Abstract
The Internet provides a general communication environment for resource sharing and Web is an example of a communication medium used for information representation and exchange on the Internet. Many organizations maintain Web servers which can collect the access logs of the Web users. These access logs provide a valuable source of information about the visitors' access patterns in the Web site, but Web robots may mislead the analysis of the access logs. Despite of advances in Web robot detection techniques, this paper propose a set of key components necessary for robot classification based on the access logs and provides a case study of the rule-based classification based on the proposed robot classification algorithm.