com                 This would be nice to have Vocabulary or Definition
amazon.com
archive.org
audiogalaxy.com
dmoz.org (which is google)
itunes radio
yahoo.com




                             This would be nice to have Vocabulary or Definition



                             This is a road block for me XSD or Schema or Structure



                             This is a stumbling block   Tag or Element
                             for me




                             This would be nice to have Tag or Element,
                                                                 Vocabulary or
                                                        Definition




                             This would be nice to have Tag or Element,
                                                                 Vocabulary or
                                                        Definition




                             This would be nice to have Tag or Element,
                                                                 Vocabulary or
                                                        Definition
                                 This is a stumbling block   Tag or Element,
                                 for me                                 Vocabulary or
                                                             Definition




                                 This is a stumbling block   Tag or Element,
                                 for me                              Vocabulary or
                                                             Definition,
                                                             XSD or Schema or Structure




                                 This is a stumbling block   Tag or Element,
                                 for me                              Vocabulary or
                                                             Definition,
                                                             XSD or Schema or Structure




http://en.wikipedia.org/wiki/ISO This is a stumbling block   Vocabulary or Definition,
_639-3                           for me                                    XSD or Schema
                                                             or Structure


http://en.wikipedia.org/wiki/ISO This is a stumbling block   Vocabulary or Definition,
_639-3                           for me                                    XSD or Schema
                                                             or Structure
                               This is a road block for me XSD or Schema or Structure




                               This is a stumbling block   XSD or Schema or Structure
                               for me


MARC 21 Bibliographic          This would be nice to have Tag or Element
Subject Access Fields:

http://www.loc.gov/marc/bibliogr
aphic/bd6xx.html,
             PBS DLL
Metadata User Guide:
http://metadata.pbsdigitallearnin
http://www.pbcore.org/PBCore/f This is a stumbling block   Vocabulary or Definition
ormatEncoding.html                for me




http://www.pbcore.org/PBCore/f This is a stumbling block   Vocabulary or Definition
ormatEncoding.html             for me




                               This would be nice to have Vocabulary or Definition


                               This would be nice to have Vocabulary or Definition

                               This would be nice to have Vocabulary or Definition




                               This would be nice to have Vocabulary or Definition




                               This would be nice to have Vocabulary or Definition
                                This would be nice to have Vocabulary or Definition




                                This would be nice to have Vocabulary or Definition


                                This would be nice to have Vocabulary or Definition




 http://valueaddednews.org/     This is a stumbling block   Tag or Element
http://www.mediastandardstrust for me
.org/projects/transparency.aspx




                                This is a road block for me XSD or Schema or Structure




                                This would be nice to have Other




                                This is a stumbling block   XSD or Schema or Structure
                                for me




                                This is a stumbling block   XSD or Schema or Structure
                                for me
                                 This is a stumbling block   XSD or Schema or Structure
                                 for me




                                 This would be nice to have Tag or Element




The full recommendation can
be found at
http://www.pbcoreresources.org
/article/pbcore_recommendatio
n_pbcorecollection/.
             http://www.pbcoreresourc This is a road block for me XSD or Schema or Structure
             es.org/article/pbcore_and
             _the_semantic_web/
             http://www.pbcoreresourc
             es.org/P10/




             http://www.pbcoreresourc This is a road block for me XSD or Schema or Structure
             es.org/article/pbcore_and
             _the_semantic_web/
             http://www.pbcoreresourc
             es.org/P10/




dated 7/27/10)
                                         Required by CPB
Required by CPB




Required by CPB




Required by CPB
                                 Required by CPB




                                 Required by CPB




                                 Required by CPB




Similar to "feed" request from   This is a stumbling block   XSD or Schema or Structure
DD




Add another request re         This is a stumbling block     XSD or Schema or Structure
reviewing format pickilsts -
outdated and how to add new
formats (this was addressed in
1.2)
I would think total formatdata    This is a stumbling block   Tag or Element
rate more important than a + v
(this was addressed in 1.2)


add a dateCreated in the          This is a stumbling block   Tag or Element
intellectual content container




Not yet addressed. Either         This is a stumbling block   XSD or Schema or Structure
create 2 records (part 1 and 2)
or add a way to order
instantiations within a record
Not yet addressed. Can't you      n/a
just repeat instances?
Not yet addressed.                                            Other




Not yet addressed.                Nice to have                Other
Create a preservation profile
for PBCore with this mapping
and different required fields.

Not yet addressed.                n/a                         Other - tools


Not yet addressed.                n/a                         Other - tools


Not yet addressed.                Stumbling block             XSD or Schema or Structure,
Review structure, revisit                                     Vocabulary or Definition
controlled vocabs and update.

                                                              XSD or Schema or Structure
Not yet addressed.Write some                      XSD or Schema or Structure
samples for how to use
extensions.

Review use of containers -
make better use of
conventional XML structures.

http://www.iso.org/iso/catalogue
_detail.htm?csnumber=41521

http://www.imsglobal.org/acces
sibility/



Not yet addressed.                 Nice to have   Vocabulary or Definition




                                   Nice to have   XSD or Schema or Structure




                                                  Vocabulary or Definition



Dave M: This can be an             Nice to have
identifier

Dave M: FormatIdentifier is
already a repeatable field
within instantiation.



(DM)
 <pbcoreCreator>

<creator>Deane, Elizabeth
(WGBH)</creator>
              <creatorRole>
Producer</creatorRole>
           </pbcoreCreator>
  <pbcoreCreator>
              <creator>Smith,
John (Nature)</creator>

<creatorRole>Producer</creato
Dave M: We are looking at this.
http://www.loc.gov/standards/m
ods/changes-3-4.html




Asset Type element needed.       Required by CPB   Schema Change
Also improved documentation
and example best practices.
Use cases: Inventory, All




Consider all audiences in        Required by CPB   Vocabulary Addition
adopting, suggesting or
developing asset type
controlled vocabulary.
Use cases: Inventory, All.
Already included-slight change Required by CPB      Schema Change
to XML schema - no change to
dictionary.
Better documentation and best
practices is needed. Remove
instantiation as requirement
within the PBCore schema.
Use cases: DONE.




Application of PBCore.           Required by CPB    Schema Change
Feed is out of scope for 1.3,
but will add example
documentation mapping of how
to do this. Wrapper is in within
scope and detailed elsewhere.
Use cases: n/a




PBCore 1.2 is able to express     Required by CPB   Schema change
these now but will re-visit and
review for missing
relationships. More specific
user scenarios needed. Use
cases: DONE


Already included in schema,       Required by CPB   Schema Change
may need new doc, PBCore
1.2 is able to express this
already, Needs better
documentation and example
best practices. Use cases:
What is the use case here?        Required by CPB   Schema Change
Need clarification. Will
consider for 2.0. Not necessary
for AA Inventory. Record fields
could apply to instantiations,
fields could move, or use
existing annotation element.
Comment: A use case would
be if an asset is represented by
tapeA from local station and
tapeB from PBS. Then the AA
digitizes tapeA to acquire fileC.
In PBCore there needs to be a
means to document the
specific relationship between
tapeA and fileC. It will be
important to know what
instantiations the AA managed
content is derived from. This is
an example of apply
pbcoreRelation to
pbcoreInstantiation (although
this could be addressed with
other means).



Already included in schema,     Required by CPB     Implementation
may need new doc. Issue with                        Recommendation
use of controlled vocabularies.
Improve/add documentation
and best practices. Use cases:
DONE.




Already included in schema,     Required by CPB     Implementation
may need new doc. . Issue with                      Recommendation
use of controlled vocabularies.
Improve/add documentation
and best practices.

Comment: This may be a
schema issues as well. For
instance in a producer one
contributor may have more
than one ‘role’, but PBCore
only allows 0 or 1 roles per
contributor. Use cases: DONE
Already included in schema,       Required by CPB   Implementation
may need new doc. Issue with                        Recommendation
use of controlled vocabularies.
Improve/add documentation
and best practices.Use cases:
DONE.




Update descriptions and           Required by CPB   Implementation
examples. Not a required field                      Recommendation
but we'll enhance
documentation.




Update documentation about        Required by CPB   Implementation
descriptions and vocabularies                       Recommendation
Assess how to link to known     Required by CPB
schemas. Not necessary for
inventory project.
Rights can be handled with
other known schemas – we’ll
look at linking to those
schemas

Comment: This is true, but
managing one metadata
standard is already a burden
for most stations, handling two
standards and the relationship
between the two can be
challenging work. Additionally
there is need to include better
rights metadata in records to
facilitate web access
(stream/download permissions,
creative commons, etc).
Currently there is no
standardized or recommended
means to express rights within
PBCore, just a broad
container. Also PBCore does
not allow for storage of an
external rights xml standard
inline within PBCore.
Potentially external rights
standards could be used inline
within the Rights element if the
schema allowed this flexibility.
Will look at documentation.      Required by CPB   Implementation issue, not a
                                                   schema change.




Documentation, invite           Required by CPB    Implementation
tool/profile development from                      Recommendation
users. Ask for definition of
needed profiles
Will add example of how to do Required by CPB   Implementation
this to documentation.                          Recommendation
Out of scope for PBCore - this
is an American Archive project.




                             Required by CPB    Schema Change
                                          Organization                                  Do you
If other category,                                                                      currentl
                   Organization:          Type:             Location:
explain:                                                                                y use
                   [Organization]         [organizationtype [Location]
[categoryother]                                                                         PBCore
                                          ]
                                                                                        :

                                          PBS Station             55116                 No




                                          PBS Station             55116                 No

                                          PBS Station                                   No


                  Illinois Public Media   PBS Station, NPR        Urbana, Illinois 61801 Yes
                                          Station, Other public
                                          media organization

                  Walter J. Brown         Other organization      Athens, GA            No
                  Media Archives &
                  Peabody Awards
                  Collection




                  Walter J. Brown         Other organization      Athens, GA            No
                  Media Archives &
                  Peabody Awards
                  Collection
Walter J. Brown    Other organization   Athens, GA 30602   Yes
Media Archives &
Peabody Awards
Collection


Walter J. Brown    Other organization   Athens, GA 30602   Yes
Media Archives &
Peabody Awards
Collection

Walter J. Brown    Other organization   Athens, GA 30602   Yes
Media Archives &
Peabody Awards
Collection


Walter J. Brown    Other organization   Athens, GA 30602   Yes
Media Archives &
Peabody Awards
Collection
Walter J. Brown    Other organization   Athens, GA         No
Media Archives &
Peabody Awards
Collection




Walter J. Brown    Other organization   Athens, GA         No
Media Archives &
Peabody Awards
Collection
                       Walter J. Brown    Other organization   Athens, GA       No
                       Media Archives &
                       Peabody Awards
                       Collection




Proper terse subset.                      Other public media   Washington, DC   No
                                          organization
Problem could be       Walter J. Brown    Other organization   Athens, GA     No
solved in several ways Media Archives &
                       Peabody Awards
                       Collection




                      Audiovisual                                             Yes
                      Preservation
                      Solutions




                      Audiovisual                                             Yes
                      Preservation
                      New York University Other organization   New York, NY   Yes




                      New York University Other organization   New York, NY   Yes




                      New York University Other organization   New York, NY   Yes
channelAustin   Other public media   Austin, TX 78741     Yes
                organization




WITNESS         Other organization   Brooklyn, NY 11217   Yes



WITNESS         Other organization   Brooklyn, NY 11217   Yes



WITNESS         Other organization   Brooklyn, NY 11217   Yes




WNYC-WQXR       NPR Station          New York, New York   Yes
Radio




WNYC-WQXR       NPR Station          New York, New York   Yes
Radio




WNYC-WQXR       NPR Station          New York, New York   Yes
Radio
WNYC-WQXR   NPR Station          New York, New York   Yes
Radio




WNYC-WQXR   NPR Station          New York, New York   Yes
Radio




WNYC-WQXR   NPR Station          New York, New York   Yes
Radio




            Other organization                        Yes




            Other organization                        Yes
Audiovisual                      New York, NY      Yes
Preservation
Solutions




                                                   Yes



Louisiana Public   PBS Station   Baton Rouge, LA   Yes
Broadcasting




KCTS Television    PBS Station   Seattle, WA       No




KCTS Television    PBS Station   Seattle, WA       No




KCTS Television    PBS Station   Seattle, WA       No


KCTS Television    PBS Station   Seattle, WA       No

KCTS Television    PBS Station   Seattle, WA       No




KCTS Television    PBS Station   Seattle, WA       No




KCTS Television    PBS Station   Seattle, WA       No
                      KCTS Television   PBS Station          Seattle, WA       No




                      KCTS Television   PBS Station          Seattle, WA       No


                      KCTS Television   PBS Station          Seattle, WA       No




                      Hearing Voices    Other public media   Internet          No
                                        organization




                      iotaCenter                             Los Angeles, CA   Yes




vision and scope of   EURIX Group       Other organization   Turin, Italy      Yes
the format




                      EURIX Group       Other organization   Turin, Italy      Yes




                      EURIX Group       Other organization   Turin, Italy      Yes
EURIX Group   Other organization   Turin, Italy   Yes




WNYC-WQXR     NPR Station          New York, NY   Yes
Radio




NPR
Illinois Public Medi    PBS Station, NPR   Urbana, Illinois 61801 yes
                        station




Illinois Public Media   PBS Station, NPR   Urbana, Illinois 61801 Yes
                        station




NPR
no funding for
tweaking cataloging
tool
no funding for
tweaking cataloging
tool
If yes use PBCore, how:
[howusepbcore]




Our media cataloging system is based on
PBCore.
TRYING (operative word) to use PBCore
as model for our instantiation data.




