摘要
为解决传统的协同过滤算法不能准确理解用户的喜好,影响推荐准确率和推荐效果,提出基于社会化标签语义相似度的协同过滤算法.算法以标签语义相似度为基础,将项目资源和相关标签的语义信息纳入,显著提高了推荐系统的预测性能.研究结果表明:与以具体评分数据为基础的算法相比,该算法较好地解决了词相似度和句子相似度计算问题,推荐准确度和性能较以往的协同过滤算法有明显提高,改善了推荐效果.
In order to solve the traditional collaborative filtering algorithm can not accurately understand the user's pref- erences, affect the recommendation accuracy and recommendation effect, a collaborative filtering algorithm based on social tags semantic similarity is proposed. Based on the semantic similarity of tags, the semantic information of project re- sources and related tags is included, and the prediction performance of the recommendation system is significantly im- proved. Research results show that: compared with the algorithm based on the user rating, the proposed algorithm can solve the problem of word similarity and sentence similarity computation, and the recommendation accuracy and recom- mendation effect, as well as the performance of the proposed algorithm is significantly improved compared with the previ- ous collaborative filtering algorithm.
出处
《华侨大学学报(自然科学版)》
CAS
北大核心
2016年第1期84-87,共4页
Journal of Huaqiao University(Natural Science)
基金
广东省高等学校学科与专业建设专项(2013LYM_0110)
广东省教育科学规划专项(14JXN060)
关键词
协同过滤
推荐系统
社会化标签
语义相似度
预测性能
collaborative filtering
recommendation system
social tags
semantic similarity
prediction performance