Discovering Paths Traversed by Visitors in Web Server Access Logs

摘要

One of the most difficult and complicated problems for the Web Masters is obtaining feedback for their web sites. Web server access logs are originally intended for being used by Web Masters. In fact these files are the ideal source of feedback. On the other hand these files may reach to extreme sizes in a very short time periods and it becomes impossible to draw results without using advanced tools. As a result web server access logs are nice raw material for Data Mining applications but advanced techniques are necessary. In this paper, the problem of discovering paths traversed by visitors of a web site is handled. The path discovery tool, PathFinder2K mining web server access logs for the paths traversed by visitors, is developed. The design of the PathFinder2K is presented, the performance of the algorithm is analyzed, and some outputs of the tool are given.