Latent Semantic Analysis: An Approach to Understand Semantic of Text
Pooja Kherwa, Poonam Bansal · 2017 International Conference on Current Trends in Computer, Electrical, Electronics and Communication (CTCEEC) · 2017
Latent semantic analysis (LSA) is a method for analyzing a piece of text with certain mathematical computation and analyzing relationship between terms in the documents, between the documents in the corpus.Various application of intelligent information retrieval, search engines, internet news sites requires an accurate method of accessing document similarity in order to carry out classification, clustering, summarizing or search tasks. So in this paper we are studying latent semantic analysis based on single value decomposition. The aim of Latent semantic analysis is to exploit the global structure of documents. The emphasis of latent semantic analysis is to find hidden relationship in document for better understanding the relationship between terms and document in dataset. In this paper, we have conducting a study using Latent semantic analysis (LSA) to find correlation of terms in a dataset consisting of research papers of various natural language processing applications.LSA shows that single value decomposition collapse multiple terms with same semantic and can identify terms with multiple meaning and represent documents in lower dimensional conceptual space.