Document Sample
narrative Powered By Docstoc
1. Assessment of Need

         Organizing and providing access points to geospatial information such as maps
and atlases has been challenging for librarians and even more daunting for those without
expertise in geography, who must access them for their jobs and/or everyday life. While
we are already using driving direction services or tour map services frequently, everyday
life information seeking behaviors in geospatial need has been rapidly growing with great
interests and concerns because of natural disasters, such as hurricanes, tsunami, and
earthquakes. After the hurricanes of Katrina and Rita of 2005 in the state of Louisiana,
there has been an urgency and need to access geospatial information. Such interest is
expected, considering the relocation, reconstruction and recovery activities in the disaster

        Since serious devastation has occurred in New Orleans and in other parts of the
state, all types of physical and intellectual resources are now occurring in the city of
Baton Rouge, the state capital. It has become a crucial point for recovery activities,
future business investments, and educational arrangement. One of these resources,
geospatial information at Louisiana State University (LSU) has a tremendous potential
for serving the entire state.

       Geospatial information at the LSU campus is located in three places, Middleton
Library, Hill Memorial Library, and the Cartographic Information Center (LSUCIC). The
Middleton Library the main library holds geospatial information in the Government
Materials Section, which is distributed by the U.S. Government. Historical maps, atlases,
and other materials about Louisiana are in special collections, which are housed in the
Hill Memorial Library. The LSUCIC has maps, globes, journals, monographs,
photographs, slides, and atlases for faculty and students of the Geography Department as
well as the general public, government agencies, and business firms. Considering the
growing need by diverse groups of user population for geographic information, we
believe the digitization of these materials on the Web in the near future will make these
documents easily accessible to users. As evidence of this, a FEMA supported project for
constructing a geodata clearinghouse by LSUCIC was begun following the two
hurricanes in 2005. In order to make digitized geospatial information on the Web more
accessible and usable to users, we believe the existing geographic metadata should be
reconsidered for improvement based on users’ information behaviors such as their actual
information needs and seeking behaviors.

        Since the early nineties, the number and size of digital maps, and cartographic
information from the U.S. government through the Depository Library Program, and the
introduction of GIS (Geographic Information System) has increased in libraries. In order
to cope with this trend in libraries, a significant amount of efforts to cope with collecting,
organizing, complying with related technologies, and sharing and distributing electronic
spatial information have been made. User training also has been a part of their efforts to

facilitate access to electronic geographic data. Nevertheless, a survey (Kinikin & Hench,
2005) showed that most GIS use in smaller academic libraries was done by only faculty
in geography/geology for teaching or research, and little use by surrounding communities.
Moreover, describing and accessing spatial data is one of the most imminent issues
libraries have to solve to facilitate access points to casual users. As McGlamery and
Lamone (1994) pointed out, real time access to electronic data on Web and other digital
libraries will depend of the quality of the metadata. The previous studies by Westbooks
(2004), Nelson (1997), Mangan (1997) indicate efforts to combine or adjust existing
metadata schemes.

        Spatial metadata could be defined as digital information that allows the potential
user to understand data (Onsrud & Rushton, 1992), and allows users to determine if the
data is suitable to be integrated with other data (Flowerdew, 1991). Since spatial data
can represent many complex concepts and interrelationships, metadata should convey to
the data searchers the concepts and relationships represented within the data (Frank,

        It is fortunate for us to have tremendous amount of geospatial data available on
the Internet, but most of this geospatial data are represented by a metadata intended for
professional geographers. Gluck (1997) pointed out that the metadata standard has the
potential to improve the selection of datasets by providing improved expressions for
quality, accuracy, and availability. One of the barriers to the access and use of the current
available geographic data is the intellectual access that is highly influenced and guided by
geographic metadata. Studies by Hill (1990) and Holms (1990) showed that the use of
coordinate systems to represent geographic data are unambiguous to certain groups of the
use population such as geographers, while Woodruff and Plaunt (1994) demonstrated in
their study that users would like to retrieve information not only by explicit spatial
locations such as place names but also by more descriptive geographic characteristics.
They also pointed out that the current system does not adequately support these types of
indexing. Considering that diverse and large groups of users need geospatial information,
such as natural resources managers, historians, journalists, librarians, tourists, and policy
makers, everyday life geospatial information cannot sufficiently meet the current way of
data description. Therefore, the need for a new metadata that is usable and useful for
average users is evident.

