Bad webpage recognition method based on URL

Abstract

The invention discloses a bad webpage recognition method based on URL. The method judges whether a URL is the URL of a pornographic website through the semantic analysis to URL primary domain and the structure analysis to the whole URL. When judging, the two characteristics-sensitive string characteristic and structure characteristic contained in URL are extracted to be the basis for judging, and discriminator final comprehensive characteristic combined with SVM algorithm is adopted to perform secondary classification and obtain the judge result. The bad webpage recognition method based on URL of the invention can assist other recognition methods so as to fast recognize bad websites and provide healthy Internet environment; and the judgment can be performed without obtaining web contents so as to provide a high effective new idea for the recognition of pornographic websites.

Description

technical field [0001] The invention relates to a method for filtering bad information on the Internet, in particular to a method for identifying bad webpages based on URLs. The method involves the field of machine learning, and the final discrimination is accomplished by applying feature extraction and classification techniques in the field of machine learning. Background technique [0002] With the rapid development of the Internet, bad Internet culture is also flooding it, and the emergence of a large number of pornographic web pages has seriously affected the healthy development of young people. In recent years, research on automatic identification of pornographic content has made remarkable achievements. After a novelty search, the applicant retrieved two patents related to the present invention on the automatic identification of pornographic content, which are: [0003] 1. Multifunctional management system for network pornography and bad information detection [000...

AI Technical Summary

Problems solved by technology

[0014] 2. The method based on image and streaming media recognition has a wide range of applications, but the method handles a large amount of data, is highly complex, has a large delay, consumes a lot of bandwidth resources, and is not suitable for real-time recognition and processing in a network environment. ;

The disadvantage of this method is that it has poor flexibility and cannot cope with new sites;

[0016] 4. At present, there is no literature on the identification of bad web pages through URL analysis and semantic understanding, so this invention makes up for the vacancy in this regard and provides a new idea for quickly identifying bad web pages

Images