TRYING (operative word) to use PBCore
as model for our instantiation data.



TRYING (operative word) to use PBCore
as model for our instantiation data.




TRYING (operative word) to use PBCore
as model for our instantiation data.
...cannot for reasons (above)
Preservation Repository descriptive and
technical metadata schema.




Preservation Repository descriptive and
technical metadata schema.




Preservation repository descriptive and
technical metadata standard.
We are adopting PBCore for our
implementation of the Open Media
Project. See www.openmediaproject.org




our catalog is based on PBCore, and
exportable as PBCore XML


Our catalog is loosely based on PBCore,
and is exportable as PBCore (with a
workaround for the instantiation problem)

Our catalog is based on PBCore




We are currently implimenting a PBCore
compliant SQL archive database.




We are currently implimenting a PBCore
compliant SQL archive database.




We are currently implimenting a PBCore
compliant SQL archive database.
Dave Rice just launched our new SQL
database.




We are currently implimenting a PBCore
compliant SQL database.




We are currently implimenting a PBCore
compliant SQL database.




for describing assets, what else?




for describing assets, what else?
We first used PBCore during the
American Archive Pilot Project and plan to
continue using it as our in-house archive
project develops.
Working on mapping current collections
database to PBCore. Will likely use it as a
data structure standard (i.e. not use the
cataloging tool) in conjunction with
CollectiveAccess, so we can customize
the installation profile more easily.
Research




Research




Research
Research




New PBCore compatible catalog now in
service.
PBCore is our core standard for media
cataloging




PBCore is our core standard for
media cataloging
 Chg     v1.                  WGBH PBCore Team
             v2.0 v3.0 n/a                                   Name
Req. #    3                     Evaluated Type



  5           x              same as 2nd half of 46, and Mary Miller
                             same as 72. Schema change.




 8a           x              Schema Change - may add         Mary Miller
                             optional annotation attribute
                             to string elements
 9    x   Similar to 46a. Possible        Mary Miller
          schema change, parsing
          challenge




19a   x   See                           Andy Lanset
          http://pbcore.org/PBCore/form
          atFrameRate.html. Will
          consider a field name change.



24    x   Schema change                   dave rice




25    x   Schema change and               Leslie Bourgeois
          documentation. Thinking of
          adding 3 attributes for 2.0 -
          source, version, and
          annotation. Need to structure
          annotations, or add type
          attribute.

31    x   Could use annotation field. Or Jabran Soubeih
          create specific relationship
          type. Either solve or create
          example of how to do this in
          documentation.




43    x   similar to 78. Schema change Daniel Jacobson
          or a documentation/use case
          solution
46       x   Similar to 5.                 Hannah
                                           Sommers




53   x   x   Similar to 78. 1.3 provides the WGBH -known
             mechanism to solve this.         needs
             Need to define a collection
             profile and/or create
             documentation. Review
             discussions on how to create
             collections and feeds. May
             use METS wrapper or Jack
             Brighton's collection structure.

63       x   rework extension structure.   WGBH (Dave M)
             Need either more fields in
             extension or attributes.

67       x   Schema issue DM and PP        WGBH - Peter P
             need to work out. Add
             structure or not add
             structure.....


72       x   similar to 5 and 46 Could     WGBH -
             add field for affiliation.    Inventory




78       x   Similar to 43. Schema
             change or a
             documentation/use case
             solution
81        x   Rights per instantiation.




 1    x   x   Controlled vocab with source   Bruce Jacobs
              reference. This could be a
              PBCore picklist.



 6        x   Already included in 1.3.       Mary Miller
              Documentation update
              needed. Need to review
              controlled vocabulary.




7a        x   Controlled Vocabulary review Mary Miller
              - formatGenerations




7d        x   Controlled Vocabulary review Mary Miller
              - formatGenerations


12b       x   Need clarification about       Dave Rice
              empty tag issue. May be
              schema change
15    x   Add radio genres to          Stefan Wray
          vocabulary for genre




16    x   Similar to #1 and #6. Need to Yvonne Ng
          review these cvs.
          (formatStandard is PBCore
          1.1)
19b   x   Review genre controlled       Andy Lanset
          vocabulary




20    x   Review controlled            Andy Lanset
          vocabularies for sound.
21a   x   CV review                    Andy Lanset




26    x   Done but review controlled   Jabran Soubeih
          vocabularies. Update
          documentation




27    x   Done but review controlled   Jabran Soubeih
          vocabularies. Update
          documentation



28    x   Review frame rate CV         Jabran Soubeih


29    x   Review formatPhysical CV     Jabran Soubeih

30    x   Review framesize CV          Jabran Soubeih




32    x   Review formatbitdepth cv     Jabran Soubeih

36    x   will review for controlled   Barrett Golding
          vocabulary input.




42    x   Controlled Vocabulary review Andy Lanset
          - formatGenerations and
          contributor
48   x   controlled vocab review         Hannah
                                         Sommers




66   x   controlled vocabulary issue -   WGBH
         stripped down subset of
         vocabuary, limited. Also a
         use case - how to develop a
68   x   controlled vocabulary issue.    WGBH-MLA
69   x   Either an asset type or in
         controlled vocabulary issue.    WGBH -
         Either an asset type or in      Inventory
71   x   another element.
         controlled vocabulary           WGBH -
73   x   Same as 26, 29. Controlled      Inventory
                                         WGBH -
         vocabulary review and           Inventory
         documentation



2    x   Similar to 49. Address in       Bruce Jacobs
         documentation and training
3    x   similar to #2                   Bruce Jacobs


6    x   Already included in 1.3.        Mary Miller
         Documentation update
         needed. Need to review
         controlled vocabulary.
10        x   use minimal required fields    Rich Rarey
              for PBCore, add extensions
              for domain specific data.
              Could also use a style sheet
              to transform PBCore
              metadata. Use case and
              training issue.




11    x   x   Can make an annotation on       Mary Miller
              instantiations now. Addressed
              also in 21, 37 --- This is
              really 2 Questions: 1. how to
              record 1 program that spans
              multiple tapes (part 1, part 2,
              etc.), 2. need clarification -
              does she mean multiple types
              of assets for one work or
              multiple parts for one work?
              Whichever one, we need
              better documentation and use
              case because this is now
              possible with 1.3. Same as
              #23.




12a       x   Update documentation,          Dave Rice
              include ISO 639-3 as a
              language recommendation.
              Need clarification about
              empty tag issue.

14        x   Solved by 1.3. assetType    Kara Van
              allows description of       Malssen
              segments. Also format
              timestart and duration
              describe the segment. Need
              new documentation and
              training.
17    x   x   Same as 11. Need better     Yvonne Ng
              documentation and use cases


18        x   Will give examples for rights - Yvonne Ng
              1) allowing for a link to
              external document and 2)
              American Archive use case
              example spelling out use of
              rights schema as extension
19c           x   Can be done with                Andy Lanset
                  formatPhysical. Documents
                  can be instantiations. (include
                  as example in documentation)

21b   x   x       Same as 11. Need better     Andy Lanset
                  documentation and use cases




22a       x       same as 12a. Update             dave rice
                  documentation




23        x       Same as 11. Need better     dave rice
                  documentation and use cases




25        x       Schema change and               Leslie Bourgeois
                  documentation. Thinking of
                  adding 3 attributes for 2.0 -
                  source, version, and
                  annotation. Need to structure
                  annotations, or add type
                  attribute.


26        x       Done but review controlled      Jabran Soubeih
                  vocabularies. Update
                  documentation




27        x       Done but review controlled      Jabran Soubeih
                  vocabularies. Update
                  documentation
31   x   Could use annotation field. Or Jabran Soubeih
         create specific relationship
         type. Either solve or create
         example of how to do this in
         documentation.




39   x   Update documentation - use     Walter Allasia
         essenceTrackType.




40   x   Similar to #2 - segmentation   Walter Allasia
         issue. Use assetType from
         1.3. Documentation needed.




41   x   Same as #18                    Walter Allasia




43   x   similar to 78. Schema change Daniel Jacobson
         or a documentation/use case
         solution
49       x   similar to #2. relation defines   Hannah
             relationship but need to          Sommers
             research sequencing
             capability.




51   x   x   Transcripts can be a             Hannah
             description type. Possible to Sommers
             insert html into a pbcore
             string using escape
             characters. We should
             provide documentation on
             how to do this. You can also
             create a pbcore record for
             your transcript and relate it to
             the media asset. WGBH does
             with TEI and SMIL to
             synchronize transcript to
             media.
52       x   Same as 18 and 41                Hannah
                                              Sommers




53   x   x   Similar to 78. 1.3 provides the WGBH -known
             mechanism to solve this.         needs
             Need to define a collection
             profile and/or create
             documentation. Review
             discussions on how to create
             collections and feeds. May
             use METS wrapper or Jack
             Brighton's collection structure.

54       x   This is addressed in 1.2.         WGBH -known
             Need to document                  needs
             essencetrack and how to
             repeat.


57       x   Similar to 11, 21, 37.            WGBH -known
             Documentation issue               needs
70           x   can already be done,         WGBH -
                 misunderstanding             Inventory


73       x       Same as 26, 29. Controlled   WGBH -
                 vocabulary review and        Inventory
                 documentation
75   x           done, need better
                 documentation




76   x           done, need better
                 documentation




77   x           done, need better
                 documentation
78       x   Similar to 43. Schema
             change or a
             documentation/use case
             solution




79       x   Similar to 53. Documentation.
             Now that we have assetType
             (1.3) we can combine with
             existing relationtypes. Also
             relates to collection needs.



80   x       done, need better
             documentation




82   x




83   x       done, need better
             documentation
84   x   done, need better
         documentation




85   x   done, need better
         documentation




86   x   done, need better
         documentation
87   x   Same as 18, 41, and 52




88   x   Documentation and review
         common practices.




89   x   Documentation.
90   x   Will add example of how to
         do this to documentation.
         Show how to use OAI to
         harvest PBCore metadata.




91   x   Documentation. Same as 2,
         3, and 49.




10   x   use minimal required fields    Rich Rarey
         for PBCore, add extensions
         for domain specific data.
         Could also use a style sheet
         to transform PBCore
         metadata. Use case and
         training issue.
11    x   x   Can make an annotation on       Mary Miller
              instantiations now. Addressed
              also in 21, 37 --- This is
              really 2 Questions: 1. how to
              record 1 program that spans
              multiple tapes (part 1, part 2,
              etc.), 2. need clarification -
              does she mean multiple types
              of assets for one work or
              multiple parts for one work?
              Whichever one, we need
              better documentation and use
              case because this is now
              possible with 1.3. Same as
              #23.



