Abstract
Finding the key topics in a large amount of short texts in social networks is a hot research point in data mining. There are a lot of models and algorithms to solve this problem, but they are not designed for social networks where many new words and nonstandard writings, grammars exist. It's more difficult to detect key topics in social networks because of these characteristics. In this paper, we propose a framework for detecting key topics in social networks. First, we get the posts in social networks using a focused crawler. Then we introduce the Word Segment Merging (WSM) method to identify new phrases in short texts and represent a document with the vector space model (VSM). At last, we model the life cycle of topics for clustering and popularity computing. Experiments on three datasets of SINA Weibo show that our method is better than existing state-of-arts models.