一种基于用户动态兴趣和社交网络的微博推荐方法

发布时间：2018-06-25 19:59

本文选题：动态兴趣 + 社交网络　；参考：《电子学报》2017年04期

【摘要】：针对为微博用户推荐符合其兴趣取向的个性化微博信息的问题,结合LDA主题模型,提出了一种基于用户动态兴趣和社交网络(DISN)的微博推荐方法.DISN方法首先引入时间函数,推断出用户的兴趣向量,通过对新发布的微博数据内容进行聚类分组,以用户兴趣向量筛选与用户最匹配的分组,随后以网格索引的形式对选定的分组中微博进行查询,计算微博发布者被目标用户关注的可能性并进行排序,最终形成推荐列表.实验验证了DISN方法较之传统方法更具有效性和高效性.
[Abstract]:Aiming at the problem of recommending personalized Weibo information according to their interest orientation for Weibo users, this paper proposes a Weibo recommendation method based on user dynamic interest and Social Network (DISN), which firstly introduces the time function. The user's interest vector is inferred, the newly published Weibo data content is grouped by clustering, the user's interest vector is used to filter the most matched group, and then the Weibo in the selected group is queried in the form of grid index. Calculate the likelihood that the Weibo publisher will be noticed by the target user and sort it to form a list of recommendations. Experiments show that the DISN method is more effective and efficient than the traditional method.
【作者单位】：南京工业大学计算机科学与技术学院;中国人民解放军73677部队;
【基金】：国家自然科学基金(No.61203072) 江苏省重点研发计划(社会发展)(No.BE2015697)
【分类号】：TP391.3

【相似文献】