User Types and Filter Effectiveness: A University Case Study

Abstract

As the amount of content on the Web grows almost exponentially, one of the new growth industries is that of filtering products. The effectiveness of Web-filtering software depends on a number of factors including the architecture of the software itself, and the sophistication of the users operating within its application domain. The main use of filtering software is to “block” access to controversial content such as pornography. This paper reports an investigation of the effectiveness of a filter called squidGuard in the real-world environment of an Australian University. The product is used to “block” pornographic material. This investigation simulates three classes of web users in trying to access pornography. While squidGuard did have limited success in blocking such material from novice users, the blocking rate dropped dramatically for the more experienced users using access lists. In all cases, however, access to supposedly filtered material was gained in seconds. Under such testing, the effectiveness of squidGuard as a specific-content filter for “pornographic” material can only be seen as superficial approach at best. The use of anonymous proxy servers was found to be an easy means to by-pass the filter.