Hi! I'm Bobocicada, your mentee through the Growth Team Program. My CS research at the University is closely related to what you do — anti-vandalism and content patrolling.

I'm building a tool to help newcomers understand policy violations. The idea: I collect reverts where the patroller cited a specific policy (like "rv WP:NOR" or "unsourced per WP:V"), then pair the removed text with that policy label. After training on thousands of these pairs, the model might be able to learn what policy-violating content looks like, and could warn a newcomer before they save their edits.

I know Edit Check already handles some checks like tone and citations. What I'm trying to cover is the broader range of content policies — verifiability, original research, reliable sources — which don't have automated detection yet.

There are three things I can't figure out from the data alone:

1. How reliable are the policy labels in revert summaries? When a patroller writes something like "rv WP:NOR" or "unsourced per WP:V", in your experience, is that usually the correct policy? Or do some patrollers tend to use a catch-all like "WP:V" even when the real issue might be something else? Also, are there specific editors or groups whose policy labels you'd consider especially accurate, and others you'd take with a grain of salt?

2. When newcomers are confused about why their edit was reverted, how do you usually explain which policy they violated? I ask because my tool would try to do something similar — point to the specific policy. But the model won't always get it right. In your experience, how much harm does a wrong policy explanation do compared to no explanation at all?

3. As a mentor and a patroller, do you think this tool would actually help? Or is the bigger problem that new users just don't engage with feedback?

I'd love to stay in touch and hear more about your patrolling work and experience as my project develops.

Best, Bobocicada --Bobocicada (talk) 23:08, 19 February 2026 (UTC)[reply]

Bobocicada, I see your question. This is just to let you know that I have seen it, and due to the nature of your question, I need to formulate a good answer for you. This may take me a while. I'm just letting you know that I have seen it and will get back to you ASAP. Yoshi24517 (Chat) (Very Busy) 06:26, 20 February 2026 (UTC)[reply]

Thank you so much for the acknowledgment, Yoshi24517! I really appreciate you taking the time to look into this, especially given how busy you are.

In the meantime, following suggestions from other scholars in the community, I’ve created a formal research page on Meta-Wiki to centralize the project’s goals, methodology, and the specific policies I'm focusing on (like WP:NOR, WP:LLM, and WP:PROMO):

Research:WikiPolicyBench: A Multi-Policy Violation Detection Benchmark from Wikipedia Editor Annotations

The page also includes an updated "Seeking Community Input" section that expands on the questions I asked you. Please don't feel any rush to reply, but if you have a chance to look at the research page, I would deeply value any input you might have on whether these priorities align with your daily experience as a patroller.

Looking forward to your thoughts whenever you have a moment! Best, --Bobocicada (talk) 22:15, 21 February 2026 (UTC)[reply]

Bobocicada, Just an FYI you do not need to ping me on my own user talk page, we get notified for any edit on our user talk page. I ping you so you get notified that there is a message here, because this is not your user talk page. We get notified automatically on our own talk page. In the meantime, after giving it some thought, I have thought about your questions for a bit, and I would like to include my response below:

1. "How reliable are the policy labels in revert summaries? When a patroller writes something like "rv WP:NOR" or "unsourced per WP:V", in your experience, is that usually the correct policy? Or do some patrollers tend to use a catch-all like "WP:V" even when the real issue might be something else? Also, are there specific editors or groups whose policy labels you'd consider especially accurate, and others you'd take with a grain of salt?" For the most part, for people like me, it can be reliable. Sometimes people make mistakes, and put the wrong reason, or other reasons where somebody puts the wrong thing. For more experienced editors, like me with about 60,000 edits, people should know how or what the policy violation is. If it is say, a brand new editor with only say, like 10 edits, they are not familiar with the policies on Wikipedia yet, and are prone to putting the wrong policy violation when they revert something.

2. "When newcomers are confused about why their edit was reverted, how do you usually explain which policy they violated? I ask because my tool would try to do something similar — point to the specific policy. But the model won't always get it right. In your experience, how much harm does a wrong policy explanation do compared to no explanation at all?" This is something that all Wikipedia editors struggle with, we have a policy called "Don't bite the newcomers". However, we have very many templates that we use to explain things on user talk pages. For example, when we are in our anti-vandalism tool, and we see an edit that we don't like, or constitutes vandalism, or is unsourced, or whatever, we usually press one button, and the program will revert the edit, and also leave a message on the user talk page of the person that made the edit, letting them know their edit was reverted, and for what reason. In my opinion a wrong policy explanation is better than no explanation at all, because the user would get all mad that their edit was reverted without any explanation as to why. Having at least a reason helps to calm the user down, that they know there is a reason why something was reverted.

3. "As a mentor and a patroller, do you think this tool would actually help? Or is the bigger problem that new users just don't engage with feedback?" I think for the most part, and this is just my opinion only, that users do not engage with feedback because they received a response that was "unhelpful" to them. What I mean is, for example, a person new to Wikipedia decided to make an edit, later on, it gets reverted, then the person is mad about that revert because it didn't stay, and they decide to not edit Wikipedia ever again. However, there are some people who will understand that the edit that they made was wrong, and try and fix it. There are people with good intentions out there. Sometimes I do see people that go onto experienced users' talk pages and ask why a certain edit was was reverted, and they do get helpful advice from the editor, and the user understands and moves on. Sometimes, people get really angry and start yelling at people because their edit didn't stay. People need to learn that sometimes, edits are reverted for a reason. If you calm down and understand why, it can help you become a better editor. Don't just leave Wikipedia because your edit was reverted, instead, take the time to learn and become a better editor.

I hope this has been helpful for you. If you have any more questions, do not hesitate to ask. I also ask if there are experienced editors who are watching my user talk page, who would like to chime in on this discussion as well, I invite you to do so. Good luck! Yoshi24517 (Chat) (Very Busy) 02:12, 23 February 2026 (UTC)[reply]

How do I insert in-text citations, and do they need to be placed in the exact same position when comparing English to my native language? --Ishak8 (talk) 22:46, 21 February 2026 (UTC)[reply]

Hey I just made an edit to Mrs Grundy and decided to make an account after seeing it made a temporary account, but it doesn't show that I made an edit on my account. Is there a way I can get that edit on my account? --Caseychiella (talk) 00:39, 23 February 2026 (UTC)[reply]

Caseychiella, as far as I know, and any editor can correct me if I'm wrong, but edits made using a temporary account cannot be reattributed to an actual user account. You would have had to make that edit using your full user account. Sorry about that! Yoshi24517 (Chat) (Very Busy) 02:18, 23 February 2026 (UTC)[reply]