Does anyone have any idea how long this experiment will last? Just curious. Also, would there be a way to estimate how effective this experiment is? I.e. comparison of number of deleted new pages, etc. There should be a way to measure the effectiveness of this experiment.
Thanks!
Flcelloguy
From Wikipedia, the Free Encyclopedia.
_________________________________________________________________ Express yourself instantly with MSN Messenger! Download today - it's FREE! http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/
On a similar topic -- are there any tool for automatic monitoring new page creation?
I was thinking about this in the shower today, and it didn't seem that it would be very technically difficult (for a guy who knows some VB and PHP) to hook up an (external) program which would monitor the Special:Newpages RSS feed, then check the page content itself and potentially flag it as something to attend to if it met a set of variable characteristics -- i.e., is it less than 5 words long, does it contain the word "fart" or "gay" or "penis", was it created by an anon, is it wikified, and so forth. More complicated though still quite feasible operations could involve automatically putting a sample of the content through Google and seeing if anything comes up, or potentially checking for incoming links, etc. At the end of the day you could ideally run something which would check all of those marked as "potential problems" to see if they had been edited extensively or deleted, and then make it easy for the editor running the program to take a look at what remained. All in all it wouldn't put any more stress on the servers than a user who actually checked these things manually, and would potentially catch things that were missed by other diligant admins.
Anything like this exist? If not, I might try to cobble one together in my (meager) spare time, though I warn you it'll be written in Visual Basic... Seems like it would help with at least one problem in relation to new pages, if not the more insidious one of false claims disguised as encyclopedia articles.
FF
--- Fastfission fastfission@gmail.com wrote:
On a similar topic -- are there any tool for automatic monitoring new page creation?
I was thinking about this in the shower today, and it didn't seem that it would be very technically difficult (for a guy who knows some VB and PHP) to hook up an (external) program which would monitor the Special:Newpages RSS feed, then check the page content itself and potentially flag it as something to attend to if it met a set of variable characteristics -- i.e., is it less than 5 words long, does it contain the word "fart" or "gay" or "penis", was it created by an anon, is it wikified, and so forth. More complicated though still quite feasible operations could involve automatically putting a sample of the content through Google and seeing if anything comes up, or potentially checking for incoming links, etc. At the end of the day you could ideally run something which would check all of those marked as "potential problems" to see if they had been edited extensively or deleted, and then make it easy for the editor running the program to take a look at what remained. All in all it wouldn't put any more stress on the servers than a user who actually checked these things manually, and would potentially catch things that were missed by other diligant admins.
Anything like this exist? If not, I might try to cobble one together in my (meager) spare time, though I warn you it'll be written in Visual Basic... Seems like it would help with at least one problem in relation to new pages, if not the more insidious one of false claims disguised as encyclopedia articles.
I don't think anything like that exists yet, so please code it! We may even be able to convice Google to exceed the 1000 search a day limit per IP.
-- mav
__________________________________________ Yahoo! DSL Something to write home about. Just $16.99/mo. or less. dsl.yahoo.com
On 12/5/05, Fastfission fastfission@gmail.com wrote:
On a similar topic -- are there any tool for automatic monitoring new page creation?
I was thinking about this in the shower today, and it didn't seem that it would be very technically difficult (for a guy who knows some VB and PHP) to hook up an (external) program which would monitor the Special:Newpages RSS feed, then check the page content itself and potentially flag it as something to attend to if it met a set of variable characteristics -- i.e., is it less than 5 words long, does it contain the word "fart" or "gay" or "penis", was it created by an anon, is it wikified, and so forth. More complicated though still quite feasible operations could involve automatically putting a sample of the content through Google and seeing if anything comes up, or potentially checking for incoming links, etc. At the end of the day you could ideally run something which would check all of those marked as "potential problems" to see if they had been edited extensively or deleted, and then make it easy for the editor running the program to take a look at what remained. All in all it wouldn't put any more stress on the servers than a user who actually checked these things manually, and would potentially catch things that were missed by other diligant admins.
Anything like this exist? If not, I might try to cobble one together in my (meager) spare time, though I warn you it'll be written in Visual Basic... Seems like it would help with at least one problem in relation to new pages, if not the more insidious one of false claims disguised as encyclopedia articles.
