基于微博的主题社区发现

MICROBLOG-BASED THEME DISCOVERY

为了满足微博营销寻找投放目标的需求,提出结合面向内容及连接关系分析的微博主题社区发现方法。创造性地加入了领袖发现、文本分类以及最大流社区发现的链接分析技术,同时采用多种剪枝策略,设计出一个高效准确的微博主题爬虫。实验经过真实数据的采集,并且从不同的维度对结果数据进行了实验分析。

In order to meet the demand of microblog marketing in hunting the delivery target, we propose a discovery method of microblogging theme community which combines the content-oriented and linking relationship-based analytical methods. In the paper, the link analysis technologies of authority discovery, text classification and max-flow community discovery are creatively added, multiple pruning strategies are employed simultaneously as well, we design a quite effective and precise microblogging theme crawler： Our experiments are passed with the collection of real data, and the result data are made experimental analysis from different dimensions.