Jump to content

Submissions/The story of building ORES review tool and how it impacted workload of patrollers

From Wikimania


Info

This is an Open submission for Wikimania 2017 that has not yet been reviewed by a member of the Programme Committee.

Submission no. 3063 - T1
Title of the submission

The story of building ORES review tool and how it impacted the workload of patrollers

Type of submission (lecture, panel, tutorial/workshop, roundtable discussion, lightning talk, poster, birds of a feather discussion)

lecture

Author of the submission

Amir Sarabadani (User:Ladsgroup)

Language of presentation

English

E-mail address

amir.tafreshi_ext@wikimedia.de

Username

Ladsgroup

Country of origin

Iran

Affiliation, if any (organisation, company etc.)

Wikimedia Germany

Personal homepage or blog
Abstract (up to 300 words to describe your proposal)

ORES is the platform to use AI in Wikimedia projects at scale. It has been worked on for quite some time. This talk is about one of it's most prominent applications, the ORES review tool, a beta feature allowing users to highlight and filter edits in recent changes or their watchlists. The first part of this session will be about explaining ORES and its applications. The second part will be about the history of ORES, obstacles we had to face and lessons we learned while making this happen on the scale of Wikipedia. This can be useful for future tool developers that want to build large-scale tools for Wikipedia. And lastly, I will show a comprehensive analysis of the impact ORES review tool had on the reduction of edits needing review and the number of reverts in periods of time and average speed of reverting vandalism in different wikis. This will include Wikidata and several languages of Wikipedia. The faster we revert vandalism, the more reliable Wikipedia is.

ORES filters the firehose. ORES generally reduces the counter-vandalism workload for patrollers by 90% by splitting edits that don't need review from those that do.
ORES highlights edits for review. Edits in recent changes that are likely to be damaging are highlighted.
Damage is easier to find. The diff for a damaging edit that was detected by ORES.
What will attendees take away from this session?

Attendees will learn more about how ORES works and hopefully gets motivated to bring this tool to their own wikis or build tools on top of the service.

Theme of presentation
Technology, Interface & Infrastructure
Length of session (if other than 25 minutes, specify how long)

25 minutes

Will you attend Wikimania if your submission is not accepted?

Yes

Special requests
None
Is this Submission a Draft or Final?
Info

This is a Completed submission for Wikimania 2017 ready to be reviewed by a member of the Programme Committee.

Interested attendees

If you are interested in attending this session, please sign with your username below. This will help reviewers to decide which sessions are of high interest. Sign with a hash and four tildes. (# ~~~~).

  1. Smallbones (talk) 20:11, 31 March 2017 (UTC)[reply]
  2. Amir É. Aharoni (talk) 09:39, 9 April 2017 (UTC)[reply]
  3. Jtmorgan (talk) 15:46, 10 April 2017 (UTC)[reply]
Retrieved from "https://wikimania2017.wikimedia.org/w/index.php?title=Submissions/The_story_of_building_ORES_review_tool_and_how_it_impacted_workload_of_patrollers&oldid=26196"