IR issues for digital ecosystems users

Abstract

The purpose of this research is to discuss some challenges of information retrieval, especially Web information retrieval, in digital ecosystems from a userpsilas perspective. As a dominant search tool, search engines usually return millions of search results in a long flat list in which many or even most of the results can be irrelevant. The long flat list conveys nothing about knowledge structure related to the retrieved results and personal search preferences and interests are not explored. Although some search engines try to cluster the Web results, the automatically formed titles and knowledge hierarchy is prone to mismatching the searcherpsilas human mental model. In digital ecosystems, while many different search tools are available, they are not integrated. To address these issues, a search framework which combines categorization, clustering, ontology, and personalization is proposed, and thus the quality of search results in digital ecosystems is expected to be boosted.