Automatic Classification of Uighur Web Pages

In this paper, we introduce a classification approach for Uighur web pages. It utilizes the class feature dictionary and Cosine similarity computation to classify the Uighur web pages into the predefined classes rapidly and accurately. The experimental result shows that the approach has a good classification performance for Uighur web pages classification. It is useful and helpful for the construction of the statistical and rule-based classification of Uighur texts as well as construction of high-quality Uighur corpus.