Binary Genetic Algorithm (BGA): A Hybrid Approach for Efficient Web Document Categorization

Abstract

Textual data on the web is growing at an exponential rate and millions of categories of data are available on web. In this big data environment, finding accurate information from these millions of categories is a challenging task. Web document categorization helps in classifying the web pages in to one or more predefined categories. This paper proposed hybrid Text document categorization approach. In this paper document categorization approach is improved by feature extraction and selection approach since filtering for relevant features is done twice. Here Deep Belief Network (DBN) is used for feature extraction i.e., DBN-FE, Binary Genetic Approach is used for Feature Selection i.e., BGA-FS and three different classifiers are used for evaluation of result of proposed categorization scheme. BGA Model gives the promising result for proposed approach in comparison to other approaches.