A NOVAL APPROACH FOR DOCUMENT CATEGORIZATION BASED ON LATENT SEMANTICINDEXING USING ANN

Abstract

In modern era, document categorization is most prominent and challenging problem in various domains of data mining. It helps greatly on defining interoperability in information processing systems. Document categorization institutes interoperability by deriving semantic similarity and correspondences between the information. In this paper Document Categorization Based on Latent Semantic Indexing or LSA has been discussed. For Document Categorization we follows the Various types of Techniques ,Tools and Technology such as KNN Classifier, Artificial Neural Network classifier and Tools is Newsgroup20 and Rueter21578 .In Document Categorization on LSI by using Various Techniques, Provides the Best Results In which we compare the results of various techniques. At last, we compare the result of TFIDF with LSI always LSI provides the best results as compare to TFIDF. By the way, if we find the results of all techniques after that the finally best will produce the ANN Classifier.