17    x   x   Same as 11. Need better     Yvonne Ng
              documentation and use cases


18        x   Will give examples for rights - Yvonne Ng
              1) allowing for a link to
              external document and 2)
              American Archive use case
              example spelling out use of
              rights schema as extension
21b   x   x   Same as 11. Need better         Andy Lanset
23        x   Same as 11. Need better         dave rice
              documentation and use cases




43        x   similar to 78. Schema change Daniel Jacobson
              or a documentation/use case
              solution


53    x   x   Similar to 78. 1.3 provides the WGBH -known
              mechanism to solve this.         needs
              Need to define a collection
              profile and/or create
              documentation. Review
              discussions on how to create
              collections and feeds. May
              use METS wrapper or Jack
              Brighton's collection structure.
59   x   Similar to 78 and 53. 1.2   WGBH -known
         does some of this.          needs
         StructureMap idea is like
         METS and the collection
         change requests.




65   x   Profile or use case issue with WGBH
         additional fields for
         acessibility as extensions.
         Create use case or user
         profile? NB: Change request
         has indicated to us that
         people don't know how to use
         PBCore - we need to figure
         out how to deal with this and
         prioritize list for use cases.




66   x   controlled vocabulary issue - WGBH
         stripped down subset of
         vocabuary, limited. Also a
         use case - how to develop a
         controlled vocabulary for your
         use of PBCore.
78   x   Similar to 43. Schema
         change or a
         documentation/use case
         solution
45   x       creating profiles - applications Jack Brighton
             of PBcore for different
             purposes (eg COVE, SIP for
             AA) - Question of ability to
             deliver within existing
             resources/timeline.Also,
             potential overlap with
             Broadway Video work about
             SIPs, AIPs and DIPs..




60   x       Kara did some work on this      WGBH -known
             for NDIIPP. A profile thing.    needs
             see notes. Create profiles to
             handle PREMIS.

65   x       Profile or use case issue with WGBH
             additional fields for
             acessibility as extensions.
             Create use case or user
             profile? NB: Change request
             has indicated to us that
             people don't know how to use
             PBCore - we need to figure
             out how to dea with this and
             prioritize list for use cases.




4    x       similar to 44. Use 44           Jack Brighton



7b       x   Already able to do this         Mary Miller




7c       x   Already able to do this - each Mary Miller
             file could be a record
8b        x   1.3 allows you to source your Mary Miller
              Controlled Vocabulary... don't
              have to use PBCore picklist




13a       x   Recommend stripped down          Kara Van
              PBcore (minimal fields) or       Malssen
              another instantiation data set
              (eg MPEG7).

13b   x       Solved by 1.3. Instantiation     Kara Van
              no longer required.              Malssen




22b   x       solved by authority attribute in dave rice
              1.3. Document.




33        x   Need clarification. Start time Jabran Soubeih
              plus duration = end. Frame
              specific metadata better dealt
              with by another schema. Or
              create a pbcore record per
              frame.
34        x   Done. Express with use of      Jabran Soubeih
              colons and semi-colons. See
              http://pbcore.org/PBCore/form
              atTimeStart.html

35        x   Use pbcoreRelation and           Jabran Soubeih
              assetType




37        x   multiple records can point to Stephanie
              the same formatIdentifier,    Sapienza
              each instantiation could have
              a different start time on the
              dvd
38           x   PBCore is not intended for  Walter Allasia
                 preservation metadata.
                 Recommend implementing
                 extension and using
                 PREMIS. Or create a METS
                 wrapper and include PBCore.
                 (Could show example in
                 documentation)



44       x       similar to #4.                  Jack Brighton




47   x           Need clarification of linkage   Hannah
                 from Dave Rice                  Sommers




50   x           Need clarification. Seems like Hannah
                 Hannah wants to create a        Sommers
                 record for this unique sound
                 and its source. 1.3 addresses
                 this by allowing instantiation-
                 free records.
55   x       x   In 1.2                        WGBH -known
                                               needs

56       x       Review date functions in      WGBH -known
                 PBCore                        needs



58           x   use different unique ids      WGBH -known
                                               needs
61           x   implementation issue, not a   WGBH -known
                 PBCore issue                  needs

62           x   out of scope                  WGBH -known
                                               needs


64           x   As far as we can tell an    WGBH
                 applicationissue, not a
                 schema issue. Maybe part of
                 the DD atom feeds issue
                 number 78.
74       x       Review MODS for             WGBH - Chris B
                 authorityURI protocol
Title                  Change Request


                                                      Schema
Peabody Awards         Currently, in the containers for creator and contributor, you can't indicate
Collection Archivist   the contributor's affiliation.

                       Here's an example of a type of credit I work with on a regular basis:

                       Producer(s): Ric Burns (Executive producer)
                       Margaret Drain (Executive producer, American Experience)
                       Judy Crichton (Executive producer, WGBH)
                       Lisa Ades
                       Ric Burns
                       Steve Rivo

                       PBCore does not provide a way to document the person's affiliation.
                       One solution:

                       <pbcoreCreator>
                       <creator>Burns, Ric</creator>
                       <creatorRole>Executive producer</creatorRole>
                       </pbcoreCreator>

                       <pbcoreCreator>
                       <creator>Drain, Margaret</creator>
                       <creatorRole>Executive producer</creatorRole>
                       <creatorAffiliation>American Experience</creatorAffiliation>
Peabody Awards         Here is an excerpt from a bibliographic record for "The Autobiography of
Collection Archivist   Miss Jane Pittman."

                       511 Cast: Cicely Tyson (Jane Pittman), Odetta (Big Laura), Josephine
                       Premice (Madame Gautier), Michael Murphy (Quentin Lerner).

                       PBCore does not provide a way to include the role played by each cast
                       member; this is not irrelevant information. Here is one solution:

                       -<pbcoreContributor>
                       <contributor>Tyson, Cicely<contributor/>
                       <contributorRole>Actor<contributorRole/>
                       <contributorRoleElaboration>Miss Jane
                       Pittman</contributorRoleElaboration>
                       </pbcoreContributor>

                       -<pbcoreContributor>
                       <contributor>Odetta<contributor/>
                       <contributorRole>Cast<contributorRole/>
                       <contributorRoleElaboration>Big Laura</contributorRoleElaboration>
                       </pbcoreContributor>

                       etc. etc.
Peabody Awards           Currently, in the containers for creator and contributor, you can't repeat
Collection Archivist     the creator/contributor role element. You can repeat the container, but
                         this can be messy.

                         From a recent record:
                         executive producer, producer, writer, director, Eric V. Tait, Jr.
                         leads to this:

                         <pbcoreCreator>
                         <creator>Tait, Eric V.</creator>
                         <creatorRole>Executive producer</creatorRole>
                         </pbcoreCreator>

                         <pbcoreCreator>
                         <creator>Tait, Eric V.</creator>
                         <creatorRole>Producer</creatorRole>
                         </pbcoreCreator>

                         <pbcoreCreator>
                         <creator>Tait, Eric V.</creator>
                         <creatorRole>Writer</creatorRole>
                         </pbcoreCreator>

                         <pbcoreCreator>
Director of Archives -   The field for 'Frame Rate' could simply be a generic term: 'speed' or you
New York Public          could have a specific field for 'Analog Audio Speed' that would include
Radio (WNYC-             speeds for both disc and tape formats. Discs would have the option list
WQXR)                    of being RPM (revolutions per minute) at 16 2/3, 331/3, 45, 78, and
                         Other. Tape would have the option list of being IPS (inches per second)
                         at 15/16, 1 7/8, 3 3/4, 7 1/2, 15, and 30.
allow more than one sometimes an instantiation has more than one role, it would be helpful to
express of          state more than one formatGenerations. At WNET their prior database
formatGenerations   allowed multiple formatGenerations per instantiation, which make
                    mapping to PBCore challenging.
Archivist           There needs to be a way to break up or label (possibly through
                    attributes?) the pbcoreSubject field to indicate the type of subject being
                    referenced. For instance in MARC 21 Bibliographic, the subject added
                    entry fields in the 600s are different for personal names, corporate
                    names, chronological terms, topical terms, and geographic names, etc.
                    Also, PBS DLL v. 1.2 has something similar within the Coverage fields,
                    where there is a separate field for Coverage Event, Coverage
                    Geographical, Coverage Organization, Coverage People, and Coverage
Exec Dir of         25.00 pbcoreInstantiation
Engineering         Is there a way to indicate that an asset has been converted from its
                    native format to the current format. We often have two different
                    instances of the same asset one in native NTSC and the other in HD
                    which was upconverted from NTSC. If someone wanted a different
                    format than the current HD format it might be better to go back to the
                    original NTSC to do a "better" upconversion. Just knowing that this HD
                    asset was not native HD but was upconverted from something else would
                    be helpful. The next bit that would be nice is what was the native or what
Director of App.    iswould like to see PBCore evolve asset so it could be referenced the
                    I the GUID of the "parent" native into a feed-based structure, allowing
Dev., NPR Digital        natively more than one document to be distributed in a single request
Media                    (without the need for other protocols or standards). I have previously
                         suggested a wrapper XML element called that will allow multiple
                         elements to be bundled inside of it.
Librarian   Contributor (Element 16) / Named entity w/ role NPR’s librarians
            often cite the name/role metadata as the most valuable we collect.
             The metadata gives us the ability to locate tape quickly on deadline
            because we capture the name of the newsmaker with a code for the
            context (direct interview, public statement, singing, reading, acting,
            about, etc.) This metadata supports a very specific use case
            though, that we don’t expect to be replicated elsewhere. But would
            this metadata be useful to the American Archive? I think the answer
            The way content collections are to harmonize the critical to lay the
            is yes. Now the question is how handled, which is metadata withthe
            groundwork for a central archive of public broadcasting content.




            (Dave MacCarn): PBCore needs to have extension structure analyzed.
            Concerned that we not turn PBCore into MPEG 7, or SMPTE metadata
            dictionary – way too many fields. Could poll for controlled vocabulary
            (many requests are for CV changes, not elements).
            The schema should not require a particular sequence of element tags.
            This is unusual in my experience for schema, and I believe it
            unnecessarily increases the complexity of creating valid PBCore XML. I
            also don't see any upside to people who are writing PBCore parsers to
            have the tags in a particular order. I believe most parsers simply read
            the XML and create a data object which – how to differentiate between
            Creator/contributor/publisher and roles may no longer preserve the
            person and organization? Add contributor type? What if you want to
            express “Elizabeth Deane of WGBH as Producer”? How to do this?




            S1.04 Feed Creation
            Gain the ability to package PB Core into a feed or XML wrapper
            Users of PB Core should be able to package metadata into a feed or
            XML wrapper to facilitate transfer between systems and
            organizations. One approach could be through the use of wrappers
            like ATOM.
            ATOM provides a supported standard that can be used to package
            and distribute Public Media assets. The Atom Syndication Format is
            an XML language used for web feeds, while the Atom Publish
            Protocol (AtomPub or APP) is a simple HTTP-based protocol for
            creating and updating web resources. The advantage of using this
            approach for Public Media is in sustainability (a large community
            shares the cost of maintenance of standards) and compliance with
            best practices (Public Media avoids the risk of developing a custom
            standard that is not reflective or compliant with broader industry
                       S03.01 Applying containers to an instantiation
                       An Instantiation may have specific rights, relation or title
                       information that applies only to it and not to the entire asset.
                       Currently containers cannot be applied to Instantiations. A structure
                       is needed in order to relate Instantiation-specific metadata.
                       This issue should be addressed through the expansion of content
                       types (S02.01) and relationship definitions (S02.02), which will foster
                       structures that can be used to relate containers to instantiations.
                       Type: Schema Change




                                         Controlled Vocabulary
Chief Technologist     Much more tightly constrain the media taxonomy, so there is hope for
                       machines to be able to read the data.

                       For example, within PBCore currently, I could use the terms "HD", "High
                       Definition" or "High Def", "1080i" or "720p" interchangeably. A human
                       could read these. A machine could not.
Peabody Awards         Maybe we just aren't speaking the same language, but we cant' find a
Collection Archivist   formatDigital value for the streaming flash files we are creating, unless
                       you are using video/vnd.sealed.swf to cover flash.

