With the proliferation of objectionable world wide web (WWW or web) materials such as pornography and violence, there is an increasing need for effective web content filtering tools to protect unsuspecting users from the harmful effect of such materials. This paper aims to discuss this issue.
Using pornographic web materials as a case study, the authors have developed an effective filtering solution that uses machine intelligence to perform offline web page classification into allowed and disallowed web pages.
The results are stored in a database for fast online retrieval whenever access to a web page is requested.
The separation between offline classification and online filtering ensures fast blocking decisions are made from the user's viewpoint.
There is an urgent and continued need for effective measures against the proliferation of objectionable materials on the web. In this paper, the authors describe a possible solution in the form of a complete working system. Future research will focus on adding appropriate modules to tackle other types of objectionable materials than the type described. The basic framework, however, should be applicable to a wide range of materials.
