Feedback from InooLabs
Open, Needs TriagePublic
Actions

Assigned To

None

Authored By

	KinneretG
	Feb 8 2022, 12:17 PM

Description

Hello Platform Team!
I am adding a request for support that we received from InooLabs, a Swedish edtech company. Would appreciate your assistance in providing them with answers.

Share documentation/Roadmap notes about the new versions of the API that are in development.
Is there any way (using any API’s today, old or new) with which inooLabs could get hold of the reference indicators in the plaintext, without parsing through wikitext.

For example, when searching for Hawaii the first paragraph contains this sentence:
Hawaii is also one of four U.S. states that were once independent nations along with Vermont, Texas and California.[11]
But using API’s we get this sentence:
Hawaii is also one of four U.S. states that were once independent nations along with Vermont, Texas and California.

Same scenario as above, but for hyperlinks. Is there a way to find out which words in the text have other Wikipedia articles?

Related Objects
Search...

		Status	Subtype	Assigned	Task
		Open		None	T302284 API Platform User Support
		Open		None	T301239 <API User Support> Request/Feedback from InooLabs

Event Timeline

KinneretG created this task.Feb 8 2022, 12:17 PM

Restricted Application added a subscriber: Aklapper. · View Herald TranscriptFeb 8 2022, 12:17 PM

• DAbad renamed this task from Request/Feedback from InooLabs to <API User Support> Request/Feedback from InooLabs.Feb 22 2022, 1:35 PM

• DAbad edited projects, added API Platform; removed API Platform (API Platform Roadmap).

• DAbad moved this task from Incoming to Must do now on the API Platform board.

• DAbad added a parent task: T302284: API Platform User Support.Feb 22 2022, 1:37 PM

Additional Notes:

Currently using our data dumps as training data and a third-party service to fetch real-time content to summarize on the fly (https://pypi.org/project/Wikipedia-API/), but that service doesn’t do everything they want/need, so they came to us.
looking to fetching page content + metadata on contributors) is doable & documented in the API portal (saw the catalog just went up – nice!)

– except getting references on articles (which get stripped by PyPI, and we don’t seem to have clean/easy ways of serving up, either). I saw that Evan started to think about this, but it was backlogged after he left WMF: https://phabricator.wikimedia.org/T247858.