By:

This talk examines two different approaches to web statistics collection and analysis; with an emphasis on, 'doing it yourself'.

Firstly, collecting web statistics using an approach similar to Google Analytics. How to insert a tracker in your web pages (and how to bring down all your web sites at once if you get this wrong!) How then to cope with the large volume of data generated, and be able to report on the results.

There will be a demonstration of an application created using the open source Flash compiler, MTASC, communicating through a PHP script to a MySQL database: extracting data to show people navigating web sites, in real time.

Secondly, doing offline analysis of web server log files: parsing and summarising the data using Python; and storing this information in a MySQL or SQLite database. How to use Python, and the Python Imaging Library to generate a map of the world (derived from bathymetric data, processed with a custom map projection). Then overlay web statistics to create images, and an animation to illustrate global web usage.

These are large topics, so the 30-minute coverage will not be deep. The presentation will aim to give you some ideas to use in your own work, and illustrate the potential for informative (and even entertaining!) web statistics reporting.

Stream: Web & Online Technologies

Paper: DIY Web Statistics, DIY Web Statistics

I'm currently a web developer with the Antarctic Division, of the Department of Environment and Water Resources. I've been working on various large government web sites for the past seven years. Prior to that I worked as a TAFE teacher in the far north-west of Western Australia. I'm interested in web site programming and data visualisation, using open-source tools.