Note: I (@SSalgaonkar-WMF) am writing this request on behalf of @KStoller-WMF. Please consider Kirsten the submitter and primary owner of this request.

Scoping details

• Use case: As described in T389871, the goal is to develop and deploy a structured task to help newcomers add incoming links to orphan articles, improving the visibility and integration of these orphan articles within Wikipedia’s content network. An orphan article is defined as "an article with no links from other pages in the main article namespace".
• At a high level, we expect the user flow to look like the following: (1) A newcomer goes to the newcomer homepage, and they see the Suggested Edits module. (2) The newcomer sees a suggested edit that contains a structured task. (3) The structured task shows an orphan article, as well as a word or phrase from a related article that we suggest should be linked to the orphan article. (4) The newcomer decides whether to accept or reject the suggestion. (5) If the newcomer rejects the suggestion, they can provide a rejection reason.

• Model purpose: The goal of the model is to surface specific words or phrases that should be linked to orphan articles. We think, but have yet to prove, that the existing Add-a-Link models can be re-used for this purpose, perhaps with a filter added on the client side to only display results that link to orphan articles. If we find that the existing model cannot be reused for this use case, we should rename this ticket/work stream.

• Goal: Our main goal is to impact constructive activation, by giving newcomers easy and accessible structured tasks that will help them make successful first edits. We also think that this structured task will help with content visibility. Orphan articles lack incoming links, making them hard to discover and navigate. Addressing this issue aligns with research highlighting the importance of internal linking for article visibility and engagement. Studies have also shown structural biases in linking, particularly affecting articles about women and underrepresented topics. Machine learning models can assist editors by identifying high-quality link recommendations.

• Prior art: As mentioned above, we believe that the Add-a-Link models can be reused for this application. The broader UX around Suggested Edits already exists, but we'll need to make some design decisions about how this specific type of task gets surfaced to newcomers.

Prioritization details - Coming soon!

• Timing: When are you hoping to launch an experiment or feature using this model? How flexible is your timeline? Is there any other planned work that's blocked by this experiment or feature?
• KR impact: Which KRs are enabled by this project, and how critical is this project for moving the needle on those KRs?

Other comments

• [Optional] Model requirements: If you have any specific concerns around model performance (latency, cost, etc.) or model output quality (likelihood of false positives, ability to detect all possible instances, etc.), please note them here.
• [Optional] Is there anything else you'd like to share?