摘要
社交网络中存在密切的朋友圈子关系,而目前社交网络由于用户推荐方法单一,出现用户流失、朋友圈不明确等问题,通过用户标签的语义相似度的计算,对用户与其关注者之间的关系程度进行分析并按兴趣分类,根据分类进行关联关注用户的再推荐。通过实验验证了分类的有效性,提高了推荐效率。
There are exist a close relationship of friends among the social network. The current recommended method of social network users is single and lead to customer churn and relationship of friends is not clear. By the calculation of semantic similarity of the labels, analyse the relationship between user and attention degree and have a classification based on interest. According to the classification associated with attention to the user and then recommended. The validity of the classification is verified by experiment, on the other hand enhance the efficiency of recommendation.
出处
《贵州师范大学学报（自然科学版）》
CAS
2013年第1期94-97,共4页
Journal of Guizhou Normal University：Natural Sciences
基金
国家重点实验室开放课题(20090029)
安徽省高校优秀人才基金(2012SQRL227)
关键词
标签
相似度
关注用户
分类
推荐
label
similarity
concern users
classification
recommended
