Home

---

WikiTracer explained

About

WikiTracer is a service aiming at providing platform-independent analytics and comparative statistics for wikis. By tracking the evolution in content, population and structure of wikis over time, this service aims to provide measurable indicators of the overall performance of individual wikis as well as visual representations of the growth of the wikisphere as a whole.

Background

WikiTracer builds on the successful example of Flickr Group Trackr, a service tracking and analyzing 15,000+ public Flickr groups since Jan 2007, offering statistics on the evolution of each of these groups. Since its launch, it has been adopted by a large number of group administrators from the Flickr user community. Flickr Group Trackr has been featured by ProgrammableWeb ("best new mashup"), Yahoo! Gallery ("editor picks"), LifeHacker, ResearchBuzz, Code & Form among others.

Extracting statistics from wikis is more complicated than Flickr groups just because different wikis run on different platforms: there is no such a thing as a global API for wikis. WikiTracer aims at providing a unified framework to track wiki evolution in a platform-independent way.

Why we care about wiki metrics

At WikiSym 2008 we presented a first case study showing a number of empirical predictions that can be made on the evolution of wikis by analyzing their structure and growth over time. By tracking data extracted from MediaWiki-based wikis, we were able to identify general patterns (such as access control, administrator density etc.) that boost or hinder the growth of a wiki. More accurate analyses are only possible if more fine-grained metrics are available from a variety of wiki engines: this is the gap WikiTracer aims to fill.

How will this affect you

Whether you are interested in building a sustainable wiki community or in studying wikis for research purposes, WikiTracer might be the essential tool you have been looking for.

• wiki owners, will be able to understand more about their wiki by monitoring its growth and potential problems
• wiki developers, will be able to assess the adoption of their product and the market share of specific versions
• wiki practitioners, will be able to obtain valuable insights into the performance of a wiki and to build best practices to spread wiki adoption against quantitative analyses
• wiki researchers, will be able to use the data made available by this service to perform large-scale studies on the evolution of wiki-based comunities

How it works

WikiTracer pulls statistics from registered wikis at regular intervals and processes this data generating analytics and comparative reports for wiki users. Before a public beta of this service is released, we need to hear from you about the WT plugin specifications.

Registration

1. Wiki admins will need to register to the service by providing basic information about their wiki. On registering, they will obtain a key that the service will use to pull data from the wiki.
2. A plugin will need to be installed by wiki admins to expose statistics about their wiki to WikiTracer via a RESTful interface.
3. WikiTracer will pull data on a regular basis from each registered wiki and store them in a central DB for analysis.
4. Detailed analytics and comparative reports will be made available on the basis of the history of each wiki. WikiTracer data will be available under an open license in the form of widgets, feeds, data dumps as well as via a dedicated API.

Privacy

1. Wiki admins will be able to configure their plugin so as to send to the server only the kind of information that they are happy to expose.
2. Corporate wikis or wikis behind a firewall will be able to benefit of WikiTracer services in a private way. Detailed wiki statistics will only be accessible to wiki owners whereas aggregate statistics will include data from private wikis in an anonymous way.