作者: Fady Draidi , Esther Pacitti , Bettina Kemme , Patrick Valduriez
DOI:
关键词:
摘要: In this paper, we propose P2Prec, a recommendation service for P2P content sharing systems that exploits users' social data. The key idea is to recommend user high quality documents in specific topic using ratings of friends (or friends) who are expert topic. To manage data, rely on Friend-Of-A-Friend (FOAF) descriptions. P2Prec has hybrid architecture work top any system. It combines efficient DHT indexing the FOAF files with gossip robustness disseminate topics expertise between friends. our experimental evaluation, CiteSeer dataset, show ability get maximum recall very good performance. Furthermore, it increases and precision by factor 2 compared centralized solutions.