Recommendation systems, and specifically Social Filtering (SF) systems, play a significant role in reducing information overload and providing users with information relevant to their specific interest. For over a decade now, the ad-hoc standard in social filtering employed an approach, where recommendations were generated by computing "shared interests" based on users' preferences for items. The rapid growth in online social networks presents an opportunity for a new social filtering approach. The main thrust of our work is in identifying the relevant relationship characteristics among participants who know each other and use these characteristics to improve the quality of the recommendations generated. This paper develops a model that incorporates users'
|Title of host publication||Proceeding of the 1st Design Science Research in Information Systems and Technology conference (DESRIST)|
|Number of pages||20|
|State||Published - 2006|