融入情感差异和用户兴趣的微博转发预测

发布时间：2018-04-03 12:15

本文选题：微博转发　切入点：用户兴趣　出处：《图书情报工作》2017年09期

【摘要】：[目的 /意义]微博转发是实现微博信息传播的重要方式,对用户转发行为进行研究可以更好地理解微博信息传播机制,对热点话题检测、舆情监控、微博营销等具有重要意义。针对以往研究中用户兴趣表示不够全面准确以及未考虑情感差异对用户转发行为的影响,提出一个融入情感差异和用户兴趣的微博转发预测模型。[方法 /过程]该模型首先从维基百科中提取概念语义关系构建维基知识库,将其作为语义知识源对微博文本进行语义扩展,解决语义稀疏问题;对语义扩展后的用户历史微博进行聚类,提取用户兴趣主题和主题对用户的影响力;然后计算微博中各类情感的情感强度,提取情感差异特征;最后结合用户行为特征、用户交互特征、微博特征、用户兴趣特征和情感差异特征,运用SVM实现微博转发预测。[结果 /结论]在新浪微博真实数据集上进行实验,验证了所提模型的有效性。
[Abstract]:[objective / significance] Weibo retweeting is an important way to realize the information dissemination of Weibo. The research on the behavior of user forwarding can better understand the information dissemination mechanism of Weibo, detect hot topics and monitor public opinion.Weibo marketing and so on has the important significance.Aiming at the influence of user's interest expression on user's forwarding behavior in previous studies and without considering the influence of emotional difference on user's forwarding behavior, a Weibo forwarding prediction model incorporating emotional difference and user's interest is proposed.[method / process] this model firstly extracts the concept semantic relation from Wikipedia to construct the wiki knowledge base, which is used as the semantic knowledge source to extend Weibo's text to solve the problem of semantic sparsity.Cluster the user history Weibo after semantic expansion, extract the theme of user interest and its influence on the user; then calculate the emotional intensity of all kinds of emotions in Weibo, extract the emotional difference feature; finally combine the user behavior characteristics,User interaction feature, Weibo feature, user interest feature and emotion difference feature are used to realize Weibo forwarding prediction by SVM.[results / conclusion] the validity of the proposed model is verified by experiments on the real data set of Sina Weibo.
【作者单位】：武汉大学信息管理学院;武汉大学信息系统研究中心;
【基金】：国家自然科学基金项目“基于文本和Web语义分析的智能咨询服务研究”(项目编号:71673209)研究成果之一
【分类号】：TP391.1;TP393.092

【相似文献】