Publication

Abstract : Due to large amount of textual data in the in the age of internet there is a need for quality information retrieval systems. Text mining or text analytics provides us such a method, usually by comparing patterns and arrangements of words. Applications of text mining include topic modelling where we intend to extract the topics or keywords from the given text. Another application would be document summarization. Summarizing any document(s) using algorithms has become common these days, and in order to gain information from large amount of data quickly, it is very much essential. This paper proposes an algorithm to combine document summarization with topic modeling and evaluating them based on various metrics available. Summarization algorithms such as Text Rank and Maximal Marginal Relevance (MMR) are combined with one of the topic modeling techniques known as the Latent Dirichlet Allocation (LDA). Both the summarization algorithms give better performance when applied with topic modeling and their performance is determined using precision, recall and F-measure, a set of evaluation Measures. © 2018, Institute of Advanced Scientific Research, Inc.. All rights reserved.