2. National Impact and Intended Results
       The purpose of this project is to generate a new metadata for digital maps by
combining revised Dublin Core and National Spatial Data Infrastructure (NSDI)’s GIS
metadata to serve non-geographers as well as professional geographers. The two
metadata sets will be combined by RDF (Resource Description Framework) using
namespaces. With this new metadata set, digitized maps in the nation will be more
accessible to broader audiences and not only to coordinate-based geographers.

       Specifically, this project will
          1. identify and define the everyday geographic information needs and
              seeking behaviors of casual users;
          2. identify shortcomings in the current geographic metadata standard;
          3. propose a new metadata scheme with revised Dublin Core; and
          4. evaluate a prototype with a new scheme through a series of user testing.

        Our research goal is to conceptualize users’ geographic information needs and
seeking behaviors, improve the current metadata standards based on users’ actual
information situation and seeking behaviors, and to evaluate a proposed metadata format.
The revised version will be implemented with RDF, where namespaces are used to
identify each metadata scheme. This project will provide useful and valuable findings on
how geographic metadata should be described to serve users as appropriate representation
of data itself.

        The impact of the results of this project is expected to be significant in both
national and international levels. Nationally, all types of libraries and information centers
such as special libraries, academic libraries, and public libraries could apply the findings
of this project for developing and organizing their digital library collections of maps to
serve their users more effectively. Internationally, it would initiate an international effort
to establish a global standard for geo-spatial materials for various purposes.

        Since the new metadata scheme is based on qualitative and quantitative data from
real people’s everyday life, this project will provide researchers and practitioners in both
geography and library & information science with a better understanding of users’
geospatial information needs and seeking behaviors.

        This project will attempt to solve the problem by providing a new metadata
scheme, in which the current GIS metadata is usable only by professional geographers,
but not by many other people. In order to make the new metadata efficient and effective,
we will try implementing real users’ geo-spatial information needs and seeking behaviors
as much as possible, by asking them what they want to find. Based on the findings from
user surveys and interviews, a new metadata scheme will be proposed. If successful, the
new metadata set will facilitate other professionals to use GIS and digital cartographic
information more effectively in many areas such as public health, politics, sociology, or

        As a research project, internal and external validity (generalizability) of this
project work will be verified through rigorous research methods to answer our research
questions. In addition, our plans for iterative rounds of our prototype evaluation with
users will ensure process and findings will be beneficial to all the stakeholders such as
information researchers and practitioners as well as the everyday users.

3. Project Design and Evaluation Plan

1) Project Design
Objectives and Research Questions: The purpose of this study is to construct a
conceptual framework for developing procedures and guidelines for digitizing map
collections to the Web and for constructing geospatial clearinghouse in general. In order
to do this, we pose two research questions for investigators to focus on throughout this
project. Under each focus, there are several subsequent questions that are proposed and
we will attempt to answer.

Research Question 1: Geospatial information needs and seeking behaviors of the casual

          What are the geospatial information needs of casual users?
          How do casual users seek geospatial information and what kind of access
           points do they use most frequently?

Research Question 2: Evaluation of the existing metadata standards and proposing a
new way to describing information

          What are the shortcomings of existing metadata descriptions from the users’
          Are there any ways to improve the currently used metadata?

First, we try to investigate Research Question I in Phase I, and Phase II will be conducted
based on what we find in the Phase I.

Phase I: Identifying users' needs on everyday life geospatial information seeking

        According to our preliminary literature review, a limited number of research has
been done in defining geospatial information needs in terms of users (non-geographers)’
perspectives (Gluck, 1995, 1996). In order to identify and define geospatial information
needs from users’ perspectives, task performance and interviews will be conducted to
collect user data. A total of forty subjects, twenty from Baton Rouge, LA and twenty
from Milwaukee, WI will be recruited for the experimental sessions. Subjects from both
campuses will exclude students and faculty members from the geography department.
Each subject will be given 3 - 4 real life geographic information searching on the Web.
After task performance, open-end interview sessions will be held to collect verbal data
from the subjects. Subjects will describe their search processes retrospectively, and these
will be recorded and transcribed. The transcribed data will be analyzed by using concept

       Concept mapping is a structured process of constructing a topic or an interest and
producing an interpretable pictorial view of concepts and their interrelations (Novak &

Gowin, 1984; Tromchin, 2001). In this phase of the project, steps are:
      Step1 (preparation): Generation of statements; in our case, statements from user
                            interviews about information needs and searching.
      Step2: Sort and categorize the statements made by subjects into similar ones
      Step3: Rate (or make a ranking) to the each statement based on importance and
             representing it in map form
      Step4: Interpretation of map

Phase II: Re-describing a metadata for everyday life geospatial information seeking

   1. It does not mean excluding existing geographers' metadata. Rather we will expand
      the existing metadata with the findings in Phase I.
   2. It will be done in RDF combing multiple metadata with namespaces.
   3. To evaluate the new metadata, a series of user tests with a map retrieval system
      prototype will be conducted.

