Toward the exploitation of social access patterns for recommendation
Jill Freyne, Rosta Farzan, Maurice Coyle · 2007
The size and diversity of the Web has been the root cause of the poor performance of many retrieval systems, with little navigational support provided by many large online formation repositories. The online information retrieval process cross different repositories shares similarities with content access facilities and user behaviors even when containing inherently different content types. In this work, we introduce our social recommender system called ASSIST. The recommendation framework in ASSIST can be applied to any online information retrieval service with key information access components, search and browsing. ASSIST exploits multiple forms of social implicit feedback in order to generate well-informed user recommendations in the online information retrieval domain.