An Effective Hybrid Classifier Based on Rough Sets and Neural Networks

Due to the exponential growth of documents on the Internet and the emergent need to organize them, the automated categorization of documents into predefined labels has received an ever-increased attention in the recent years. This paper describes a method developed for the automatic clustering of documents by using a hybrid classifier based on rough sets and neural networks, which we called as Rough-Ann. First, the documents are denoted by vector space model and the feature vectors are reduced by using rough sets. Then using those feature vectors we reduced that are training set for artificial neural network and clustering the documents. The experimental results show that the algorithm Rough-Ann is effective for the documents classification, and has the better performance in classification precision, stability and fault-tolerance comparing with the traditional classification methods, Bayesian classifiers SVM and kNN, especially for the complex classification problems with many feature vectors

1. Introduction

With the exponential growth of textual information available from the Internet, there has been an emergent need to find and organize relevant information in text collections. For this purpose, automatic text categorization becomes a significant tool to utilize text information efficiently and effectively. Text categorization aims to automatically place the documents on previously classes. It is an active research area in information retrieval, machine learning and natural language processing. At present there are many textual classification methods, such as the support vector machine (SVM), the Nearest Neighbor, the Decision Tree, the K Nearest Neighbor method (KNN) and others.