    Metadata has many uses on the Web, including organizing, searching, filtering, and
personalizing Web sites. Accurate metadata should make it much easier to find the Web
sites you want while ignoring the Web sites you do not want. To achieve these benefits,
however, Web sites, search engines, and directories must agree to use a standard format
for metadata. The Resource Description Framework is a World Wide Web Consortium
(W3C)-recommended XML application for encoding, exchanging, and reusing structured

         RDF only defines an XML syntax for encoding “resource-property type-property
value” triples in XML. It does not define the actual vocabularies used to describe
resources and properties. Efforts have been made to produce standard vocabularies for
digital library catalogs (Dublin Core) and news syndication (RSS).

      However, the more important aspect of RDF is that we can use multiple
vocabularies at the same time. In that case, namespaces are used to distinguish between
RDF elements and elements from other vocabularies in property types and values.

The following example shows how multiple schemas can be combined in RDF.

<? xml version="1.0" ?>
<RDF xmlns    = ""
     xmlns:DC    = ""
     xmlns:AGLS = "" >

  <Description about = "" >
    <DC:Title> The Future of Metadata </DC:Title>
    <DC:Creator> Jacky Crystal </DC:Creator>
    <DC:Date> 1998-01-01 </DC:Date>
    <DC:Subject> Metadata, RDF, Dublin Core </DC:Subject>
    <AGLS:Function> Information Management - Internet </AGLS:Function>

For digital geographical resources, the Federal Geographic Data Committee National
Geospatial Data Clearinghouse, also known as the National Spatial Data Infrastructure
(NSDI) established metadata for GIS1. Even though ESRI White Paper (ESRI, 2002)
noted that it can provide access to GIS data for users who are not necessarily GIS users
because the NSDI Clearinghouse is based on protocols for data exchange established by
the Library of Congress (Z39.50), still the metadata set is mainly for professional
geographers. It is far beyond everybody’s everyday geospatial information needs. For
example, as seen in the following example2, while “Keywords” element is too general to
meet non-professional geographers, “Spatial Domain” element specified 4 bounding
coordinates that are not so critical to everyday geo-spatial information needs.

                West_Bounding_Coordinate: -84.129
                East_Bounding_Coordinate: -75.535
                North_Bounding_Coordinate: 36.542
                South_Bounding_Coordinate: 33.757
                       Theme_Keyword_Thesaurus: None
                       Theme_Keyword: Roads
                       Theme_Keyword: Highways
                       Theme_Keyword: Transportation
                       Theme_Keyword: Streets
                       Theme_Keyword: framework
                       William S. Powell, The North Carolina GAZETTEER, A Dictionary of
                       Tar Heel Places, (Chapel Hill: University of North Carolina Press),
                       August 1984.
                       Place_Keyword: North Carolina

        From the data collected at Phase I, we will identify common aspects in everyday
geospatial information needs. Those aspects will be implemented with Dublin Core,
which is defined flexibly. We will not make any new element (tag) in Dublin Core, but
utilize as many existing attributes as possible. The modified Dublin Core set will be
combined with NSDI’s GIS metadata following RDF rules.

The following   example shows what the combined RDF may look like.
<? xml version="1.0" ?>
<RDF xmlns    = ""
     xmlns:DC    = ""
     xmlns:FGDC = "">

  A comprehensive template for the GIS metadata is available at .

       For developing a map retrieval system prototype for this project, selected digital
maps that were converted from LUSCIC map collections will have their metadata based
on the new RDF. The image filenames and their RDF metadata will be stored in MySQL
database, and the retrieval interface will be created by PHP. The usability of the interface
will be modified several times with repeated usability testing. The final version of the
prototype will be evaluated with a web form by the public.

This project will be accomplished through the following specific steps:

   1. Experiment design and setup (geographic information needs): includes developing
      tasks for experiments, developing interview protocols, and other materials. In this
      stage, we will recruit participants from Baton Rouge and Milwaukee.

   2. Conduct Experiments: Forty experimental sessions will be conducted. Each
      session will have an orientation session, task performance, after-task, and
      interview. All verbal data will be recorded. It is anticipated that each session will
      take approximately one and a half hours to complete.

   3. All verbal data will be transcribed.

   4. Concept mapping will be used to generate categories and structure those
      categories. Representation of structured data into various maps.