                       We currently create .flv files using an on2 codec and f4V files using an
                       H.264 codec.

                       Maybe you are using video/H264 for f-4v files but what about .flv?

                       If the values I need are there, maybe what you need is a dictionary rather
Peabody Awards         than just aneed updates. Now thatadd the terms. actually use PBCore to
                       Term lists term list. If not, please I am trying to thanks
Collection Archivist   catalog our collection, I am stumped repeatedly. I have, for one particular
                       title (originally on 16mm) a dupe sound negative, a dupe picture negative,
                       and a duplicate preservation print. I can't find anything in
                       format/generations to accommodate this.

Peabody Awards         As compared to video, audio options on format generations are severely
Collection Archivist   limited. There is no plain "audio master" or "listening master" or
                       "preservation master", for example.
                       Relevant URLs:
Senior Consultant      re: language (see 12a) Additionally in the schema, I think it may make
                       PBCore easier to use if an empty language tag was supported by the
                       regex (currently the tag must be suppressed if empty).
Communications        Request adding the following vocabularies for genres:
Director
                      Alternative
                      Blues
                      Children's
                      Classical
                      Country
                      Eclectic
                      Electronica
                      Experimental
                      Folk
                      International
                      Jazz
                      Latin
                      R&B and Soul
                      Rap and Hip Hop
                      Rock

                       This list was derived from looking at the common denominator terms from
Archivist              Improve controlled vocabulary for formatStandard and formatDigital, with
                       better distinction between what goes in each element. Also, there are
                       some values in formatDigital that are confused with values that should be
                       in formatEncoding.
Director of Archives - Additionally - the GENRE drop down list is bloated with specific sports -
New York Public        not generally a public broadcasting focus. I would eliminate the specific
Radio (WNYC-           sports (i.e. baseball, figure skating, etc..) and just have "Sports" .
WQXR)                  Additionally it would be helpful to add the following to the GENRE list:
                       Unknown; Station Identification; Sound Effects; Public Hearing; Oral
                       History; Mood Music; Lecture; Soundtrack (film or filmstrip); Field
                       Recording; Concert Rehearsal; and Advertisment or Commercial which
                       would refer to conventional commercials as well as political or campaign
                       ads.
Director of Archives - Forgive me if I'm repeating myself here. I would really suggest that you
New York Public        folks may want to review the audio side of the IASA cataloging rules
Radio (WNYC-           manual. I don't know if it has been updated, I have a 1999 edition.
WQXR)
                       If I have not already mentioned it, BWF or Broadcast Wave File needs to
                       be added to the list of formats digital -audio (5.04).

                      If there is no separate listing for analog audio speed, under frame rate
                      (25.18) I would add 15/16ths ips to the audio speed list. While this is not
                      great sound, there are a fair number broadcast logger tapes around that
                      were recorded at this speed by stations for legal purposes and FCC
                      requirements.

                      Under "Format Generations" (25.07) it would be good to have
                      "Audio/Preservation Master" and "Audio/Preservation Copy" and
                      "Compilation Copy"

                      Under "Format Standard" (25.08) it would be good to have some listing
                      for analog audio that indicates "lateral" and "vertical" for the disc cutting.
                      Both types of cutting were used for broadcast transcription discs. Both
Director of Archives - 1)    Contributor Role controlled vocabulary should include: Senior
New York Public        Broadcast Engineer; Senior Producer.
Radio (WNYC-
WQXR)                  2)    The recommended formatGenerations controlled vocabulary
                       applies separate vocabularies for each mediaType, so even though
                       ‘Moving Image/Preservation Master’ is a recommended value there is no
                       equivalent ‘Audio/Preservation Master’.

Exec Dir of            25.08 formatStandard
Engineering            Remove MPEG - MPEG is a compression not a format as NTSC, PAL or
                       SECAM. A video could be MPEG and NTSC. If that was the case the it
                       should have MPEG under fromatEncoding and NTSC unser
                       formatStandard. A user might pick MPEG in both and then we wouldn't
                       know if it is PAL or NTSC. Remove NTSC Progressive
                       I don't think there is anything such format and it would just lead to
                       incorrect data entry.
Exec Dir of            25.08 formatStandard
Engineering            There is no choice for Dolby E under Audio. I'd also suggest that many
                       of the choices are actually compression not format. I'd suggest you move
                       compression of audio to 25.09 formatEncoding and put mono, audio, 5.1,
                       etc in 25.08 formatStandard unless that is dealt with somewhere else
                       then maybe just remove audio from 25.08
Exec Dir of            25.18 formatFrameRate
Engineering            There isn't a choice for 50 fields/sec (25fps) this is used and we convert
                       to it for international delivery.
Exec Dir of            25.03 formatPhysical
Engineering            There isn't a choice for HDCAM SR. I think that should be added.
Exec Dir of            25.16 formatFrameSize
Engineering            There isn't a choice for 1440x1080 which is a common sub-sampled
                       frame size for 1920x1080 that's been compressed. A tape like HDCAM
                       which is compressed at 1440x1080 would still be 1920x1080 but a file
                       that's compressed at 1440x1080 may need to be indicated as such?
Exec Dir of            25.14 formatBitDepth
Engineering            There's no choice for 12 bit. 4:4:4 can be 12 bit.
HearingVoices.com      Hoping PBCore will incorporate the journalistic metadata innovations of
                       the Value Added News project:
                       http://valueaddednews.org/

                       (part of the Media Standards Trust: Transparency Initiative
                       http://www.mediastandardstrust.org/projects/transparency.aspx
                       fronted by Tim Berners-Lee)
Director of Archives - Change Request: I can't recall if this is a repeat, so forgive me if it is.
New York Public
Radio (WNYC-           Format generations - It would be good to have "Access Copy" or "User
WQXR)                  Copy" - We have a large number of duplicate CDs for our future
                       'listening room' for producers and others doing research where they can
                       come and listen to CDs on their own.

                       Format generations - It would be good to have "Compilation Copy" . To
                       save shelf space we used to compile up to ten news features on a single
                       shelf reference/master CD while also maintaining the original story CD.

                       Format generations - Concert production here generally results in to
                       types of tapes. "Mix-Minus" which is the original concert without the
                       broadcast announcer or inserted production interviews and "Mix -Plus"
                       which is what is produced for broadcast with the announcer and
                       intermission production(s).
Librarian              Genre (Element 5) / Format
                       Thanks for referring me to the documentation on this one. The
                       definition for PBCore’s Genre element states that it reflects the
                       “structure of the presentation, as well as the topical nature of the
                       content in a generalized form.” This is another place where I think
                       overall searchability and metadata collection could benefit from a
                       separation of two characteristics – and it seems that the creation of
                       two separate controlled lists would also be fairly straightforward. I
                       would be interested to hear the arguments though, for keeping them
                       together, as I’m sure that there were use cases that lead to this
                       reality.
                       NPR has a format field that I wasn’t sure would appropriately map to
                       the Genre element. We’re capturing metadata about the structural
                       presentation in that field.
                       [HS update: Our discussion of 7/14 provided helpful explanation of
                       the current state of the Genre picklist. Seems the American Archive
                       will want to define a specific set of useful genres for its purposes.
                       NPR expresses structure of presentation metadata, both at the
                       (Eli Ingraham/Forum Network): Needed culture and identity, region of
                       world but wasn't in PBCore. Reduced high level topic categories (from
                       PBCore & 10 others incl. TV Guide) from 20 to 10. Created
                       intermediate layer of a preset tag layer from PBCore/PBS/NPR
                       Ability to describe elements: interview, b-roll, stock footage
                       Episode # - where to put this? Is it a title type? Can it be an
                       identifier? (they include it in EBUObjectTypeCode but as subset?)
                       Enhance role vocabularies with audio roles – see EBUCore and
                       DAM
                       FormatDigital and FormatEncoding – streamline these vocabs –
                       confusion and overlap between file types, codecs and mimetypes

                                       Documentation/Training
Chief Technologist     Provide a formal structure for series of episodes - insuring that they
                       remain linked together and not bound by series name.
Chief Technologist     Provide a formal structure for chapters within an episode - vital for web
                       use. Each chapter must have separate meta-data, while the episode also
                       can have the same meta-data structure.
Peabody Awards         Maybe we just aren't speaking the same language, but we cant' find a
Collection Archivist   formatDigital value for the streaming flash files we are creating, unless
                       you are using video/vnd.sealed.swf to cover flash.

                       We currently create .flv files using an on2 codec and f4V files using an
                       H.264 codec.

                       Maybe you are using video/H264 for f-4v files but what about .flv?

                       If the values I need are there, maybe what you need is a dictionary rather
                       than just a term list. If not, please add the terms. thanks
Manager, Strategic     PBCore is many fine things, but terse it is not. In developing our PAIS
Technology             project (HD Radio tagging for visually impaired listeners), I looked at
Applications,          using the PBCore structure for the basis of the tagging and determined
NPRLabs                that it was too verbose, and useful tags could not be used in the time-
                       sensitive, limited bandwidth available in the HD Radio Program
                       Associated Data (PSD) channel.
                       My request is this:
                       Create a terse subset of PBCore that is specifically designed for
                       transmission in bandwidth challenged environments, such as HD Radio
                       PSD channels, and FM RDBS streams (even more bandwidth limited).
                       This subset would allow me to get the data to the receiver quickly, and
                       could have hooks to allow the listener to access the richer PBCore trove
Peabody Awards         for that data item using other [internet-based] means. whose duration
                       Currently, PBCore is inadequate for dealing with a work
Collection Archivist   is such that it is split between multiple carriers. An example would be a 2-
                       hour made for TV movie from the 1970s that is split between two 3/4 inch
                       U-Matic tapes. Analog and digital copies of the original may also be
                       divided into two parts, or may have been edited together for storage on a
                       single analog carrier or as a single digital file.

                       PBCore is currently structured upon the presupposition of a 1-1
                       correspondence between intellectual content/property metadata and
                       instantiation metadata. There is nothing to distinguish an instantiation
                       that holds part of the content of a work to one that holds the entire
                       content. In the event of several instantiations that each old part of the
                       work and together hold the entire content, there is no way to show the
                       relationship between the parts.

                       It would seem that this could easily be resolved by nesting multi-part
                       instantiations in an additional element container.

                       I know that I am not the only would-be PBCore user who has struggled
                       with this problem. Your solving it would remove a significant obstacle to
                       more widespread PBCore implementation.
Senior Consultant      yay the change request form!
                       This is mainly to test this out, but I think updating the data dictionary to
                       support a recommendation for ISO 639-3 in the language elements would
                       be helpful for some more multi-lingual applications of pbcore. Also the
                       data dictionary's recommendation for handling multiple languages
                       (eng;fre) is not valid within the schema.
Metadata Specialist    Support for segment-level description. Currently all descriptive metadata
                       is limited to the asset level. It is important to be able to describe
                       segments of an asset and relate those to the whole, rather than have to
                       create new PBCore records for each segment (and have flimsy
                       relationships between segments and assets).


