Web page Classification Using Ant Colony Algorithm

Abstract

In this paper we describe the new classification algorithm for web page classification that is ant colony optimization algorithm. The algorithm’s aim is to solve for discrete problem and discreteness of text documents’ features. In this paper, the system consists of two parts for classification: training processing and categorizing processing. In training process, the system removes the unnecessary part of the web page in preprocessing step. After preprocessing step, each text is represented by vector space model using TF-IDF formula. In the categorizing process, the testing web page is tested to classify appropriated class label by using ant colony algorithm. Ant colony algorithm works to find the optimal path or optimal class for text features by matching during iteration in the algorithm. Our proposed system is more robust and flexible than other traditional machine learning because it is based on swarm intelligence behaviors. The satisfactory accuracy of classification will get in this proposed system.