Fuzzy data mining based automatic classification method of Chinese web pages

Abstract

The invention discloses a fuzzy data mining based automatic classification method of Chinese web pages. A classification word description library construction part, a new web page preprocessing part, a fuzzy classification matcher part and a classification result fuzzy judgment part are provided in the method. By applying a method of fuzzy comprehensive judgment, the fuzzy classification operation is carried out on a training set characteristic word vector fuzzy matrix generated when a classification word description library is constructed and a web page characteristic word fuzzy vector generated by the new web page preprocessing part, and finally the fuzzy classification of the Chinese web pages is obtained through classification result fuzzy judgment. The method can effectively improve the classification efficiency and solve the classification fuzziness problem and has favorable expandability, simple and convenient operation and easy popularization.

Description

technical field [0001] The present invention is aimed at the research of fuzzy data mining and the automatic classification method of Chinese web pages based on fuzzy data mining. The knowledge of data mining and the method of fuzzy classification of web pages involve technical fields such as automatic web page acquisition, Chinese web page preprocessing, Chinese word segmentation and keyword frequency analysis, and fuzzy classification of Chinese web pages. Background technique [0002] With the rapid development of Internet technology and Web technology, the number of web pages on the Internet is constantly increasing. The increasing popularity of the network and the explosive growth of the number of Internet users make the behavior of network users appear complex and diverse. How to properly analyze, manage and warn the behavior of network users is an urgent problem to be solved. Faced with the massive amount of information on the Internet, how to filter out the informat...

AI Technical Summary

Problems solved by technology

[0005] 2) Chinese webpages contain a lot of "noise", many webpages are not standardized, and contain a lot of advertisements, annotations and other information. It is necessary to preprocess the webpages, evolve the content of the webpages, and extract the information that users are interested in

[0006] 3) Most of the current research on Chinese webpage classification focuses on classifying webpages through feature selection and vector representation of Chinese webpages, using the KNN (k-Nearest Neighbor algorithm) classification algorithm to classify webpages. This method realizes the time cost and complexity of webpage classification high sex

Images