A tool to analyze traffic on Wikipedia
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Failed to load latest commit information.


Wikistats - Wikipedia article traffic statistics

Wikistats is a tool to visualize how viewership of Wikipedia articles has changed over time. It ran on http://toolserver.org/~emw/wikistats/, but is now decommissioned. This repository archives the tool's source code.

It uses flot to plot page view data for individual Wikipedia articles. The underlying data is from the stats.grok.se tool created by Wikipedia user Henrik, and that data in turn is from official, anonymized raw page view data released for all Wikimedia projects.

This project augments the popular tool at stats.grok.se in a few ways:

  1. Page view data can be viewed over an arbitrarily long time span in daily increments. With Wikistats, you can see how viewership for an article has changed for over 4 years, rather than only 1, 2 or 3 months. This longitudinal data reveals trends that do not appear in shorter time spans. If you only want a week of data, you can get that too.
  2. Data for up to three different pages can be plotted simultaneously. You can plot three different pages in the same project (e.g. "Influenza", "Allergy" and "Insomnia" on the English Wikipedia), or the same page on multiple different projects (e.g. "Influenza" on the English, French and German Wikipedias).
  3. Summary statistics like the number of days plotted, total views, average number of views, and standard deviation are shown for all plots. These statistics can be plotted for each graph by double-clicking on the plot's respective row in the legend.
  4. The cause of traffic spikes (e.g. for the article on Michael Jackson) is inferred by analyzing metadata about the subject (e.g. from infoboxes) and searching for links to the subject on Wikipedia's main page on the dates around the traffic spike. The inferred cause of any spikes is displayed in the plotted data.
  5. The raw data is available as A) a pre-formatted, sortable HTML table, B) a CSV file or C) over a JSON API.