Re: Adaptation and Personalization

These are good points, Adam,
That's why we created Txikipedia at the Basque Wikipedia, to serve a younger audience: https://eu.wikipedia.org/wiki/Txikipedia:Azala. We don't rely on AI summaries, while they are interesting, but on the same spirit of Wikipedia. The results are very interesting, as we are having younger editors writing for their peers.

Best,

Galder
________________________________
From: James Heilman via Wikimedia-l <[email protected]>
Sent: Monday, April 6, 2026 5:36 AM
To: Wikimedia Mailing List <[email protected]>
Cc: James Heilman <[email protected]>
Subject: [Wikimedia-l] Re: Adaptation and Personalization

A few of us over many years tried to shift the reading level of the leads of EN WP's medical articles and we managed to succeed.

https://en.wikiversity.org/wiki/WikiJournal_of_Medicine/Readability_of_English_Wikipedia%27s_health_information_over_time

However there was a great deal of pushback with some insisting the reading level be increased again. This was one of the reasons a few of us switched our efforts over to mdwiki.org<http://mdwiki.org/> so we could more easily continue writing for a general audience.

James

On Mon, Apr 6, 2026 at 5:15 AM Adam Sobieski via Wikimedia-l <[email protected]<mailto:[email protected]>> wrote:
Wikimedia-l,

For whom should large-scale repositories of knowledge, e.g., Wikipedia, phrase their content: article-subject-matter experts, laypeople, or students? Similarly, what about technical documentation? Digital textbooks?

While English Wikipedia provides a vast and accurate source of information, its reading level is said to be generally high, requiring a college-level comprehension. For easier access, Simple English Wikipedia was created, simplifying vocabulary and sentence structures to improve readability and comprehensibility for a broader audience.

What if English Wikipedia and Simple English Wikipedia could be the same Web resource? What if end-users could choose to send site- or article-specific "adaptation parameters", e.g., via HTTP request headers to English Wikipedia servers, to receive one of a multitude of article phrasings available at each article's URL?

That is, what if users could direct their Web browsers to https://en.wikipedia.org/wiki/Photosynthesis , for example, and receive an article custom-tailored to their preferred reading levels, language fluencies, and background knowledge (or to a proximate user-modeling cluster or stereotype)? What if these technical features simply worked as users navigated hyperlinks connecting articles to one another, exploring a vast adaptive encyclopedia?

As considered, there would be opportunities for finer-grained adaptations. Articles could have more than two phrasings, e.g., English and Simple English.

While there are multiple technical approaches for delivering these features, each approach presenting its own technical challenges, I am envisioning that generative artificial-intelligence technologies can be of use for adapting, personalizing, and custom-tailoring audience-independent wikitext documents into cachable, reusable, audience-dependent articles served to those end-users who opt to configure and send their "adaptation parameters" to Wikipedia’s Web servers. Other users would, as considered, receive articles' default phrasings.

I am excited about fine-grained adaptation and personalization, in particular as these pertain to educational resources. For these reasons, I recently started a new Adaptation and Personalization Community Group [1].

I am reaching out to invite anyone interested in these topics to browse and participate in our ice-breaker technical discussion forum [2] and to join the new group [1]. Thank you!


Best regards,
Adam Sobieski

[1] https://www.w3.org/community/adaptation/
[2] https://github.com/w3c-cg/adaptation/issues

_______________________________________________
Wikimedia-l mailing list -- [email protected]<mailto:[email protected]>, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and https://meta.wikimedia.org/wiki/Wikimedia-l
Public archives at https://lists.wikimedia.org/hyperkitty/list/[email protected]/message/6TY7PBVX36CKDGD7KTNPL45YHW5Y43AT/
To unsubscribe send an email to [email protected]<mailto:[email protected]>


--
James Heilman
MD, CCFP-EM, Wikipedian

_______________________________________________
Wikimedia-l mailing list -- [email protected], guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and https://meta.wikimedia.org/wiki/Wikimedia-l
Public archives at https://lists.wikimedia.org/hyperkitty/list/[email protected]/message/L5DQFFGZPCTZCETQIT5SBYRS5BQ73W5J/
To unsubscribe send an email to wikimedia-l-leave-RusutVdil2icGmH+5r0DM0B+6BGkLq7r@public.gmane.org