利用概率主题模型的微博热点话题发现方法

摘要

微博具有长度短、实时传播、结构复杂以及变形词多等特点，传统的向量空间模型(VSM)文本表示方法和隐含语义分析(LSA)无法很好的对其进行建模。提出了一种基于概率潜在语义分析(pLSA)和 K 均值聚类(Kmeans)的二阶段聚类算法，此外通过定义微博热度分析和排序，有效地支持微博热点话题发现。实验表明，此方法能有效地进行话题聚类并检测出热点话题。%Microblog has the characteristic of short length, complex structure and words deformation. Therefore, traditional vector space model (VSM) and latent semantic analysis (LSA) are not suitable for modeling them. In this paper, a two stage clustering algorithm based on probabilistic latent semantic analysis (pLSA) and Kmeans clustering (Kmeans) is proposed. Besides, this paper also presents the definition of popularity and mechanism of sorting the topics. Experiments show that our method can effectively cluster topics and be applied to microblog hot topic detection.