Web Classification using Deep Belief Networks

In recent years, deep learning approaches have gained significant interest as a way of building hierarchical representations from unlabeled data. These deep learning approaches have been applied to image recognition, voice recognition and text processing. However, to our knowledge, the deep learning approaches have not been extensively studied for web data. In this paper, we apply deep belief networks to web data and evaluate the algorithm on various classification experiments by comparing its performance with that of the SVM classification algorithm. In addition, the experiments show good performance of the deep belief networks for various classification tasks.