基于信任网络的个性化推荐算法

Personalized recommendation algorithm based on trust network

以社交网络为平台的个性化推荐技术[1]已经得到了广泛的研究,但推荐系统仍然面临着若干问题,即数据稀疏性,用户冷启动等。文章提出了一种融合了信任网络的个性化推荐算法,在用信任网络信息进行推荐时,首先用余弦相似度的方法计算用户的相似度,通过相似度矩阵来对主题进行预测打分;然后计算用户与用户间的信任度;最后利用信任网络个性化推荐策略得到最优推荐结果推荐给用户。实验结果表明,提出的算法与传统的推荐算法相比,在准确率和召回率方面具有显著的提升。

The personalized recommendation technology based on social network has been extensively studied.However, several problems have still remained to be tackled in this area of technology,for example,data sparsity problem, cold-start problem and so on.In this paper,the author proposed a personalized recommendation algorithm combing trust network.Firstly,we used cosine similarity method to calculate the similarity of users,When recommending the use of trust network information; Secondly,we calculated trust metric between pairs of users; Finally,using the personalized recommendation strategy based on the trust network to get the best recommendation results to the users. The experimental results show that compared with the traditional recommendation algorithm, the proposed algorithm has a significant improvement in precision and recall rate.