Discovering Topic Evolution Topology in a Microblog Corpus

Topic evolution is an important research topic in natural language processing. Existing work has been focused on topic evolution of long text such as Web pages, news and blogs. In this paper, we aim to derive topic evolution summaries and its topology structures from short text like microblogs. We propose a microblog topic evolution algorithm based on retweeting relationships (MTERR). Firstly, we utilize topic models to generate microblog topics by incorporating temporal characteristics and retweeting features. Secondly, we build a topic correlation function to generate a topic evolution topological graph by considering both topic content and retweeting relationships. Experiments on the real-world event datasets show the effectiveness of our proposed methods.