A Chinese web page classification algorithm based on combination of bayes classifier and clustering

Allowing for shortcomings of existing Chinese web page classification algorithms, a web page classification algorithm based on the combination of Bayesian classification and clustering is present in this paper. When maximum similarity difference between page and raining positive examples and negative examples of cluster centre is greater than a given threshold, it belongs to the current class. Otherwise, call the Bayesian classifier for classification. The experiment shows that this method can reduce the train scale of classifiers and improve the training efficiency. Its precision and recall are very good, and the test speed is also very high.