Archivist              A way to handle instantiations that have more than one part (not essence
                       tracks). We have many instantiations like this due to the way our media
                       is captured (i.e. software creates multiple files in succession as it
                       captures content. All belong to same instantiation of a single intellectual
Archivist              More granular and machine-readable rights elements. E.g. an element
                       similar to METSRIGHTS Rights Category would be helpful for flagging
                       content.
Director of Archives - It would also be useful to have "Broadcast Script" or "Document" referring
New York Public        to a media item that may no longer exist but the original paper record or
Radio (WNYC-           script of the program remains.
WQXR)

Director of Archives -   3)      PBCore’s instantiations appear only to support rendtions that are
New York Public          made up of only one tape file. Most broadcast collections have items that
Radio (WNYC-             span multiple tapes or files or parts or segments. For example WNYC's
WQXR)                    Leonard Lopate and Brian Lehrer shows are each two hours long. The
                         first reel or CD or BWF of the program is typically part A or 1 of 2 and the
                         second reel or CD or BWF is part B or 2 of 2. Jonathan Schwartz’s
                         program is four hours long. Each hour is a part or segment.

update language iso PBCore recommends ISO 639-2. Since PBCore made this
639                 recommendation there is ISO 639-3. Can we update the recommendation
recommendations     to add this much more complete standard. The 639-3 codes are already
                    valid against the regex in the schema.




handling multi -         I probably already posted this, but on nearly every PBCore project I've
component                worked on this has been a major stumbling block. Most of the reasoning
instantiations           on this issue is spelled out here.
                         http://www.pbcoreresources.org/article/instantiations_components_and_e
                         ssence_tracks/
                         Relevant URLs:
                         http://www.pbcoreresources.org/article/instantiations_components_and_e
                         ssence_tracks/
Archivist                There needs to be a way to break up or label (possibly through
                         attributes?) the pbcoreSubject field to indicate the type of subject being
                         referenced. For instance in MARC 21 Bibliographic, the subject added
                         entry fields in the 600s are different for personal names, corporate
                         names, chronological terms, topical terms, and geographic names, etc.
                         Also, PBS DLL v. 1.2 has something similar within the Coverage fields,
                         where there is a separate field for Coverage Event, Coverage
                         Geographical, Coverage Organization, Coverage People, and Coverage
                         Time Period.
Exec Dir of              25.08 formatStandard
Engineering              Remove MPEG - MPEG is a compression not a format as NTSC, PAL or
                         SECAM. A video could be MPEG and NTSC. If that was the case the it
                         should have MPEG under fromatEncoding and NTSC unser
                         formatStandard. A user might pick MPEG in both and then we wouldn't
                         know if it is PAL or NTSC. Remove NTSC Progressive
                         I don't think there is anything such format and it would just lead to
                         incorrect data entry.
Exec Dir of              25.08 formatStandard
Engineering              There is no choice for Dolby E under Audio. I'd also suggest that many
                         of the choices are actually compression not format. I'd suggest you move
                         compression of audio to 25.09 formatEncoding and put mono, audio, 5.1,
                         etc in 25.08 formatStandard unless that is dealt with somewhere else
                         then maybe just remove audio from 25.08
Exec Dir of         25.00 pbcoreInstantiation
Engineering         Is there a way to indicate that an asset has been converted from its
                    native format to the current format. We often have two different
                    instances of the same asset one in native NTSC and the other in HD
                    which was upconverted from NTSC. If someone wanted a different
                    format than the current HD format it might be better to go back to the
                    original NTSC to do a "better" upconversion. Just knowing that this HD
                    asset was not native HD but was upconverted from something else would
                    be helpful. The next bit that would be nice is what was the native or what
R&D Manager         is the GUID of the Subtitles native asset so it could be referenced the
                    Change Request: "parent" and Speeches. In the case a video content
                    has subtitles (as well as in the case an automatic content analyzer
                    creates the STT for a specific audio track in the video), how PBCore can
                    support the text description? There are no elements for this. It should be
                    supported the temporal description of the Text. That will enable the
                    categorization and indexing of the shots inside the video.
R&D Manager         Change Request: Temporal Segmentation

                    The metadata describing the content must support temporal
                    segmentation, in order to express different temporal decompositions.
                    We can have a temporal decomposition describing a specific behavior of
                    the content such as chapters and shots.
                    These decompositions are available in MPEG-7

R&D Manager         Change Request: Rights description

                    Current situation:
                    "PBCore does not allow for storage of an external rights xml standard
                    inline within PBCore.
                    Potentially external rights standards could be used inline within the Rights
                    element if the schema allowed this flexibility."
                    Also the Element "pbcoreRightsSummary" is a flat text element,
                    considering only the ODRL Permissions, Constraints, and Obligations.

                    I suggest to enable have external Rights Metadata, expressed in a
                    whatever format. ODRL could be useful for some simple example of
                    BusinessToConsumer licenses.
                    Also MPEG-21 REL may be bounded to the B2C. In order to support a
                    complex description of B2B (the common behavior of a broadcaster) a
                    more detailed and complex schema is required. We are currently working
                    on MVCO (MediaValueChainOntology) trying to describe "contracts"
                    between broadcasters.
                    There are no "ultimate" standards, but I think that PBCore should enable
                    further extensions and new rights schemas.
                    This could be done allowing an external format/schema linked into a
                    specific element in the "Intellectual Property" class.
Director of App.    I would like to see PBCore evolve into a feed-based structure, allowing
Dev., NPR Digital   natively more than one document to be distributed in a single request
Media               (without the need for other protocols or standards). I have previously
                    suggested a wrapper XML element called that will allow multiple
                    elements to be bundled inside of it.
Librarian   Parent & Child metadata (in sequence)
            I’m not sure how to use PBCore to indicate the hierarchical and
            sibling relationships between stories and with the overarching
            series. The RelationType element looks like the closest match, but
            I’m still not sure where in the schema the notion of sequencing of
            story parts can be rendered. In thinking about the American Archive
            use case, I can imagine a user locating a story, and then wanting to
            see a listing of other stories from the same program (in sequence).
            Does the Relation element capture enough to be able to satisfy this
            use case?
            [HS update: Request stands: relates to Digital Dawn REQs in S02.
            In a newsmagazine program, sequence of stories is important. We
            would like to be able to leverage the story segment number as
            numerical metadata, so that the number retains its intelligence and
Librarian   Transcripts
            NPR transcripts are full-text representations of our news stories,
            and can be useful to locate exact quotations. I don’t think I’ve ever
            seen a standard schema that has a field for “transcript,” but since
            the American Archive isn’t fully designed yet, I thought I should ask
            the question of how transcripts might be handled. Is there a vision
            for transcripts?
            [HS update: I now understand that transcripts are intended to map
            to the Description element. Good. Dave Rice advises that html and
            other format tags are not supported, so we would lose some textual
            intelligence, but we could probably write an ingest script to format
            the transcript correctly on PBCore ingest.
            Digital Dawn’s REQ S05.02 should make even clearer the intended
Librarian   Rights
            We don’t have one specific way to collect information about rights
            at this time, but were we to add a field for it, it would need to collect
            metadata at the item or story level. Also, it would need to be
            searchable according to controlled vocabulary, if at all possible, in
            order to define what users with certain permissions may access.
            [HS update: Nothing to add here. NPR practice is still evolving.]
            The way content collections are handled, which is critical to lay the
            groundwork for a central archive of public broadcasting content.




            The current format elements need additional work: they currently
            apply to both video and audio without distinction. When creating
            metadata for an asset that has both video and audio there is no way
            to repeat the format elements within an instantiation. A restructure
            of the format elements in the XSD would allow for the repeat of
            some a good way to represent instantiations consisting of more than
            Need elements with the parent container of audio or video or others
            one physical item. For example, we have video of an event on
            campus, shot in Betamax. The master spans 2 Betamax cassettes.
            However, there is nowhere to indicate that the Betamax/master
 I want to repeat Identifier using NOLA code and any other id I
want to use. FormatIdentifer is more important/stable to me... Why
can’t I repeat identifier w/in an instantiation?
FormatDigital and FormatEncoding – streamline these vocabs –
confusion and overlap between file types, codecs and mimetypes
01.01 Definition of an “Asset”
Allow a means to define what the “asset” is intended to depict.
PBCore 2.0 should begin by offering a more granular definition of
what is meant by an “asset”. The “Asset” definition should not be
viewed as a passive definition driven by current content practices.
Rather, “asset” definition should be a conscious decision to define
the digital building blocks that will meet Public Media’s business
objectives.
Public Media generally talks in terms of “episodes” as its
fundamental asset, however, Public Media’s business objectives
clearly call for a much m ore granular asset categorization that
allows episodes to be de-assembled into audio tracks, closed
S01.02 Varied Content Types
Support a wider group of Asset content types
PBCore 2.0 should also support a wider understanding and
definition set for content type, specifically content types should:
• Be flexible enough to support methods for defining new “content
type profiles”
Include some of the common content types that are used in Public
Media, including but not limited to:
o TV: series, episode, clip,
o Radio: show, story
Type: Schema Change


S01.03 Abstract Assets
Support the definition and creation of “abstract” assets – assets
which do not include a corresponding instantiation or are only
represented as temporal portions of another asset’s instantiations
without their own discrete instantiations, such as segments or clips
or episodes.
Support for abstract assets is important for re-purposing segments
of existing content into new products. Definition of Public Media’s
foundational asset types is an important step in formulating how the
Public Media information architecture will address abstract assets
that have no instantiation (for example “series”). Once the
foundational asset types have been defined different approaches
can be evaluated for packaging abstract assets.
One approach for creating abstract assets is through the use of
XML-based wrappers like ATOM (see below).
Type: Schema Change
S1.04 Feed Creation
Gain the ability to package PB Core into a feed or XML wrapper
Users of PB Core should be able to package metadata into a feed or
XML wrapper to facilitate transfer between systems and
organizations. One approach could be through the use of wrappers
like ATOM.
ATOM provides a supported standard that can be used to package
and distribute Public Media assets. The Atom Syndication Format is
an XML language used for web feeds, while the Atom Publish
Protocol (AtomPub or APP) is a simple HTTP-based protocol for
creating and updating web resources. The advantage of using this
approach for Public Media is in sustainability (a large community
shares the cost of maintenance of standards) and compliance with
best practices (Public Media avoids the risk of developing a custom
standard that is not reflective or compliant with broader industry
S02.01 Asset Relationships
Enable Relationship of Assets Outside of Field Entries
Improve relationship handling and expression between assets. This
should occur in a way that allows for and encourages support for
inheritance from one asset to others.
Please see note on XML-based wrappers in S01.03, as one potential
solution.
Type: Schema Change
S02.02 Relationship Definitions
Define Relationships for Content Types
Develop a mechanism to define the relationships that a content type
might have (ex. a Series will have Episodes, and Episode may have
Clips).
Type: Schema Change
S04.01 Source ID
Ensure that Source ID fields (identifierSource, subjectIdentiferUsed,
etc.) are retained through implementation and Source ID values
retained through workflow
Source IDs are essential to understanding identifiers, descriptors,
and identifying related assets. PBCore records that rely on multiple
sources for their values must have an accompanying field for
Source ID that is filled and carried between systems.
Type: Implementation Workflow
S04.02 Multi-Source Identifiers
Source Identifiers from Multiple Sources and Terms from Multiple
Controlled Vocabularies
Ensuring that repeating data sets can be applied to single assets
(S03.01) will support structurally the requirement for expressing
both a term/unique identifier with an accompanying field defining
the source of the term/ID. User training may be required to ensure
consistent application of both elements in a record.
Type: Implementation Recommendation
S04.03 Multiple Controlled Vocabularies
Support terms from multiple controlled vocabularies.
PBCore 2.0 should support multiple controlled vocabularies in a
structured manner. Specifically, it should allow users to define a
“profile” or “extension”, which relates its metadata sets to a
specific controlled vocabulary and mappings to PBCore.
Provide a means for the community to nominate new terms for
PBCore’s recommended controlled vocabulary. Provide additional
guidance for the selection or refinement of PBCore’s existing
controlled vocabularies to meet specific use cases. Often the full
PBCore vocabularies are too extensive for specific use, divided
existing lists by use cases may allow users to apply lists more
relevant to their application.
S05.01 Genre Element Set
Revise the Genre controlled vocabulary to become more user
friendly and expand documentation and examples.
The purpose of Genre element set (per PBCore User Guide) requires
review and revision, specifically:
• Clarify the use of Genre in the schema and the PB User Guide.
Currently the User guide explains that the element set can be used
for both topical descriptors and editorial format descriptors. This
leads to confusion.
• Remove outdated descriptors from picklist
• Add new terms added as needed by users and by the use cases
defined above.
S05.02 Description Element Set
Revise the Description element set to be more user friendly.
The description element set requires review in terms of its original
design and the current practices of users in filling it. The number of
acceptable descriptions should be carefully analyzed with an eye to
evaluating whether the description types available to users should
be limited, whether Description should be divided into separate
fields for different types of Descriptions, or whether to maintain the
current Description element set and apply more effective user
training.
S06.01 Rights Management
Define an approach for rights management.
PBC 2.0 should include a more standard approach to rights
management, either through PBC itself (similar to the discussion of
“packaging” above) through and open standard for rights
management. A few specific requirements are important to note:
• Rights metadata should apply to assets at all levels of granularity,
from Clip to Series. This becomes especially important when
repurposing content.
• Once an abstract asset has been defined for a particular
distribution purpose (for example, assembling music and clips from
previous broadcasts to create a new asset) the rights should be able
to be defined for the new asset – as opposed to attempting to define
rights within a series of individual PBCore records.
• Apply a standard vocabulary to rights so that it can become a sort
able field to facilitate accessibility and decision-making.
As a note, one candidate approach for modeling digital rights is the
Open Digital Rights Language (ODRL) for standardizing the
expression of rights information over content. ODRL provides
flexible and interoperable mechanisms to support use of digital
resources in publishing, distributing and consuming of electronic
publications, digital images, audio and movies, learning objects,
computer software and other digital creations.
Type: Schema Change




