融合多关系社交信任网络的推荐算法

Recommendation Algorithm Based on Multi-relationship Social Trust Network

推荐系统经常只引入单一社交信任关系,导致推荐准确率与覆盖率不高。笔者提出一种融合多社交信任关系的推荐算法,该算法利用多子网复合复杂网络模型构建多关系社交信任网络,将多元社交信任信息引入推荐系统以提高推荐的准确率与覆盖率。从数据集FilmTrust上的实验结果来看,本文算法的推荐效果有明显提升。

Recommendation systems often only a single social trust relationship is introduced,resulting in low recommendation accuracy and coverage.This paper proposed a recommendation algorithm based on multi-relational social trust network.This algorithm uses the multi-subnet composite complex network model to build a multi-relational social trust network,and introduces multiple social trust information into a recommendation system to improve the accuracy and coverage of recommendations.The experimental results on the dataset FilmTrust show that this paper algorithm recommendation effect has been significantly improved.