Wikipedia:Bots/Requests for approval/NNBot


 * The following discussion is an archived debate. Please do not modify it. Subsequent comments should be made in a new section. The result of the discussion was Symbol neutral vote.svg Request Expired.

NNBot
Operator: Nn123645 (talk)

Automatic or Manually Assisted: Automatic, with occasional supervision.

Programming Language(s): PHP Hypertext Preprocessor

Function Summary: Tag new articles that meet certain criteria with the stub template (or a template similar to it) and check for catagories and if they don't exist add uncategorized.

Edit period(s) (e.g. Continuous, daily, one time run): Continuous

Already has a bot flag (Y/N): N

Function Details: The bot will make use of the IRC Recent Change Feed to log new page creations to a database table. As pages are deleted it will remove rows from the database table accordingly. After a certian time (Currently it looks like it will be 12 hours) has passed the bot will go through the page and analyze them for certian stubbish criteria (string length, whether the page is a disambig, if the page has formatting, if the page has wikilinks, if the page has images, if the page has references, etc.) and edit the page if it meets enough of those criteria.

As long as I'm not stepping on anyone's toes I'd like to check for uncategorized and possibly some other templates down the line. I will try to see if I can't contact Wikiproject Categorization and see how they feel about the idea, though that project seems to be fairly inactive.

I'd like to add remove old inuse and underconstruction templates from pages which are obviously neither inuse or underconstruction. Since I will be checking for these templates anyways it I figure I might as well use the data.

Discussion
Just as a general note this bot is not yet complete in the code. I coded the IRC part of the bot and some of the API portion. I ran this by Wikiproject Stub Sorting, and they were ok with the idea. --Nn123645 (talk) 18:46, 7 July 2008 (UTC)

A full list of what I have come up with as possible stubbish criteria is available here. --Nn123645 (talk) 18:46, 7 July 2008 (UTC)


 * Things like word length and sentence length might be difficult due to intricacies with wikitext. A few other suggestions: don't tag pages with inuse or underconstruction or pages that have been edited in the past hour, as this could mean someone is actively working on it. I assume it will also skip pages that already have a stub tag? Mr.Z-man 21:15, 8 July 2008 (UTC)


 * Yes, disambiguation issues were a fairly big concern at wikiproject stub sorting, the way it is looking at now is this is the criteria it will have for checking a page.
 * If the page is at least 12 hours old
 * AND the page hasn't been edited by the article creator in the last 8 hours
 * AND the page hasn't been edited by anyone else in the last 2 hours
 * AND the page has no prod templates on it
 * AND the page is not tagged for speedy deletion
 * AND the page is not inuse
 * AND the page is not underconstruction
 * THEN analyze and tag the page
 * If the bot sees a page that is inuse or undercontsruction it will move it from the pending table of the database to the pending_inuse, or pending_underconstruction table. I might file a later BRFA for permission to remove old inuse templates or underconstruction templates on pages that are not obviously under construction (example: have not had an edit by the page creator in the last month two weeks and have not had an edit from anyone else in the last week).--Nn123645 (talk) 21:21, 8 July 2008 (UTC)


 * Sounds good, no objections it seems. Whenever you're ready. Mr.Z-man 23:42, 12 July 2008 (UTC)


 * As a status update I'm making progress but am not yet ready to run the trial. I will be on a trip next week until August 2nd or 3rd so expect a long delay in me replying to any questions here. --Nn123645 (talk) 21:37, 26 July 2008 (UTC)

I'd be interested to see an exact list of the criteria used and the weighting being given to each criterion. The page linked above doesn't really cut it for me I think. Martinp23 15:04, 31 July 2008 (UTC)
 * I ask this because it is very difficult to actually come up with a foolproof mechanism for identifying stubs (I remember issues trying to come up with some rules on another tool). One has to remember that different subjects have different amounts of info available to put into an article, which makes things all the more complicated.  I think having as many eyes as possible look over your algorithm can only help.. Martinp23 15:07, 31 July 2008 (UTC)
 * I am not yet done writing the heuristics. I will provide more thorough documentation and publish the source before I run the trial though I may not do it on wiki. --Nn123645 (talk) 23:17, 1 August 2008 (UTC)

Any progress? – Quadell (talk) 18:06, 8 August 2008 (UTC)
 * Yeah, though its coming along slowly. --Nn123645 (talk) 14:06, 9 August 2008 (UTC)


 * It's been over a month since you were approved for trial. Do you want to withdraw this for the time being, and reopen it when you have more time? – Quadell (talk) 13:28, 18 August 2008 (UTC)


 * I'm not totally sure, I have been making progress but as mentioned before its coming along slowly. --Nn123645 (talk) 00:18, 20 August 2008 (UTC)

Feel free to reopen this when you are ready to run the trial. BJ Talk 13:01, 1 September 2008 (UTC)


 * The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made in a new section.