S07.01 Linked Data
Supported Linked Data to supplement metadata from structured and
unstructured fields.
PBC 2.0 should make some accommodation for supporting linked
data to harvest information from digital sources. This topic requires
additional exploration and definition, as well as more specific
requirements.
Type: Schema
S08.01 ProfilesChange
Support the use of profiles as a mechanism for extending PBC to
customized scenarios.
PBC2.0 should allow extension of the schema to custom scenarios
through the use of profiles (“applications”). These should:
• Be used in the case of content providers that ingest metadata into
the American Archive and distribution partners which receive
content from the American Archive
• Allow certain metadata elements to be defined on a per-entity
basis, including but not limited to: Description, element sets,
controlled vocabularies, content types and required fields
Type: Implementation Recommendation
                     S08.02 Basic APIs
                     Publish a set of basic APIs for retrieving information from the
                     American Archive.
                     Since the American Archive may make its content available to
                     entities which do not require complex metadata and which do not
                     have trained metadata experts on staff, the American Archive
                     should create a basic set of APIs that support a few basic queries in
                     native PBC format, including but not limited to:
                     • Retrieving lists of available content with genre, content type
                     • Retrieving specific assets for display or playout
                     Type: Implementation Recommendation
                     S09.01 Segmentation
                     Define the method for accomplishing segmentation in common use
                     cases.
                     Segmentation should be addressed structurally in prior
                     requirements. However, there are still a variety of options about how
                     segmentation could be handled. PBC 2.0 should explicitly define the
                     way to do this in a way that satisfies common use cases, including
                     but not limited to:
                     • Creating chapter points within media files
                     • Creating clips from media files
                     • Supporting cataloguing of time based metadata, such as bottom
                     thirds
                     • Supporting time code based rights management
                     Type: Schema Change

                                                 Use Cases
Manager, Strategic   PBCore is many fine things, but terse it is not. In developing our PAIS
Technology           project (HD Radio tagging for visually impaired listeners), I looked at
Applications,        using the PBCore structure for the basis of the tagging and determined
NPRLabs              that it was too verbose, and useful tags could not be used in the time-
                     sensitive, limited bandwidth available in the HD Radio Program
                     Associated Data (PSD) channel.
                     My request is this:
                     Create a terse subset of PBCore that is specifically designed for
                     transmission in bandwidth challenged environments, such as HD Radio
                     PSD channels, and FM RDBS streams (even more bandwidth limited).
                     This subset would allow me to get the data to the receiver quickly, and
                     could have hooks to allow the listener to access the richer PBCore trove
                     for that data item using other [internet-based] means.
Peabody Awards           Currently, PBCore is inadequate for dealing with a work whose duration
Collection Archivist     is such that it is split between multiple carriers. An example would be a 2-
                         hour made for TV movie from the 1970s that is split between two 3/4 inch
                         U-Matic tapes. Analog and digital copies of the original may also be
                         divided into two parts, or may have been edited together for storage on a
                         single analog carrier or as a single digital file.

                         PBCore is currently structured upon the presupposition of a 1-1
                         correspondence between intellectual content/property metadata and
                         instantiation metadata. There is nothing to distinguish an instantiation
                         that holds part of the content of a work to one that holds the entire
                         content. In the event of several instantiations that each old part of the
                         work and together hold the entire content, there is no way to show the
                         relationship between the parts.

                         It would seem that this could easily be resolved by nesting multi-part
                         instantiations in an additional element container.

Archivist                A way to handle instantiations that have more than one part (not essence
                         tracks). We have many instantiations like this due to the way our media
                         is captured (i.e. software creates multiple files in succession as it
                         captures content. All belong to same instantiation of a single intellectual
Archivist                More granular and machine-readable rights elements. E.g. an element
                         similar to METSRIGHTS Rights Category would be helpful for flagging
                         content.



Director of Archives -   3)     PBCore’s instantiations appear only to support rendtions that are
handling multi -         I probably already posted this, but on nearly every PBCore project I've
component                worked on this has been a major stumbling block. Most of the reasoning
instantiations           on this issue is spelled out here.
                         http://www.pbcoreresources.org/article/instantiations_components_and_e
                         ssence_tracks/
                         Relevant URLs:
                         http://www.pbcoreresources.org/article/instantiations_components_and_e
                         ssence_tracks/
Director of App.         I would like to see PBCore evolve into a feed-based structure, allowing
Dev., NPR Digital        natively more than one document to be distributed in a single request
Media                    (without the need for other protocols or standards). I have previously
                         suggested a wrapper XML element called that will allow multiple
                         elements to be bundled inside of it.
                         The way content collections are handled, which is critical to lay the
                         groundwork for a central archive of public broadcasting content.
Version 2.0 should take a much more structural approach. Two big
issues to address: Recursive description for components of an asset –
the “structure map” issue. The need to support recursive, linked
descriptions for subsets of a production is an issue you are already
considering. The cleanest way to approach it would be to create a
second schema, “pbcoreStructureMap” and change the current
schema to pbcoreDescriptionDocument”. The current Schema 1.0
would become the pbcoreDescriptionDocument and Schema 2.0
would reference both the pbcoreStructureMap and the
pbcoreDescriptionDocument schemas Schema 2.0 may include
structureMap and descriptionDocument subschemas and define
how they relate to each other, a la the METS document schema, or
it may create the structureMap within the 2,0 schema by
(Madeleine Rothberg/NCAM)- Accessibility/ISO Access for All
considerations (based on XML, moving away from container to flat
based structure). Metadata about modalities of content access (screen
sizes, font colors, how to control, etc.) Teachers Domain.




(Eli Ingraham/Forum Network): Needed culture and identity, region of
world but wasn't in PBCore. Reduced high level topic categories (from
PBCore & 10 others incl. TV Guide) from 20 to 10. Created
intermediate layer of a preset tag layer from PBCore/PBS/NPR
metadata to ensure reliability.
S1.04 Feed Creation
Gain the ability to package PB Core into a feed or XML wrapper
Users of PB Core should be able to package metadata into a feed or
XML wrapper to facilitate transfer between systems and
organizations. One approach could be through the use of wrappers
like ATOM.
ATOM provides a supported standard that can be used to package
and distribute Public Media assets. The Atom Syndication Format is
an XML language used for web feeds, while the Atom Publish
Protocol (AtomPub or APP) is a simple HTTP-based protocol for
creating and updating web resources. The advantage of using this
approach for Public Media is in sustainability (a large community
shares the cost of maintenance of standards) and compliance with
best practices (Public Media avoids the risk of developing a custom
standard that is not reflective or compliant with broader industry
                             Profiles
Director of New        The recommendation of adding Profiles seems very important to me.
Media & Innovation     For example, a PBCore record conforming to a COVE profile would be
                       able to validate against COVE metadata requirements. The COVE
                       metadata requirements include a character-count limit on Title and
                       Description fields, and a specific COVE taxonomy. If I could feed COVE
                       a PBCore record that COVE knows is COVE-compliant, I wouldn’t have
                       to enter metadata into COVE by hand which is the current workflow.
                       You could specify a Profile by adding a namespace attribute in the
                       PBCore document root, and attributes inside the relevant elements. So
                       COVE would need a schema to validate the document, but that’s easy.
                       I’m guessing Public Media Publisher or whatever is coming down the
                       pike could function in the same way. PBCore documents based on
                       Profiles would remain valid and parse-able by applications that don’t
                       care about Profiles. All we’d be doing is adding functionality for
                       PREMIS. Another next step would be to map to the PREMIS schema
                       and probably extend the Rights and Instantiation elements to insure
                       that required elements are supported. There aren’t that many
                       required elements, but there are some elements specific to
                       preservation that are currently lacking in PBCore.
                       (Madeleine Rothberg/NCAM)- Accessibility/ISO Access for All
                       considerations (based on XML, moving away from container to flat
                       based structure). Metadata about modalities of content access (screen
                       sizes, font colors, how to control, etc.) Teachers Domain.




                                                        Other
Attributes             The addition of attributes to elements such as subject, genre, and
                       contributor to allow inclusion of UIDs for the values they contain. This
                       would enable PBCore to be very useful in the emerging linked data
                       universe.
Peabody Awards         Working with more recent items, I have a Website that is an html file
Collection Archivist   saved on a DVD. There is no option for format/physical for something
                       like this.


Peabody Awards         I have a CD with logos in .jpg format and credit sheets and biographies in
Collection Archivist   .pdf format. Again, there is no format/physical for a disc containing
                       anything other than audio or video, and no way to address multiple file
                       types with format/digital.
Peabody Awards         Please ALSO note that "Actor" is not included in the drop-down menu for
Collection Archivist   creators. Without Cicely Tyson, this production would NOT exist as we
                       know it. Should be possible to include her as creator, not contributor.

                       While my solution may not be the best one, this IS a problem that needs
                       solving.

Metadata Specialist    Modularity of PBCore Content Classes. There are often circumstances
                       where I have needed to use only the Instantiation class, but not the
                       Intellectual Content or Intellectual Property (these are served by other
                       standards such as Dublin Core or MODS).

Metadata Specialist    (see 13a) On the reverse, there are situations where I'd like to only use
                       the Intellectual Content and Intellectual Property elements to describe
                       content at a level where there aren't physical/digital items. An example
                       would be the description of a Series, which isn't embodied Instantiations
                       (while Episodes of that Series are). This could be achieved by changing
                       the requirement for elements such as formatLocation from Mandatory to
                       Mandatory if Applicable (if using Instantiation, for example), and adding
                       the ability to attribute identifiers to content classes.

update language iso Or for extra credit add an authority attribute like MODS has, so that the
639                 authority could be stated and to possibly allow use of other (or no)
recommendations     language standard.

                       This was a use case at Link TV where their music and culture department
                       has some fairly significant language requirements. IOS 639-2 didn't cut it.
                       Here is where we also need to use the semicolon delimited list for codes
                       that the data dictionary recommended but was invalid in the 1.2.1
                       schema.
Exec Dir of            25.11 formatTimeStart
Engineering            I would recommend the name of this change to formatSOM and another
                       descriptor formatEOM be added. Assuming that the EOM will be the
                       same as the duration is inaccurate when describing a scene. Having
                       EOM would also allow for scene or even frame specific metadata which I
                       feel has been missing from pbcore.
Exec Dir of            add field for formatTimecode
Engineering            It may be useful to have a field to describe the format of the timecode:
                       smpte non-drop, smpte-drop, etc.


