Community health initiative/Interaction Timeline/UIA research

From Meta, a Wikimedia project coordination wiki

This page will record my anonymized findings of how the User Interaction Analyser ( is used on English Wikipedia.

My methodology[edit]

  • Using Special:LinkSearch I found several instances of users discussing the tool. (query)
  • I also highlighted links to the tool using the following JS:

Sockpuppet investigations[edit]

The EIA is used very commonly in sockpuppet investigations. It's great to see this tool serving multiple purposes, but for the sake of of the User Interaction History tool I will ignore Sockpuppet investigation uses.


  • In one instance (α) a user reported to ANI that they were being hounded and harassed by another user. They provided a few links to diffs and article histories. A second user suggested they use the tool. "I would like to see a list of diffs where they started editing a page only after you did. The Editor Interaction Analyzer at is a good tool for doing that sort of research." The reporting user then provided several cherry-picked timeline links for specific articles, and three times later in the discussion provided more timeline links. In the end a mutual IBAN was enacted.
  • In another instance (β) User:A reported on User:B's talk page that User:C was hounding them. User:D provided a link to the EIA: "To see whether your complaint was actionable, I took the liberty of comparing your interactions using this tool. In the resulting page, there are links to various "timeline" interactions on one page. I saw two timelines that show recent instances of User:B following you around: article_one, and article_two. [...] It's clear User:B followed you to that discussion by looking at your contributions."
  • In another instance (γ) User:A reported User:B to ANI. User:C came to cast User:A in a bad light and User:A replied, posting an EIA link for User:A and User:C to dispute their attack: "Anyone interested in the truth [...] can also look at this Editor Interaction Analyzer tool."


  • Single article abuse vs. the big picture
    • Sometimes there is the need to just see how two users interaction on a single wiki page, but the whole story is also important.
    • Cherry-picking timeline links can be nonconstructive in an investigation. How can the timeline link page remind visitors that these users have also co-edited on other pages?
    • Question to investigate: When viewing just two users' activity on the same page, is it important to see other editors? Maybe we could visually show this other information.