WebTrace

David Hopp, Cato Research

General Description of Web Tracer

The goal of this project is to create a program, using Java and Perl, that will analyze web server logs to produce various information in a way that makes it possible for the site manager to make the site better.

The project was initiated by David Hopp and David Stotts. The idea was to produce a program that was able to supply the site manager with more information then the usual hit counts, etc. The piece of information most interesting is WHAT the visitor does when he/she visits the site.

The idea is to use Perl to parse the server logs into some sort of datastructure and then use Java to generate the visual display of the site. The program presents the site as a diagram with nodes, pages, and edges, transitions between pages, in 2 or 3 dimensions.

The information available in sever logs are remote host ( user's IP ), date, request ( HTTP ), status and size of the document.