Topic modeling of small sequential documents: Proposed experiments for detecting terror attacks

Research is proposed for improving the human-interpretability of topic models; specifically, for topic models of small sequential documents. Experiments are proposed for evaluating the usefulness of topic modeling. The proposed experiments will model the topics of a diverse set of social media content and attempt to correlate the presence of topics related to terror attacks with actual attacks; additionally, correlations between terror attacks and other topics will be checked for.