社区热点微博推荐研究
发布时间：2018-11-21 14:54
【摘要】：分析并总结了影响用户对特定微博兴趣的若干因素,在此基础上基于潜在因素模型提出了1个融合显式特征和潜在特征的社区热点微博推荐算法(community micro-blog recommendation,CMR),并将其用于发现微博兴趣社区热点信息.算法在3个兴趣社区上进行了实验,结果表明:1)融合2种特征信息的微博推荐效果好于使用单一特征信息的推荐;2)CMR的推荐效果好于基于转发次数的对照实验(micro-blog repost rank based recommendation,MRR);3)通过分析各个算法所推荐的微博内容,发现CMR倾向于为用户推荐兴趣社区相关微博,而MRR倾向于为用户推荐公共热点微博.
[Abstract]:This paper analyzes and summarizes some factors that affect users' interest in specific Weibo. Based on the model of potential factors, a community hot spot Weibo recommendation algorithm (community micro-blog recommendation,CMR) is proposed, which combines explicit and potential features. And use it to discover Weibo interest community hot spot information. The algorithm is tested in three communities of interest. The results show that: 1) the recommended effect of Weibo with two kinds of feature information is better than that with single feature information; 2) the recommendation effect of CMR is better than that of micro-blog repost rank based recommendation,MRR; 3) by analyzing Weibo content recommended by various algorithms, it is found that CMR tends to recommend community of interest to users, while MRR tends to recommend common hot spot Weibo for users.
【作者单位】： 中国科学院软件研究所基础软件国家工程研究中心;计算机科学国家重点实验室(中国科学院软件研究所);
【基金】：国家自然科学基金项目(61433015,61272324) 国家“八六三”高技术研究发展计划基金项目(2015AA015405) 网络文化与数字传播北京市重点实验室开放课题(ICDD201204)
【分类号】：TP391.3;TP393.092
本文编号：2347304
[Abstract]:This paper analyzes and summarizes some factors that affect users' interest in specific Weibo. Based on the model of potential factors, a community hot spot Weibo recommendation algorithm (community micro-blog recommendation,CMR) is proposed, which combines explicit and potential features. And use it to discover Weibo interest community hot spot information. The algorithm is tested in three communities of interest. The results show that: 1) the recommended effect of Weibo with two kinds of feature information is better than that with single feature information; 2) the recommendation effect of CMR is better than that of micro-blog repost rank based recommendation,MRR; 3) by analyzing Weibo content recommended by various algorithms, it is found that CMR tends to recommend community of interest to users, while MRR tends to recommend common hot spot Weibo for users.
【作者单位】： 中国科学院软件研究所基础软件国家工程研究中心;计算机科学国家重点实验室(中国科学院软件研究所);
【基金】：国家自然科学基金项目(61433015,61272324) 国家“八六三”高技术研究发展计划基金项目(2015AA015405) 网络文化与数字传播北京市重点实验室开放课题(ICDD201204)
【分类号】：TP391.3;TP393.092
【参考文献】
相关期刊论文 前1条
1 彭泽环;孙乐;韩先培;石贝;;基于排序学习的微博用户推荐[J];中文信息学报;2013年04期
【共引文献】
相关博士学位论文 前1条
1 段亚娟;微博搜索的关键技术研究[D];中国科学技术大学;2014年
相关硕士学位论文 前2条
1 吴磊;基于社会网络分析的微博用户研究[D];安徽大学;2014年
2 余紫丹;基于信任关系与主题分析的微博用户推荐技术[D];华东理工大学;2015年
，本文编号：2347304
本文链接：https://www.wllwen.com/guanlilunwen/ydhl/2347304.html
