微博演化网络的负信息分类方法

发布时间：2018-02-15 03:28

  本文关键词： 序列最小优化(SMO) 支持向量机(SVM) 演化网络 UCI数据集 负信息　出处：《计算机科学与探索》2017年01期 　论文类型：期刊论文

【摘要】：针对Sina微博博文的转发关系,建立起用户转发博文之间的演化网络,从而利用SMO SVM(sequential minimal optimization support vector machine)分类算法对博文进行分类,筛选出恶意博文、垃圾广告、垃圾营销信息,使用户能够精确地屏蔽不想要的博文和博主。第一步基于微博转发关系的演化网络和SVM分类算法对整个Sina微博进行分类;第二步利用复杂网络等技术对经常发送恶意广告的博主进行标注,从而在网络中对他们进行屏蔽;最后找出垃圾信息的来源以及分辨出博主是不是恶意转发者,在宏观上能更好地遏制垃圾信息的传播。与用户从UCI数据集中实际反馈情况进行比较,实验结果表明,机器学习分类的实验结果吻合度达到89%。
[Abstract]:In view of the forwarding relationship of Sina Weibo's blog posts, the evolutionary network between users' forwarding posts is established, and then the SMO SVM(sequential minimal optimization support vector machine algorithm is used to classify blog posts, and to screen out malicious blog posts, spam advertisements, and spam marketing information. The first step is to classify the whole Sina Weibo based on the evolution network of Weibo forwarding relationship and the SVM classification algorithm. The second step is to use complex networks and other technologies to mark bloggers who often send malicious advertisements, so as to block them in the network. Finally, find out the source of spam and distinguish whether the blogger is a malicious retweeter. Compared with the actual feedback from UCI data set, the experimental results show that the experimental results of machine learning classification are up to 89g.
【作者单位】： 武汉大学计算机学院软件工程国家重点实验室;三峡大学计算机与信息技术学院;
【基金】：国家重点基础研究发展计划(973计划)~~
【分类号】：G206;TP393.092
，

本文编号：1512316