Automatical storyline generation with help from Twitter

Storyline detection aims to connect seemly irrelevant single documents into meaningful chains, which provides opportunities for understanding how events evolve over time and what triggers such evolutions. Some previous work have been proposed, however, most of them generated the storylines through unsupervised technics such as clustering, which can hardly reveal underlying factors driving the evolution process. This paper introduced a three-layer Bayesian model to generate storylines from massive documents and infer the hidden relations and topics inside the documents. Besides, our model is also the first attempt that utilizes Twitter data as human input to ``supervise'' the generation of storylines. Through extensive experiments, we demonstrate our proposed model can achieve significant improvement over baseline methods and can be used to discover interesting patterns for real world cases.