Project Fragmentation by pengxiang

VIEWS: 105 PAGES: 13

									Project Fragmentation
  The Project Fragmentation
  Problem in Personal Information
  Management
  Bergman, et al
  CHI 2006 proceedings
Fragmentation
 Information stored by format, rather than by
  purpose
   Documents files in folders
   E-mail in different folders
   Bookmarks in a separate hierarchy
 When you want all the information about a
  particular project or activity, what do you do?
Research study
 Basic question:
    Do people tend to work on materials based on format
     (as supported by current systems) or by purpose (as
     the researchers assume.)
 Participants:
    20 adults
    Mix of PC (most) and Mac (1) and combination
     experience
 Materials collected:
    Interviews, screenshots, questionnaires
Research Questions
 How do personal computer users tend to talk about
  their information organization -- in terms of
  technological format or in terms of projects?
 To what extent do users work on projects involving
  information items of different formats
 How much overlap is there between the three folder
  hierarchies
 Do users tend to classify their information according
  to format or porjects
 To what extent does interface design affect the
  project fragmentation problem?
 What are users’ attitudes towards integration of the
  different hierarchies?
Method
 Each subject was interviewed for about 90
  minutes, giving the interviewer a tour of the
  subject’s computer system, explaining the
  organization.
 Screenshots were captured during the “tour”
 The interview was taped, transcribed, and
  analyzed.
 A questionnaire was administered after the
  interview.
Results: How do users talk
aobut information organization?
 Data: the interview transcript
   Two judges reviewed it, looking for references
    to project or format that were not results to
    direct questions from the interviewer. Only
    cases where the judges agreed were counted
 Results: On average, 70.52% (SD= 16.35) of
  the references were to projects, 28.26% (SD=
  15.22) were to formats
 Conclusion: Users think of their information
  organization in terms of projects rather than
  formats
Results: Projects involving
multiple formats
 Data: Screen captures from the day before
  the interview (using history and recent
  documents files) Participants annotated each
  reference with the relevant project.
 Result: on average, 55.57% (SD = 32.61%) of
  items referred to a project which also had an
  item in a different format set.
 Conclusion: Users have information related to
  a particular project in more than one format.
Result: Extent of overlap among
the three types of format
 Data: Screen capture of the folder
  hierarchies. Overlap defined as folders of the
  different formats that refer to the same project
   Only root level folders examined
 Results: 19.79% of folders overlap (SD =
  19.38)
 Conclusion: About one fifth of the folders
  corresponded to a folder of the same project
  information in another format

 Note the difference between this and the previous one --
 overlap by project vs. overlap by information format
Result: Do users classify by
format or project?
 Data: Examination of the 968 folders found
  among the participants
 Result: Project folder names (M=79.94%, SD
  11.91) significantly higher than the proportion
  of format folder names (M = 6.16, SD = 7.3)
 Conclusion: Users tend to classify their
  information according to project more than to
  formats


 I’m not sure what is meant by “format folder names”
Result: Effect of interface
design
 Data: Questionnaire
 Results: Users mix documents of different
  types in one folder, but rarely save emails and
  bookmarks in these document folders.
 Conclusion: Interface design guides location
  of information storage.
Result: Attitudes toward
integration of the hierarchies?
 Data: Questionnaire
 Result: Average answer was 3.74 (on a 5-
  point scale)
 Conclusion?
Possible integration solutions
 Three approaches
   Integration through search
        Find items of multiple formats from one search
        Users seem to prefer navigation (see previous paper on
         that study)
   Integration through additional structure
        Shortcut to actual item stored in a separate structure.
   Single Hierarchy
        All project related items, of whatever format, stored
         together.
Proposed single hierarchy
solution
                                     All items of one
                                     project together, but
                                     separated by tabs

        QuickTime™ and a
          decompressor
  are needed to se e this picture.   What do you think of
                                     this strategy?

								
To top