Exec Dir of            identifierParent
Engineering            identifierSourceParent
                       Consider adding fields to identify the parent of a media asset. There may
                       be a video program that has video and audio. An organization may take
                       the audio track out and now has an asset that's audio only that came
                       from a parent that someone may want to identify.
                       I'm preparing a major field mapping for the transition to PBCore for a
                       collection of avant-garde and animation titles. The vast majority of these
Managing Director,     titles are very short (from 3-10 minutes), and so our library and study
iotaCenter             center DVD reference collection has units that contain several titles. This
                       creates a problem as I'm fairly sure PBCore does not allow several works
                       to be associated with one instantiation.
R&D Manager          Change Request: Preservation metadata. Currently content descriptive
                     metadata are not dealing with preservation metadata. Usually they are
                     left to the archives and very small steps have been done towards a
                     standardization. Concerning specifically the contents coming from public
                     television environment, I strongly suggest to define a minimal set of
                     metadata for managing the preservation of the content. For example,
                     near "INTELLECTUAL PROPERTY" where "creator, publisher,
                     contributor,..." are defined, we can have a "PRESERVATION
                     DESCRIPTION" where a "preservation holder" can be defined (that could
                     be a broadcaster already preserving the content) or a "master or copy" in
Director of New      order to understand what should idea is you could have a Contributor
                     On the topic of Linked Data, thebe done.
Media & Innovation   named “Hughes, Langston,” or “Smith, John” where the common name
                     may or may not be ambiguous or unique. But if you add an actual UID
                     to the name, it would be unambiguous and machine-readable. The
                     Linked Data way of doing this is to include a URI for the person. This
                     makes it possible to traverse the universe of objects containing this
                     URI, and it links our PBCore object to that universe. We could
                     optionally add URIs to place names, organizations, subjects, and
                     relationships in the PBCore record. The way to allow this would be to
                     add optional attributes to the PBCore schema, so you could include
                     UIDs inside elements. I wouldn’t make it mandatory though, because
                     many PBCore users might not be ready to go to the effort of adding
Librarian            Coverage (Element 7) / Geography
                     PBCore’s Coverage element allows metadata to be collected that
                     pertain to either spatial or temporal characteristics. I believe this
                     approach may limit the American Archive’s ability to make full use
                     of the power of both geographic and temporal metadata. It’s easy to
                     imagine users of the American Archive wanting to locate all content
                     about a specific location, or to use this metadata to generate
                     visuals, such as maps, etc. Or, users might desire a cross section
                     of stories about the 1950s, or 30s. It would be theoretically possible
                     to locate relevant content in response to these queries using the
                     current schema, but I think separating the two characteristics into
                     separate elements will make it easier to collect cleaner metadata,
                     and give greater flexibility in how the metadata is leveraged.
                     [HS update: Request to treat Geographic and Temporal metadata
                     separately stands. Allowing a linkage of the two would give even
                     greater power to researchers, as Dave Rice reports was supported
                     in previous releases. ]
Librarian            Unique Sound
                     Historically, NPR hasn’t been able to retain copies of the raw
                     materials that go into production. So, the assets we’re describing
                     are the finished product. Our unique sound field is used to describe
                     sound that comes from an external source and that could be re-
                     used in the future. I consider it to be our substitute for cataloging
                     raw elements. This is an important searchable field for us, and I
                     think it would be even more important to be able to search this
                     metadata in the American Archive because of the descriptive nature
                     of the metadata in this field. Not sure how to map it though.
                     [HS update: Relates to Digital Dawn REQ S09. We could use an
                     Extension element here except that Dave Rice advises the extension
                     doesn’t support paired data elements. Unless there is better
                     support for segmentation, we will lose some data by using the
                     extension element, but at least descriptions of clips within
May need to add an overall formatdatarate for the complete asset
(one that includes a video and and audio track) as we will have
separate elements for video formatdatarate and audio
Need a place to indicate the date the “intellectual content” was
made: for example, “Gone With the Wind” filmed in 1936. Takes
place in 1865. So coverage.temporal = 1865. Our instantiation is a
VHS version, and the VHS was made in 1990. then dateCreated =
1990. But where do you indicate that the “intellectual content” was
How to represent instantiations that are identical copies?

(Cataloging tool) A data entry list for "Subject" that is dynamically
contingent on the value of "subject authority".

(Cataloging tool) A comprehensive guide to understanding the built-
in report utility. While it's largely an optional feature, a FileMaker
developer could customize its output.

(Steve Snyder, Catherine Murphy/The World): Metadata is simple and
manual. Tag in iTunes and Public Interactive's software. Metadata not
searchable in web search. Use Audio Vault (similar to Dalet), and AP's
ENPS for description. Ideal: to have automatically generated RSS feed
LC released MODS updated).
(currently manually version 3.4, and added a couple things we might
want to parallel in PBCore 2.0

Functional changes
Add the "displayLabel" attribute to all top-level MODS elements that now
lack it. Rationale: makes MODS overall more consistent, and promotes
easier use of MODS as a back-end format for customized cataloging and
discovery systems. (March 11, 2009)

RDA-related changes
Add authorityURI and valueURI attributes to elements that have authority
attribute (October 14, 2009)
Add authorityURI and valueURI attributes to subject subelements
(October 14,
2009)
Relevant URLs//Notes   Urgency                     Category



                       This would be nice to have XSD or Schema or Structure




                       This is a stumbling block   Tag or Element
                       for me
                                    This is a stumbling block   XSD or Schema or Structure
                                    for me




                                    This would be nice to have Tag or Element,
                                                                        Vocabulary or
                                                               Definition




                                    This is a stumbling block   XSD or Schema or Structure
                                    for me


MARC 21 Bibliographic               This would be nice to have Tag or Element
Subject Access Fields:

http://www.loc.gov/marc/bibliogr
aphic/bd6xx.html,
             PBS DLL
Metadata User Guide:
http://metadata.pbsdigitallearnin
                                    This would be nice to have Vocabulary or Definition




The full recommendation can
be found at
http://www.pbcoreresources.org
/article/pbcore_recommendatio
n_pbcorecollection/.
                                 Required by CPB




Similar to "feed" request from   This is a stumbling block   XSD or Schema or Structure
DD




Not yet addressed.               Stumbling block             XSD or Schema or Structure,
Review structure, revisit                                    Vocabulary or Definition
controlled vocabs and update.

                                 Nice to have                XSD or Schema or Structure




(DM)
 <pbcoreCreator>

<creator>Deane, Elizabeth
(WGBH)</creator>
              <creatorRole>
Producer</creatorRole>
           </pbcoreCreator>
  <pbcoreCreator>
              <creator>Smith,
John (Nature)</creator>

<creatorRole>Producer</creato Required by CPB
Application of PBCore.                                       Schema Change
Feed is out of scope for 1.3,
but will add example
documentation mapping of how
to do this. Wrapper is in within
scope and detailed elsewhere.
Use cases: n/a
What is the use case here?        Required by CPB          Schema Change
Need clarification. Will
consider for 2.0. Not necessary
for AA Inventory. Record fields
could apply to instantiations,
fields could move, or use
existing annotation element.
Comment: A use case would
be if an asset is represented by
tapeA from local station and
tapeB from PBS. Then the AA
digitizes tapeA to acquire fileC.
In PBCore there needs to be a
means to document the
specific relationship between
tapeA and fileC. It will be
important to know what
instantiations the AA managed
content is derived from. This is
an example of apply
pbcoreRelation to
pbcoreInstantiation (although
this could be addressed with


                               This is a road block for me Vocabulary or Definition




                               This is a stumbling block   Tag or Element,
                               for me                            Vocabulary or Definition




                               This is a stumbling block   Vocabulary or Definition
                               for me




                               This is a stumbling block   Vocabulary or Definition
                               for me


                               This would be nice to have Vocabulary or Definition
allmusic.com                 This would be nice to have Vocabulary or Definition
amazon.com
archive.org
audiogalaxy.com
dmoz.org (which is google)
itunes radio
yahoo.com




                             This would be nice to have Vocabulary or Definition



                             This would be nice to have Tag or Element,
                                                                 Vocabulary or
                                                        Definition




                             This is a stumbling block   Tag or Element,
                             for me                                 Vocabulary or
                                                         Definition
                                This is a stumbling block   Tag or Element,
                                for me                              Vocabulary or
                                                            Definition,
                                                            XSD or Schema or Structure




http://www.pbcore.org/PBCore/f This is a stumbling block    Vocabulary or Definition
ormatEncoding.html             for me




http://www.pbcore.org/PBCore/f This is a stumbling block    Vocabulary or Definition
ormatEncoding.html             for me




                                This would be nice to have Vocabulary or Definition


                                This would be nice to have Vocabulary or Definition

                                This would be nice to have Vocabulary or Definition




                                This would be nice to have Vocabulary or Definition

 http://valueaddednews.org/     This is a stumbling block   Tag or Element
http://www.mediastandardstrust for me
.org/projects/transparency.aspx




                                This would be nice to have Tag or Element
                                  Required by CPB




Not yet addressed.                Nice to have                Vocabulary or Definition



                                                              Vocabulary or Definition
Dave M: This can be an            Nice to have
identifier

Dave M: We are looking at this.




                                  This is a road block for me XSD or Schema or Structure

                                  This is a road block for me XSD or Schema or Structure


                                  This is a stumbling block   Tag or Element,
                                  for me                            Vocabulary or Definition
see: The Technical Basis of   This is a road block for me Other
PAIS, (downloadable .pdf)
 http://www.nprlabs.org




                              This is a road block for me Tag or Element,
                                                               XSD or Schema or
                                                          Structure




                              This would be nice to have Vocabulary or Definition




                              This is a road block for me XSD or Schema or Structure




                              This is a road block for me XSD or Schema or Structure



                              This is a stumbling block   Tag or Element
                              for me
                                 This would be nice to have Tag or Element,
                                                                     Vocabulary or
                                                            Definition


                                 This is a stumbling block   Tag or Element,
                                 for me                              Vocabulary or
                                                             Definition,
                                                             XSD or Schema or Structure




http://en.wikipedia.org/wiki/ISO This is a stumbling block   Vocabulary or Definition,
_639-3                           for me                                    XSD or Schema
                                                             or Structure




                                 This is a road block for me XSD or Schema or Structure




MARC 21 Bibliographic            This would be nice to have Tag or Element
Subject Access Fields:

http://www.loc.gov/marc/bibliogr
aphic/bd6xx.html,
              PBS DLL
Metadata User Guide:
http://metadata.pbsdigitallearnin
glibrary.org/
http://www.pbcore.org/PBCore/f This is a stumbling block     Vocabulary or Definition
ormatEncoding.html                for me




http://www.pbcore.org/PBCore/f This is a stumbling block     Vocabulary or Definition
ormatEncoding.html             for me
                                 This would be nice to have Vocabulary or Definition




                                 This is a stumbling block   XSD or Schema or Structure
                                 for me




                                 This is a stumbling block   XSD or Schema or Structure
                                 for me




                                 This is a stumbling block   XSD or Schema or Structure
                                 for me




The full recommendation can
be found at
http://www.pbcoreresources.org
/article/pbcore_recommendatio
n_pbcorecollection/
                                  Required by CPB




                                  Required by CPB




                                  Required by CPB




Similar to "feed" request from    This is a stumbling block   XSD or Schema or Structure
DD




Add another request re         This is a stumbling block      XSD or Schema or Structure
reviewing format pickilsts -
outdated and how to add new
formats (this was addressed in
1.2)

Not yet addressed. Either         This is a stumbling block   XSD or Schema or Structure
create 2 records (part 1 and 2)
or add a way to order
instantiations within a record
Dave M: FormatIdentifier is
already a repeatable field
within instantiation.
Dave M: We are looking at this.



Asset Type element needed.        Required by CPB   Schema Change
Also improved documentation
and example best practices.
Use cases: Inventory, All




Consider all audiences in         Required by CPB   Vocabulary Addition
adopting, suggesting or
developing asset type
controlled vocabulary.
Use cases: Inventory, All.




Already included-slight change Required by CPB      Schema Change
to XML schema - no change to
dictionary.
Better documentation and best
practices is needed. Remove
instantiation as requirement
within the PBCore schema.
Use cases: DONE.
Application of PBCore.           Required by CPB    Schema Change
Feed is out of scope for 1.3,
but will add example
documentation mapping of how
to do this. Wrapper is in within
scope and detailed elsewhere.
Use cases: n/a




PBCore 1.2 is able to express     Required by CPB   Schema change
these now but will re-visit and
review for missing
relationships. More specific
user scenarios needed. Use
cases: DONE


Already included in schema,     Required by CPB     Schema Change
may need new doc, PBCore
1.2 is able to express this
already, Needs better
documentation and example
best practices. Use cases:
Already included in schema,     Required by CPB     Implementation
may need new doc. Issue with                        Recommendation
use of controlled vocabularies.
Improve/add documentation
and best practices. Use cases:
DONE.




Already included in schema,     Required by CPB     Implementation
may need new doc. . Issue with                      Recommendation
use of controlled vocabularies.
Improve/add documentation
and best practices.

