A Rating Algorithm for the Analysis of Pornographic Websites
Kriangkamon Khumma, Chakkrit Snae Namahoot · 2013
The internet plays an important role in such areas as education, knowledge distribution, and e-commerce services, which has both advantages and disadvantages. In particular, access to pornographic sites or sites that are not appropriate for children should be monitored. With the ever increasing number of pornographic Web sites, analyzing as well as filtering and blocking them is a complex and difficult task. Various inappropriate sites have been found to hide their contents, e.g. by advertising of other sites that are not pornographic or using words or jargon expression that do not appear to be pornographic. In this research, we have developed a simple method that can analyze pornographic websites by using a semantic list (or database) of inappropriate words for the rating of a Web site containing pornographic content. For the analysis of an inappropriate Web site we use the structural analysis of HTML tags. After that, the rating of the inappropriate words is used to perform the computation in each part of the site structure for pornography identification. The system tests were performed on 150 Web sites, from which 50 sites contain pornographic material, 50 sites contain indecent but not obscene material, and 50 sites contain appropriate material. It was found that the system can rate 97 percent of the Web sites correctly.