Topics represent the key semantics of a set of documents in a terse manner (Blei, Griffiths, & Jordan,
2010). Topic models are probabilistic graphical models, which have been built by researchers to automatically infer the
distribution over a set of topics in a document or across a document set. After the latent topic inference, the document corpus
can be organized according to the topics inferred and hence the performance of the subsequent information-retrieval tasks
can be enhanced.