基于SVM的微博转发规模预测方法
发布时间：2018-10-09 19:50
【摘要】：为了评价微博的传播效果,在分析影响用户转发行为因素的基础上,提出了采用用户影响力、用户活跃度、兴趣相似度、微博内容重要性和用户亲密程度五项特征进行转发行为预测的SVM算法,以及基于该算法的转发规模预测算法。最后给出了传播规模预测的评价方法。针对新浪微博用户数据的实验表明,预测精度达到了86.63%。
[Abstract]:In order to evaluate the transmission effect of Weibo, on the basis of analyzing the factors that affect the user's forwarding behavior, the author proposes to adopt user influence, user activity, interest similarity, etc. The SVM algorithm for predicting forwarding behavior based on the five features of Weibo's content importance and user's closeness, as well as the algorithm for predicting forwarding scale based on this algorithm. Finally, the evaluation method of propagation scale prediction is given. For Sina Weibo user data experiments show that the accuracy of the prediction reached 86.63.
【作者单位】： 国家数字交换系统工程技术研究中心;
【基金】：国家“863”计划资助项目(2011AA010603)
【分类号】：TP393.092
[Abstract]:In order to evaluate the transmission effect of Weibo, on the basis of analyzing the factors that affect the user's forwarding behavior, the author proposes to adopt user influence, user activity, interest similarity, etc. The SVM algorithm for predicting forwarding behavior based on the five features of Weibo's content importance and user's closeness, as well as the algorithm for predicting forwarding scale based on this algorithm. Finally, the evaluation method of propagation scale prediction is given. For Sina Weibo user data experiments show that the accuracy of the prediction reached 86.63.
【作者单位】： 国家数字交换系统工程技术研究中心;
【基金】：国家“863”计划资助项目(2011AA010603)
【分类号】：TP393.092
【参考文献】
相关期刊论文 前3条
1 林学民;王炜;;集合和字符串的相似度查询[J];计算机学报;2011年10期
2 施聪莺;徐朝军;杨晓江;;TFIDF算法研究综述[J];计算机应用;2009年S1期
3 张s，
本文编号：2260569
本文链接：https://www.wllwen.com/guanlilunwen/ydhl/2260569.html
