User talk:Cobi/Archives/2011/January

Re: ClueBot NG and the Cup

No, it wouldn't, but only because Cluebot would have zero points. Edit counts aren't worth anything anymore. :P Ed ^{[talk] [majestic titan]} 02:36, 1 January 2011 (UTC)

Someone made a mistake (not you)

The info you have to introduce Alchemy is not fully accurate. Arabic isn't even a language that was introduced until 1000's of years after the first people of Alchemy so you can not drive a word from a language that didn't exist. — Preceding unsigned comment added by Wonderingraven (talk • contribs) 16:11, 5 January 2011 (UTC)

Sorry about this, new user appears to think ClueBot contributed some info he's trying to replace with unsourced POV stuff. Ian.thomson (talk) 18:30, 5 January 2011 (UTC)

Unnecessary revert

I shut down ClueBot NG. He reverted my edit, which was unnecessary because the redirect should be linked to its own article instead of another. Best regards, Mgangku (talk) 23:12, 5 January 2011 (UTC)

Personally I wouldn't have shut the Bot down for making one mistake, something like that I would've reported as a false positive. --5 albert square (talk) 23:31, 5 January 2011 (UTC)

The Cluebot knowledgebase

Hi, I have seen Cluebot do some pretty nice reverts. In terms of where it should be in 5 years, I suggest looking at it as an expert system application with comprehensive rules. I have a few ideas. It may be easier if I suggest the ideas and you add them, as appropriate, rather than start coding myself.

For instance, this edit included a "random string" and could be caught by a rule of the type:

IF

user is an IP

confidence in randomness of string > 75% and

familiarility with IP < 20%

THEN

confidence in vandalism > 70%

Eventually the final confidence should be derived using inexact reasoning. It is pretty easy actually, if approached the right way. These rules can then be gradually fine-tuned over the years and in 5 years you will have a very comprehensive knowledgebase.

I would also suggest tapping into wordnet eventually and that will open a totally new door.

How do we start this conversation? Cheers. History2007 (talk) 10:53, 8 January 2011 (UTC)

Hey History2007. ClueBot NG uses an artificial neural network, which I do not believe allows for a rule exactly like that in the core of the bot, which "measures" for vandalism. If memory serves, the bot already takes into account if an edit is made by an IP user or a registered user. It also already checks against known words. In regards to that specific edit, I can't figure out why ClueBot NG did not revert it. I cannot find a log of it on IRC nor was the IP editor reverted before on that article. Since I'm just a simple talk page stalker who thinks ClueBot NG's core is mostly a black box, you can pick up a conversation with Cobi/Crispy best by hopping onto IRC on ClueIRC's #cluebotng. If you do not have an IRC client, check out Mibbit. -- SnoFox^(t|c) 18:07, 8 January 2011 (UTC)

(nonauthoritative answer) I'm sure Crispy or Cobi have a better answer, but I'd be curious if the ANN didn't basically qualify already. After all, it currently takes into account many different inputs, so why couldn't it be fed a hypothetical case that only presented information about, say 4 inputs, and produce, within some margin of error--an reasonable output of its vandalism 'score'? I suspect that History2007's endgoal of inexact reasoning is already happening, and Cluebot is an expert system (though still on its learning curve). I'm not sure what wordnet could add, but it's an interesting concept. There was some anti-vandalism work which focused on language processing out of UIowa ([1] [2]), but it was apparently pretty computationally expensive--in other words, good but slow. Ocaasi (talk) 19:02, 8 January 2011 (UTC)

This is a repost of a discussion on User_talk:Crispy1989. I've responded there. In response to your last note, Ocaasi, CBNG already uses some NLP techniques, and we're adding more. Previous approaches may have been slow due to substandard code, poor choice in programming language (ie, scripting language), and/or improper/impractical technique application. Crispy1989 (talk) 19:08, 8 January 2011 (UTC)

72.14.194.1

Hi. 72.14.194.1, which you unblocked, has been reblocked as a proxy.   — Jeff G.  ツ 18:05, 10 January 2011 (UTC)

Cluebot NG Dataset Review Interface

Hi, I just requested access to the Dataset Review Interface, but realized that my Google account needs to be reset to my current email address, as its still pointing to a very old email address I do not control anymore. I will make amends and resubmit a request with the proper email address shortly. thanks! ^Jguy_Talk^Done 19:57, 10 January 2011 (UTC)

