Research:Ideas/Automated broadcasting of newly categorized stubs

From Meta, a Wikimedia project coordination wiki
Jump to navigation Jump to search


Light-Bulb icon by Till Teenck.svg

This page documents a proposed research project.
Information may be incomplete and may change before the project starts.

Rationale[edit]

If you knew that someone just created a Wikipedia stub on the topic of your dissertation, on a disease you're familiar with as a physician, on the most recent movie by your favorite director, or on a landmark in your home town -- would you consider expanding it?

Newly created articles on niche topics often lack a critical mass of contributors to help bootstrap their development. As a result, they tend to remain in a stub or start class, when they are not nominated for deletion for lack of notability. There definitely exists a critical mass of experts on the planet for any conceivable topic of encyclopedic relevance: this expert population could be reached and invited to participate in expanding new articles, right after their creation. The problem is then how to route information about the creation of these articles to the most relevant audience of expert contributors.

Goals[edit]

This project aims to programmatically extract topical information about new stubs and craft messages to be automatically broadcast via social media to generate targeted contribution campaigns.

Methods[edit]

  • The mechanism by which newly created stub are categorized and assessed by WikiProjects (see for example WikiProject Medicine's article assessment statistics) could be used to broadcast to a large audience of subject-matter experts the existence of such articles, using the WikiProject name as an identifier.
    • All revisions to a project's talk namespace will be parsed.
    • Revisions adding or modifying a {{WikiProject *|class=<assessment class>|importance=<importance class>}} template will be extracted (alternatively, the output generated by w:User:WP_1.0_bot could also be used). The reason not to use actual the first categories added is that their granularity might not be appropriate to identify a target audience, while WikiProjects (as groups of contributors sharing the same topical interest) could provide an appropriate middle ground.
  • Alternatively, the almost-universal syntax of the stub template could be used in all languages: it would probably be enough to use the arguments of stub templates as hashtag (they're typically topic and date); or to scan the addition to stub category tree and use the names of categories.
  • Alternatively, some established method of identifying Wikipedia article topics could be used.
  • Revisions will be parsed in real time using sources like the IRC feeds or RCStream and those matching the above criteria will be used to craft a short message to be broadcast on social media like Twitter. The name of the WikiProject will be parsed and converted into a hashtag. Templates for specific topics or topic areas could be customized. For example:
Wikipedia-logo-v2.svg
Wikipedia Stub Bot
@wpstubs

Someone just created a new Wikipedia article on "Alveolitis" #Medicine. Are you an expert on this topic? Help expand it! https://en.wikipedia.org/w/index.php?title=Alveolitis&src=wpstubs

1:37 PM - 31 Jul 2014

Wikipedia-logo-v2.svg
Wikipedia Stub Bot
@wpstubs

Football fans: someone just created a new Wikipedia article on "Royal Dutch Football Association" #Football #Netherlands. Help expand it! https://en.wikipedia.org/w/index.php?title=Royal_Dutch_Football_Association&src=wpstubs

2:13 AM - 1 Aug 2014

Proof of concept[edit]

A working proof of concept was created by User:Theopolisme (source). The bot is broadcasting tweets via the @wpstubs handle.

Impact[edit]

  • Measuring the impact of this initiative can be done by monitoring traffic referred to these articles from social media (matching specific campaign parameters like wpstubs) (easy) or edit saves resulting from visits to these articles referred from social media (more complicated).
  • The Campaigns extension, currently being redeveloped, could also be used to measure the volume and quality of these edits, grouping them by topic. This is more complicated as it requires getting contributors to sign up first.

Potential stakeholders[edit]

Support needed[edit]

  • Developers with expertise monitoring real-time Wikipedia changes and feeding them into Twitter bots.


See also[edit]

References[edit]