Comparison of machine learning algorithms in Chinese Web filtering

Web filtering based on user's demand has witnessed a booming interest due to the development of Internet In the research community the dominant approach to this problem is based on machine learning algorithms. Web filtering is an inductive process which automatically builds a filter by learning from a set of pre-assigned document and the description of user's interest, and then uses it to assign unfiltered Web pages. This survey compares four main machine learning algorithms (decision tree, rule induction, Bayesian algorithm and support vector machines) on Chinese web pages set of their filtering effectiveness and computer resources consumed, focusing on the influence of feature set size and training set size. It induces that support vector machines earn high score in Chinese Web filtering applications.