An Overview of Topic Representation and Topic Modelling Methods for Short Texts and Long Corpus

Abstract

Topic Modelling is a popular method to extract hidden semantic knowledge present in the collection of documents. Due to increase in textual data, topic models play significant role to infer meaningful topics in texts. In this paper, we present a survey of various topic representations and topic models used to discover topics in long and short texts. Topic modelling methods are grouped into standard topic modelling methods, clustering based methods, self-aggregating methods, and Deep Learning based methods. The conventional topic models such as Vector Space Model (VSM), Latent Semantic Analysis (LSA), Probabilistic latent semantic analysis (PLSA), Latent Dirichlet Allocation (LDA), Multinomial Mixture (MM) are discussed with their merits and limitations. Clustering based topic models and self-aggregating topic models used for short text corpus are discussed. Deep learning-based methods with traditional topic modelling to enhance the extraction of high quality topics are discussed.