Analysis of user retrieval patterns in digital information systems.

Readers, more than anyone else, know how a collection of documents should be organized. Rather than querying users for their explicit preferences, analyzing document text content or using author-defined citations, the Dean project develops methods to derive document networks from temporal patterns in the retrieval behavior of large groups of users. The generated networks can be applied to document linking and information recommendation service, as well as the scientometric study of document impact and document relations.

Researchers:
Johan Bollen, Computer Science Department, Old Dominion University.
Soma Sekhara Vemulapalli, Computer Science Department, Old Dominion University.
Weining Xu, Computer Science Department, Old Dominion University.

Collaborators:
Herbert van de Sompel, Library without Walls, Los Alamos National Laboratory.
Rick Luce, Library without Walls, Los Alamos National Laboratory.
Michael Nelson, Computer Science Department, Old dominion University.
Francis Heylighen, Vrije Universiteit Brussels, Belgium.

Projects and Applications
- Zeitgeist project
- Open Video Recommendations
- Blackbox Relationship Modules for the Los Alamos National Laboratory

Presentations:
NASA LARC (07/2002), 52 pages, PDF (3-5mb).

Publications
* J. Bollen and R. Luce. Evaluation of Digital Library Impact and User Communities by Analysis of Usage Patterns. D-Lib Magazine, 5:6.
* J. Bollen and F. Heylighen. A system to restructure hypertext networks into valid user models. The new review of Hypermedia and Multimedia, 4:189-213, 1998.
More Links
Publications - Presentations - Johan Bollen - Who's Dean? - Related Projects

Johan Bollen
Last modified: Fri Nov 7 15:21:18 EST 2003