   5. Interpret these maps, and identify specific details for metadata revision.

   6. Define and select a small scale of sample collection of maps from LSUCIC

   7. Scan and digitize the sample collection for a prototype development.

   8. Construct a new RDF vocabulary by combination of revised Dublin Core and

  9.   Implementation of prototype by MySQL/PHP and modifications with repeated
       usability testing.

  10. Evaluation of the final version of prototype with a web survey.

  11. Writing journal articles and the final reports.

2) Project Evaluation
    We will conduct a series of user testing for our prototype with the general public. We
will make the prototype site open to the public and ask for their participation in a web
survey available at the prototype site. The data will show the advantage of the integrated

metadata in a RDF form and provide valuable information for improvement of the site,
which is critical for a full scale implementation. In addition, Michael Leitner, assistant
professor in the Geography Department at LSU will serve as a consultant to this project.
He will do a heuristic evaluation of the prototype from the subject expert perspective.
The latter step will serve do ensure that the geographic concepts, GIS metadata, and
constructs are accurate and appropriate.

The division of labor between the two major personnel is the following.

PIs     Interview     Conducting    Concept       Develop   Prototype     Prototype     Web
        Preparation   Interviews    Mapping       RDF       Development   Experiments   Survey
Ju    √               √             √             √                       √             √
Jeong                 √                           √         √             √             √

4. Project Resources: Budget, Personnel, and Management Plan

1) Budget

        The major portion of the requested budget will cover personnel. All figures
related to personnel such as salaries, fringe benefits, indirect costs are calculated in
accordance with existing policy of Louisiana State University. Budget details for
subcontractor is described in Budget Justification. Other details regarding requested
budget is described in Budget Justification.

2) Personnel

Curriculum vitae for personnel are attached to this proposal.

        Boryung Ju is an assistant professor in the School of Library and Information
Science at Louisiana State University. Her research interests include information
interactions, interface design and evaluation of information systems, usability analysis,
and human factors in system design. She has been involved in several research projects
on designing interface and evaluation of information systems including a small scale of
geographic data analysis software, and has published several articles in these areas. She
teaches Human Computer Interaction, Networks for Information Centers, Website Design
and Management, and Research Methods.

       Wooseb Jeong is an assistant professor in the School of Information Studies at
University of Wisconsin – Milwaukee. He has been working on several studies on force
feedback with GIS, and video games. He has been teaching Human Computer Interaction
and Usability for 5 years. He is also teaching XML for Libraries, and offered two XML
workshops for librarians. His recent interests on adaptive technology for the blind made
him a recipient of American Library Association’s 2005 Diversity Research Grants.

        Michael Leitner is an assistant professor of the Geography Department at LSU.
He will provide advice and support to this project. His research interests’ fall into the
area of geographic information science and technology (GIS&T) applied to criminology,
medical geography, and cartographic visualization. He is the co-editor and contributor to
“Mapping Crime: Understanding Hot Spots” published by the U.S. Department of Justice
in 2005 and the co-author of “Geographic Information Systems and Public Health -
Eliminating Perinatal Disparity” published by IRM Press also in 2005. In addition,
Leitner has had several peer-reviewed articles on topics such geographic profiling of
serial offenders and the visualization of the location of confidential (personal)
information, among others. He has taught primarily GIS&T courses, including computer
cartography, map design, GIS, GPS, aerial photo interpretatimn, and spatial analysis.

       John Anderson, Map Librarian and Director of LSU Cartographic Information
Center will serve as resource person for actual usage statistics of geospatial materials at
LSUCIC. He will also arrange and provide support maps and atlases from his collections
when we develop a prototype of digital map retrieval system with our proposed metadata

        Two research graduate assistants will be needed for assisting data collection, data
transcription, conducting experiment sessions, and secretarial work.

3) Management Plan

        The overall monthly management of the project will come through the School of
Library and Information Science, Louisiana State University, and federal government
regulations which will provide administrative procedures, money, facilities, and
equipment related to this project.