Archiving error

You probably missed my message since there's all of the stuff with false positives going on, but I'm having an issue with my user talk page being archived. ClueBot has archived all of the 2010 stuff without a problem, but it hasn't done anything to stuff dating before 2009. Do you know what the problem is? --Kevin W./Talk•CFB uniforms/Talk 07:09, 15 January 2011 (UTC)

I don't see anything on your talk page from 2009. ClueBot III can't archive something that's not there. :) -- SnoFox^(t|c) 19:43, 16 January 2011 (UTC)

What are you talking about? There are six headings from 2009 on my talk page. --Kevin W./Talk•CFB uniforms/Talk 19:59, 16 January 2011 (UTC)

Oh, I didn't realize the talk link in your signature lead to a subpage. I'll be going away then... -- SnoFox^(t|c) 20:28, 16 January 2011 (UTC)

Yep, they're two separate pages. The second has been archived without a problem, but it's fairly new without anything before 2010. --Kevin W./Talk•CFB uniforms/Talk 22:47, 16 January 2011 (UTC)

Cobi, I don't understand why the max archive size had anything to do with it, since the stuff that has yet to be archived isn't going in the existing 2010 archive anyway. --Kevin W./Talk•CFB uniforms/Talk 23:01, 16 January 2011 (UTC)

It archives by CURRENT_DATE - age, where age is the age you specified in the ArchiveThis template. So, technically, it will dump it in the 2010 archive unless you manually archive it. -- Cobi^(t|c|b) 23:03, 16 January 2011 (UTC)

Ok, so be it. --Kevin W./Talk•CFB uniforms/Talk 04:33, 17 January 2011 (UTC)

It still hasn't worked. This is very weird. --Kevin W./Talk•CFB uniforms/Talk 19:04, 17 January 2011 (UTC)

Screw it, I've gone ahead and manually archived everything. It's just under 25K bytes so I've set that as the max size. Anything new should get archived without a problem. One more question, though: I've got the format set as just Y, so what happens if an archive goes over the limit? How would it differentiate the new archive, title-wise? --Kevin W./Talk•CFB uniforms/Talk 20:57, 19 January 2011 (UTC)

It won't -- it simply won't archive if it would go over the limit. If you want numbered archives you have to use the format for that. -- Cobi^(t|c|b) 21:13, 19 January 2011 (UTC)

Ok, that's fine. I'll just have to do some finagling then. Thanks for all of your help. --Kevin W./Talk•CFB uniforms/Talk 21:29, 19 January 2011 (UTC)

Cluebot-NG and Pending Changes

AN/I post you might be interested in here. -- Ocaasi (talk) 11:09, 22 January 2011 (UTC)

difference btw pinscreen and pinscreen animation

Hi Cobi, why my edit on Pinscreen page had been reverted??? i am in the process to improve the accuracy of pinscreen, / pin art. i have many reasons to clarify the difference between pinscreen and pinscreen animation!!! please read the fact of ward fleming and his patented invention, which i added in the origin section in pinscreen animation thank you! Nip888 (talk) 04:41, 24 January 2011 (UTC)

A response has been posted at the first place you posted. -- SnoFox^(t|c) 01:12, 25 January 2011 (UTC)

ACC access

Hi Cobi,

I would like to have access to ACC. I sent my reconsideration request to accounts-enwiki-l@lists.wikimedia.org. But I have not gotten any response and it's been almost a week since I sent it.

So I wanted to appeal this decision because although my edit count is low I feel I can be trusted with this tool. It would allow me to help other users and there by taken some of the weight off the shoulders of the admins. And I do met the minimum guidelines for approval found here. http://en.wikipedia.org/wiki/Wikipedia:Request_an_account/Guide

Qualifications for using the interface, Usually a user should simply be trusted by the approving administrator. To qualify as an ACC user, you should meet the following minimum requirements:

• You must not have had a recent block on the English Wikipedia. * You must not have any history of account-related abuse, such as sockpuppetry.
• You must have read, understood and agreed with the ACC guidelines.
• Your account on Wikipedia must be more than 60 days old.

If given access I will use it responsibly because I understand its a toll not a toy and access to it is a privilege.

experience." at 2010-12-22 17:05:42. Related information (please include this if appealing this decision) user_id: 775 user_name: TucsonDavid user_onwikiname: TucsonDavid user_email: tucsondavid@live.com

Thanks for your help in advance. TucsonDavidU.S.A. 22:16, 28 January 2011 (UTC)

never mind its being handledTucsonDavidU.S.A. 00:07, 29 January 2011 (UTC)