Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
403172 | Knowledge-Based Systems | 2007 | 9 Pages |
Abstract
The aim of this paper is to support user browsing on semantically heterogeneous information spaces. In advance of a user’s explicit actions, his search context should be predicted by the locally annotated resources in his access histories. We thus exploit semantic transcoding method and measure the relevance between the estimated model of user intention and the candidate resources in web spaces. For these experiments, we simulated the scenario of comparison-shopping systems on the testing bed organized by twelve online stores in which images are annotated with semantically heterogeneous metadata.
Keywords
Related Topics
Physical Sciences and Engineering
Computer Science
Artificial Intelligence
Authors
Jason J. Jung,