FF
Gmaxwell has a bot written in Python that usually runs in #wikipedia-en-suspectedits which reads every diff as it comes through. A diff gets announced if it contains various words considered offensive (profanity, ethnic slurs), words considered offensive if not used in context ("gay" in an article where the previous revision contained neither "homosexual" nor "gay" nor other similarly related words), added exclamation points (usually at the least poor encyclopedic style!), and speedy delete notices; he's taken a few other suggestions for improvement.
Unfortunately it's not running at the moment as toolserver is still down, but it's a pretty useful tool to aid RC patrol.
(Yes, Gmaxwell is my significant other, but that's not why I'm plugging the bot, honest. :-))
-Kat [[User:Mindspillage]]
-- "There was a point to this story, but it has temporarily escaped the chronicler's mind." --Douglas Adams
Yes, I think I recall this now that you mention. It's not quite what I had in mind, and I seem to recall someone had used it to create some sort of magical RC monitoring program?
The streaming of such information would seem to be the first step to figuring out innovative ways to use it. I made a suggestion to BugZilla that the deletion log also be set up with an RSS feed -- you could use it to keep track of which articles streamed out of Newpages you didn't have to deal with -- though I have no pretension that such a thing should be an urgent priority since there is no infrastructure to take advantage of it at the moment.
FF
On 12/5/05, Kat Walsh mindspillage@gmail.com wrote:
On 12/5/05, Fastfission fastfission@gmail.com wrote:
On a similar topic -- are there any tool for automatic monitoring new page creation?
I was thinking about this in the shower today, and it didn't seem that it would be very technically difficult (for a guy who knows some VB and PHP) to hook up an (external) program which would monitor the Special:Newpages RSS feed, then check the page content itself and potentially flag it as something to attend to if it met a set of variable characteristics -- i.e., is it less than 5 words long, does it contain the word "fart" or "gay" or "penis", was it created by an anon, is it wikified, and so forth. More complicated though still quite feasible operations could involve automatically putting a sample of the content through Google and seeing if anything comes up, or potentially checking for incoming links, etc. At the end of the day you could ideally run something which would check all of those marked as "potential problems" to see if they had been edited extensively or deleted, and then make it easy for the editor running the program to take a look at what remained. All in all it wouldn't put any more stress on the servers than a user who actually checked these things manually, and would potentially catch things that were missed by other diligant admins.
Anything like this exist? If not, I might try to cobble one together in my (meager) spare time, though I warn you it'll be written in Visual Basic... Seems like it would help with at least one problem in relation to new pages, if not the more insidious one of false claims disguised as encyclopedia articles.
FF
Gmaxwell has a bot written in Python that usually runs in #wikipedia-en-suspectedits which reads every diff as it comes through. A diff gets announced if it contains various words considered offensive (profanity, ethnic slurs), words considered offensive if not used in context ("gay" in an article where the previous revision contained neither "homosexual" nor "gay" nor other similarly related words), added exclamation points (usually at the least poor encyclopedic style!), and speedy delete notices; he's taken a few other suggestions for improvement.
Unfortunately it's not running at the moment as toolserver is still down, but it's a pretty useful tool to aid RC patrol.
(Yes, Gmaxwell is my significant other, but that's not why I'm plugging the bot, honest. :-))
-Kat [[User:Mindspillage]]
-- "There was a point to this story, but it has temporarily escaped the chronicler's mind." --Douglas Adams _______________________________________________ WikiEN-l mailing list WikiEN-l@Wikipedia.org To unsubscribe from this mailing list, visit: http://mail.wikipedia.org/mailman/listinfo/wikien-l
You would't be talking about '''CryptoDerk's VandalFighter''', would you? Why not add to that code, which already exists? That would be a huge help, and would be a big improvement. -Mysekurity http://en.wikipedia.org/wiki/User:Mysekurity
On 12/5/05, Fastfission fastfission@gmail.com wrote:
Yes, I think I recall this now that you mention. It's not quite what I had in mind, and I seem to recall someone had used it to create some sort of magical RC monitoring program?
The streaming of such information would seem to be the first step to figuring out innovative ways to use it. I made a suggestion to BugZilla that the deletion log also be set up with an RSS feed -- you could use it to keep track of which articles streamed out of Newpages you didn't have to deal with -- though I have no pretension that such a thing should be an urgent priority since there is no infrastructure to take advantage of it at the moment.
