摘要
在Web 2.0时代,人们可以通过在线社交网络快速并自由地发表自己的观点。微博作为典型的Web 2.0应用正吸引着越来越多的用户。在海量的微博信息中,充斥着大量的无用信息,这给微博内容检索提出了严峻的挑战。在微博发言中,由于用户的发言是由一个个单词组成的,而且那些热点单词往往出现在多个用户的发言中。将用户的发言与单词间的链接构成一个二部图,然后将该二部图转换成加权的用户发言网络结构图,并通过PageRank的排名方式将重要的内容提取出来。实验结果表明,该方法可以明显增加微博内容搜索的准确性。
In the era of Web 2. 0, people can post their opinions on the online social networks quickly and freely. Microblogging, as a typical Web 2. 0 application, has attracted more and more users. There are a huge amount of useless messages in massive Microblogging posts, and this poses a huge challenge for Microblogging con- tent searching. In Microblogging, users' posts are made up of words, and at the same time, one hot word usually appears in many posts. This paper first constructs a bipartite graph by the link of posts and words, then transforms the bipartite graph into a weighted post network, and at last ranks these posts according to the PageRank algorithm. Experiments show that the proposed method can improve the accuracy of Microblogging searching a lot.
出处
《科学技术与工程》
北大核心
2013年第31期9427-9430,共4页
Science Technology and Engineering
关键词
微博
搜索
排名
算法
Microblogging researching ranking algorithm