Comment: This may be a
schema issues as well. For
instance in a producer one
contributor may have more
than one ‘role’, but PBCore
only allows 0 or 1 roles per
contributor. Use cases: DONE
Already included in schema,       Required by CPB   Implementation
may need new doc. Issue with                        Recommendation
use of controlled vocabularies.
Improve/add documentation
and best practices.Use cases:
DONE.




Update descriptions and           Required by CPB   Implementation
examples. Not a required field                      Recommendation
but we'll enhance
documentation.




Update documentation about        Required by CPB   Implementation
descriptions and vocabularies                       Recommendation
Assess how to link to known     Required by CPB
schemas. Not necessary for
inventory project.
Rights can be handled with
other known schemas – we’ll
look at linking to those
schemas

Comment: This is true, but
managing one metadata
standard is already a burden
for most stations, handling two
standards and the relationship
between the two can be
challenging work. Additionally
there is need to include better
rights metadata in records to
facilitate web access
(stream/download permissions,
creative commons, etc).
Currently there is no
standardized or recommended
means to express rights within
PBCore, just a broad
container. Also PBCore does
not allow for storage of an
external rights xml standard
inline within PBCore.
Potentially external rights
standards could be used inline
within the Rights element if the
schema allowed this flexibility.
Will look at documentation.      Required by CPB   Implementation issue, not a
                                                   schema change.




Documentation, invite           Required by CPB    Implementation
tool/profile development from                      Recommendation
users. Ask for definition of
needed profiles
Will add example of how to do Required by CPB            Implementation
this to documentation.                                   Recommendation
Out of scope for PBCore - this
is an American Archive project.




                              Required by CPB            Schema Change




see: The Technical Basis of   This is a road block for me Other
PAIS, (downloadable .pdf)
 http://www.nprlabs.org
                               This is a road block for me Tag or Element,
                                                                XSD or Schema or
                                                           Structure




                               This is a road block for me XSD or Schema or Structure



                               This is a stumbling block   Tag or Element
                               for me




                               This is a stumbling block   Tag or Element,
                               This is a road block for me XSD or Schema or Structure




The full recommendation can
be found at
http://www.pbcoreresources.org
/article/pbcore_recommendatio
n_pbcorecollection/.
Similar to "feed" request from This is a stumbling block   XSD or Schema or Structure
DD
Not yet addressed.                                 Other




Not yet addressed.Write some                       XSD or Schema or Structure
samples for how to use
extensions.

Review use of containers -
make better use of
conventional XML structures.

http://www.iso.org/iso/catalogue
_detail.htm?csnumber=41521

http://www.imsglobal.org/acces
sibility/



Not yet addressed.                 Nice to have    Vocabulary or Definition




Application of PBCore.           Required by CPB   Schema Change
Feed is out of scope for 1.3,
but will add example
documentation mapping of how
to do this. Wrapper is in within
scope and detailed elsewhere.
Use cases: n/a
http://www.pbcoreresourc This is a road block for me XSD or Schema or Structure
es.org/article/pbcore_and
_the_semantic_web/
http://www.pbcoreresourc
es.org/P10/




Not yet addressed.                 Nice to have                Other
Create a preservation profile
for PBCore with this mapping
and different required fields.

Not yet addressed.Write some                                   XSD or Schema or Structure
samples for how to use
extensions.

Review use of containers -
make better use of
conventional XML structures.

http://www.iso.org/iso/catalogue
_detail.htm?csnumber=41521

http://www.imsglobal.org/acces
sibility/




http://www.pbcoreresources.org This would be nice to have Tag or Element,
/article/pbcore_and_the_seman                                  XSD or Schema or
tic_web/                                                  Structure

                                   This is a stumbling block   Vocabulary or Definition
                                   for me



                                   This is a stumbling block   Vocabulary or Definition
                                   for me
                                 This is a stumbling block   Tag or Element
                                 for me




                                 This is a stumbling block   Tag or Element
                                 for me




                                 This is a stumbling block   Tag or Element
                                 for me




http://en.wikipedia.org/wiki/ISO This is a stumbling block   Vocabulary or Definition,
_639-3                           for me                                    XSD or Schema
                                                             or Structure




                                 This would be nice to have Vocabulary or Definition




                                 This would be nice to have Vocabulary or Definition




                                 This would be nice to have Vocabulary or Definition




                                 This is a road block for me XSD or Schema or Structure
                            This would be nice to have Other




http://www.pbcoreresourc This is a road block for me XSD or Schema or Structure
es.org/article/pbcore_and
_the_semantic_web/
http://www.pbcoreresourc
es.org/P10/




                            Required by CPB




                            Required by CPB
I would think total formatdata   This is a stumbling block   Tag or Element
rate more important than a + v
(this was addressed in 1.2)
add a dateCreated in the         This is a stumbling block   Tag or Element
intellectual content container



Not yet addressed. Can't you     n/a
just repeat instances?
Not yet addressed.               n/a                         Other - tools


Not yet addressed.               n/a                         Other - tools



                                                             XSD or Schema or Structure




http://www.loc.gov/standards/m
ods/changes-3-4.html
                                     Organization                      Do you
If other category,                                                     currentl
                   Organization:     Type:             Location:
explain:                                                               y use
                   [Organization]    [organizationtype [Location]
[categoryother]                                                        PBCore
                                     ]
                                                                       :

                  Walter J. Brown    Other organization   Athens, GA   No
                  Media Archives &
                  Peabody Awards
                  Collection




                  Walter J. Brown    Other organization   Athens, GA   No
                  Media Archives &
                  Peabody Awards
                  Collection
Walter J. Brown    Other organization   Athens, GA           No
Media Archives &
Peabody Awards
Collection




WNYC-WQXR          NPR Station          New York, New York   Yes
Radio




                                                             Yes



Louisiana Public   PBS Station          Baton Rouge, LA      Yes
Broadcasting




KCTS Television    PBS Station          Seattle, WA          No




NPR
                   PBS Station          55116              No




Walter J. Brown    Other organization   Athens, GA         No
Media Archives &
Peabody Awards
Collection




Walter J. Brown    Other organization   Athens, GA 30602   Yes
Media Archives &
Peabody Awards
Collection


Walter J. Brown    Other organization   Athens, GA 30602   Yes
Media Archives &
Peabody Awards
Collection
Audiovisual                                                Yes
Preservation
Solutions
channelAustin   Other public media   Austin, TX 78741     Yes
                organization




WITNESS         Other organization   Brooklyn, NY 11217   Yes



WNYC-WQXR       NPR Station          New York, New York   Yes
Radio




WNYC-WQXR       NPR Station          New York, New York   Yes
Radio
WNYC-WQXR         NPR Station          New York, New York   Yes
Radio




KCTS Television   PBS Station          Seattle, WA          No




KCTS Television   PBS Station          Seattle, WA          No




KCTS Television   PBS Station          Seattle, WA          No


KCTS Television   PBS Station          Seattle, WA          No

KCTS Television   PBS Station          Seattle, WA          No




KCTS Television   PBS Station          Seattle, WA          No

Hearing Voices    Other public media   Internet             No
                  organization




WNYC-WQXR         NPR Station          New York, NY         Yes
Radio
                   PBS Station          55116        No

                   PBS Station                       No


Walter J. Brown    Other organization   Athens, GA   No
Media Archives &
Peabody Awards
Collection
Proper terse subset.                       Other public media   Washington, DC       No
                                           organization




Problem could be       Walter J. Brown     Other organization   Athens, GA           No
solved in several ways Media Archives &
                       Peabody Awards
                       Collection




                       Audiovisual                                                   Yes
                       Preservation
                       Solutions



                       New York University Other organization   New York, NY         Yes




                       WITNESS             Other organization   Brooklyn, NY 11217   Yes



                       WITNESS             Other organization   Brooklyn, NY 11217   Yes
WNYC-WQXR          NPR Station          New York, New York   Yes
Radio



WNYC-WQXR          NPR Station          New York, New York   Yes
Radio




                   Other organization                        Yes




Audiovisual                             New York, NY         Yes
Preservation
Solutions




Louisiana Public   PBS Station          Baton Rouge, LA      Yes
Broadcasting




KCTS Television    PBS Station          Seattle, WA          No




KCTS Television    PBS Station          Seattle, WA          No
KCTS Television   PBS Station          Seattle, WA    No




EURIX Group       Other organization   Turin, Italy   Yes




EURIX Group       Other organization   Turin, Italy   Yes




EURIX Group       Other organization   Turin, Italy   Yes




NPR
Proper terse subset.   Other public media   Washington, DC   No
                       organization
Problem could be       Walter J. Brown    Other organization   Athens, GA           No
solved in several ways Media Archives &
                       Peabody Awards
                       Collection




                      WITNESS             Other organization   Brooklyn, NY 11217   Yes



                      WITNESS             Other organization   Brooklyn, NY 11217   Yes




                      WNYC-WQXR           NPR Station          New York, New York   Yes
                      Audiovisual                              New York, NY         Yes
                      Preservation
                      Solutions




                      NPR
Illinois Public Media   PBS Station, NPR        Urbana, Illinois 61801 Yes
                        station




Illinois Public Media   PBS Station, NPR        Urbana, Illinois 61801 Yes
                        Station, Other public
                        media organization

Walter J. Brown         Other organization      Athens, GA 30602      Yes
Media Archives &
Peabody Awards
Collection

Walter J. Brown         Other organization      Athens, GA 30602      Yes
Media Archives &
Peabody Awards
Collection
Walter J. Brown     Other organization   Athens, GA        No
Media Archives &
Peabody Awards
Collection



New York University Other organization   New York, NY      Yes




New York University Other organization   New York, NY      Yes




                    Other organization                     Yes




KCTS Television     PBS Station          Seattle, WA       No




KCTS Television     PBS Station          Seattle, WA       No




KCTS Television     PBS Station          Seattle, WA       No




iotaCenter                               Los Angeles, CA   Yes
vision and scope of   EURIX Group            Other organization   Turin, Italy          Yes
the format




                      Illinois Public Medi   PBS Station, NPR     Urbana, Illinois 61801 yes
                                             station
no funding for
tweaking cataloging
tool
no funding for
tweaking cataloging
tool
If yes use PBCore, how:
[howusepbcore]
We are currently implimenting a PBCore
compliant SQL archive database.




We first used PBCore during the
American Archive Pilot Project and plan to
continue using it as our in-house archive
project develops.
TRYING (operative word) to use PBCore
as model for our instantiation data.




TRYING (operative word) to use PBCore
as model for our instantiation data.
We are adopting PBCore for our
implementation of the Open Media
Project. See www.openmediaproject.org




our catalog is based on PBCore, and
exportable as PBCore XML


We are currently implimenting a PBCore
compliant SQL archive database.




Dave Rice just launched our new SQL
database.
We are currently implimenting a PBCore
compliant SQL database.




New PBCore compatible catalog now in
service.
...cannot for reasons (above)




Preservation repository descriptive and
technical metadata standard.




Our catalog is loosely based on PBCore,
and is exportable as PBCore (with a
workaround for the instantiation problem)

Our catalog is based on PBCore
We are currently implimenting a PBCore
compliant SQL archive database.



We are currently implimenting a PBCore
compliant SQL database.




for describing assets, what else?




We first used PBCore during the
American Archive Pilot Project and plan to
continue using it as our in-house archive
project develops.
Research




Research




Research
...cannot for reasons (above)
Our catalog is loosely based on PBCore,
and is exportable as PBCore (with a
workaround for the instantiation problem)

Our catalog is based on PBCore




We are currently implimenting a PBCore
PBCore is our core standard for
media cataloging




Our media cataloging system is based on
PBCore.


TRYING (operative word) to use PBCore
as model for our instantiation data.



TRYING (operative word) to use PBCore
as model for our instantiation data.
Preservation Repository descriptive and
technical metadata schema.




Preservation Repository descriptive and
technical metadata schema.




for describing assets, what else?




Working on mapping current collections
database to PBCore. Will likely use it as a
data structure standard (i.e. not use the
cataloging tool) in conjunction with
CollectiveAccess, so we can customize
the installation profile more easily.
Research




PBCore is our core standard for media
cataloging

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:17
posted:8/8/2011
language:English
pages:231