FF
On 12/5/05, Kat Walsh mindspillage@gmail.com wrote:
On 12/5/05, Fastfission fastfission@gmail.com wrote:
On a similar topic -- are there any tool for automatic monitoring new page creation?
I was thinking about this in the shower today, and it didn't seem that it would be very technically difficult (for a guy who knows some VB and PHP) to hook up an (external) program which would monitor the Special:Newpages RSS feed, then check the page content itself and potentially flag it as something to attend to if it met a set of variable characteristics -- i.e., is it less than 5 words long, does it contain the word "fart" or "gay" or "penis", was it created by an anon, is it wikified, and so forth. More complicated though still quite feasible operations could involve automatically putting a sample of the content through Google and seeing if anything comes up, or potentially checking for incoming links, etc. At the end of the day you could ideally run something which would check all of those marked as "potential problems" to see if they had been edited extensively or deleted, and then make it easy for the editor running the program to take a look at what remained. All in all it wouldn't put any more stress on the servers than a user who actually checked these things manually, and would potentially catch things that were missed by other diligant admins.
Anything like this exist? If not, I might try to cobble one together in my (meager) spare time, though I warn you it'll be written in Visual Basic... Seems like it would help with at least one problem in relation to new pages, if not the more insidious one of false claims disguised as encyclopedia articles.
FF
Gmaxwell has a bot written in Python that usually runs in #wikipedia-en-suspectedits which reads every diff as it comes through. A diff gets announced if it contains various words considered offensive (profanity, ethnic slurs), words considered offensive if not used in context ("gay" in an article where the previous revision contained neither "homosexual" nor "gay" nor other similarly related words), added exclamation points (usually at the least poor encyclopedic style!), and speedy delete notices; he's taken a few other suggestions for improvement.
Unfortunately it's not running at the moment as toolserver is still down, but it's a pretty useful tool to aid RC patrol.
(Yes, Gmaxwell is my significant other, but that's not why I'm plugging the bot, honest. :-))
-Kat [[User:Mindspillage]]
-- "There was a point to this story, but it has temporarily escaped the chronicler's mind." --Douglas Adams _______________________________________________ WikiEN-l mailing list WikiEN-l@Wikipedia.org To unsubscribe from this mailing list, visit: http://mail.wikipedia.org/mailman/listinfo/wikien-l
WikiEN-l mailing list WikiEN-l@Wikipedia.org To unsubscribe from this mailing list, visit: http://mail.wikipedia.org/mailman/listinfo/wikien-l
On 12/6/05, Fastfission fastfission@gmail.com wrote:
On a similar topic -- are there any tool for automatic monitoring new page creation?
With the current level of NP patrol I don't think it is needed.
-- geni
On 12/5/05, geni geniice@gmail.com wrote:
On 12/6/05, Fastfission fastfission@gmail.com wrote:
On a similar topic -- are there any tool for automatic monitoring new page creation?
With the current level of NP patrol I don't think it is needed.
I fully disagree; and even if it's sufficient at this level it won't be when it's twice what it is now.
Kelly
Kelly Martin wrote:
On 12/5/05, geni geniice@gmail.com wrote:
On 12/6/05, Fastfission fastfission@gmail.com wrote:
On a similar topic -- are there any tool for automatic monitoring new page creation?
With the current level of NP patrol I don't think it is needed.
I fully disagree; and even if it's sufficient at this level it won't be when it's twice what it is now.
I'm with Kelly on this. Any tools that can assist new page patrolling is beneficial.
People doing NP patrol do an excellent job, considering what they are up against. However, the sheer volume of inbound nonsense means that even if NP patrollers eliminate most of it, significant amounts survive.
One approach I sometimes take is to go 24 to 48 hours into the new pages log and do a sort of second round patrolling. The second round can be more careful because the first round (standard) NP patrol has taken care of the bulk of the problems.
Fastfission wrote:
On a similar topic -- are there any tool for automatic monitoring new page creation?
Yeah, I created something for this purpose once but I didn't maintain it. It's part wf pywikipedia and is called followlive.py. It needs maintenance because it currently does only {{stub}}, {{cleanup}} or {{delete}}.
Gerrit.