Evolutionary Neural Network for Classification of Web Page

Abstract

As with the growth of internet and related technologies, more and more content were published on the World Wide Web. Automatic classification of web page according to their category or label is of benefit for many organization, for example focused search engine and news portal. Neural network are popular machine learning algorithm and were used for classification problem. Traditional neural network are trained by back-propagation algorithm which is a local search algorithm and can easily trapped in local optima. Genetic algorithms are types of evolutionary algorithm where multiple solution mimics natural evolution to solve the problem. This paper presents a web page classification system using neural network trained with the genetic algorithm.