一种基于自适应信任计算的推荐系统算法

摘要

Collaborative filtering algorithm based on k-nearest neighbour ( KNN) is one of the most core technologies in recommender sys-tem, and is the most widely applied technology at present .Trust-based recommender system can effectively solve the problems such as “cold start” and“vulnerability” encountered by the traditional collaborative filtering technology in its application .In this paper we present an adap-tive trust algorithm combining the above two recommender system algorithms .This algorithm has the characteristic that it can adaptively select the trusted calculation algorithm and the number of trusted nearest neighbours .%基于K近邻（ KNN）的协同过滤算法是推荐系统最为核心的技术之一，也是目前应用最广泛的技术；基于信任的推荐系统能够有效地解决传统协同过滤技术在应用中遇到的冷启动、易受攻击等问题。结合上述两种推荐系统算法提出一种自适应信任算法。该算法具有自适应选择信任计算算法和自适应选择信任近邻数目的特点。