Expected timeline for two-year research plan is: (See Appendix XX for more details)

Oct. – Dec. 2006       Identification of tasks and preparation of experiment (task
                       performance & interviews) materials by Ju

Jan. – Mar. 2007      Recruitment of subjects by Ju & Jeong
                      Conduct experiments by Ju & Jeong

Apr. - Jul. 2007      Data analysis: Concept mapping by Ju
                      Interpretation & definition of specification by Ju & Jeong

Aug. – Jan. 2008      Selection of samples from LSUCIC map collection for the
                      prototype and digitization of samples by Anderson
                      Coding & developing RDF by Ju & Jeong
                      Prototype of map retrieval system development by Jeong

Feb. - Mar. 2008      User testing by Ju, Jeong, & Leitner

Apr. – May 2008       Refinement by Ju & Jeong

June – Aug. 2008      Evaluation of outcome (Web survey)
                      By Ju, Jeong, & Leitner

Sept. 2008            Dissemination of the findings

5. Dissemination
        The findings of this research project will be disseminated mainly through annual
conferences for professional organizations such as American Society for Information
Science and Technology (ASIS&T), American Library Association (ALA), Association
of Computing Machinery (ACM), and Association of American Geographers (AAG), and
research journals such as Journal of American Society for Information Science and
Technologies, Library Quarterly, Library Trends, Journal of Geography, or Professional
Geographer. As the project progresses, technical reports and research articles will be
presented in scholarly conferences and journals and proceedings in field of library and
information science and geography. We will extend ourselves to disseminate our findings
to other related fields.

6. Sustainability
        We believe that the findings of this project will be effectively applicable to
improve the existing geospatial metadata based on users’ perspectives. If this project
receives funding and the findings and a prototype proves to be useful and usable by
actual users, then we will implement this project to a real collection such as geographic
information at LSU Cartographic Information Center. At the same time, we will make
the findings and process of this project accessible on the Web so other researchers and
practitioners may apply and modify their own settings and situations. Finally, we will
continue to make our efforts to apply the approach of this project to other areas than
geographic information.


Flowerdew, R. (1991). Spatial data integration. In M.F. Goodchild, D. W. Rhind, & D. J.
      McGuire (Eds.), Geographic information systems: principles and applications.
      (vol.1, pp. 375-387). Essex, UK: Longman Scientific & Technical.

Gluck, M.(1997). Standards for electronic access to geographic and spatial information,
       In E. Logan & M. Gluck. Electronic Publishing: application and implications.
       (pp. 27-40). Medford, NJ: Information Today, Inc.

Gluck, M., Danley, E. & Lahmon, J. (1996). Public librarians’ views of the public’s
       geospatial information needs. Library Quarterly, 66 (4). 408-448.

Gluck, M. & Smith, L. (1995). Geospatial information needs of the general public: Text,
       maps, and users’ tasks. Clinic on Library Applications of Data Processing. 32.
       University of Illinois at Urbana-Champaign.

Harold, E. R. (2001). XML Bible, 2nd ed. NY: Hungry Minds.

Hill, L. (1990). Access to geographic concepts in online bibliographic files:
        Effectiveness of current practices and the potential of a graphic interface.
        Dissertation, University of Pittsburgh, Pittsburgh, PA.

Holms, D. (1990). Computers and geographic information access. Meridian, 4, 37-49.

Kinikin, J. & Hench, K. (2005). Follow-up survey of GIS at smaller academic libraries.
       Issues in Science and Technology Librarianship. Summer 2005. Retrieved
       January 11, 2006 from

Mangan, E. (1997). Crosswalk: FGDC content standards for digital geospatial metadata
     to USMARC. Retrieved December 14, 2005 from

McGlamery, P. & Lamont, M. (1994). New opportunity and challenges: Geographic
     information systems in libraries. Database, 17 (December), 35-42.

Novak, J. & Gowin, B. (1984). Learning how to learn. Cambridge, UK: Cambridge
      University Press.

Onsrud, H. J. & Rushton, G. (1992). Institutions sharing geographic information (tech.
      rep. no. 92-5). Santa Barbara: University of California, National Center for
      Geographic Information and Analysis.

Tromchin, W. (2001). The research methods knowledge base. 2nd Ed. Cincinnati, OH:
      Atomic Dog Publishing.

Iannella, R. (1999). An Idiot's Guide to the Resource Description Framework. Retrieved
        November 23, 2005 from

ESRI (2002). Metadata and GIS. Retrieved October 24, 2005 from

W3C, (1998). Resource Description Framework (RDF) Schemas. W3C Working Draft.
      Retrieved August 31, 2005 from

W3C, (1998). Extensible Markup Language (XML). W3C Recommendation, August 31,
      1998 from

W3C, (1999). Resource Description Framework (RDF) Model and Syntax. W3C
      Proposed Recommendation. Retrieved August 31, 2005 from

Westbrooks, E. (2004). Distributing and synchronizing in heterogeneous metadata in
      geospatial information repositories for access. Ed. American Library Association,
      Metadata in practice. American Library Association, 2004. 139-157.

Woodruff, A. & Plaunt, C. (1994). GIPSY: Automated geographic indexing of text
     document. Journal of the American Society for Information Science. 45(9), 645-


Shared By: