Page MenuHomePhabricator

[Analysis] can the existing models detect a violation of norms/guidelines?
Open, LowestPublic


This is already captured via the damaging / goodfaith distinction.

We haven't done any analysis of edits that are categorized as damaging while also being goodfaith.

It would be nice to know what types of edits very commonly fall into this category of "probably damaging" and also "probably good-faith".

This task is done when an analysis of randomly sampled damaging&goodfaith edits is posted on-wiki.

Event Timeline

Halfak created this task.Jan 11 2016, 11:58 PM
Halfak raised the priority of this task from to Needs Triage.
Halfak updated the task description. (Show Details)
Halfak moved this task to Active on the Machine Learning Platform (Current) board.
Halfak added a subscriber: Halfak.
Restricted Application added subscribers: StudiesWorld, Aklapper. · View Herald TranscriptJan 11 2016, 11:58 PM
Halfak set Security to None.
Halfak triaged this task as Lowest priority.Aug 18 2016, 9:39 PM