News for hackers. Stuff that matters.
Re:Wikipedia Deploys AI to Expand Its Ranks of Human Editors by cooper12 on Thursday December 03, 2015 @08:12AM (#10669357)
A bit more of a technical treatment from the Wikimedia blog yesterday: https://news.ycombinator.com/item?id=10663208
Re:Wikipedia Deploys AI to Expand Its Ranks of Human Editors by TazeTSchnitzel on Thursday December 03, 2015 @10:47AM (#10670181)
Thank you, I was wondering how it differed from Wikipedia's existing vandalism detector systems like ClueBot.
Re:Wikipedia Deploys AI to Expand Its Ranks of Human Editors by jcla1 on Thursday December 03, 2015 @12:07PM (#10670733)
This kind of thing isn't anything new... here's a similar system that is less geared toward vandalism, but more to pointing out structural and lexical flaws in the articles: https://github.com/jcla1/wikipedia_analyser

(Disclaimer: I'm the author)

It too uses neural-nets (and an up to 3-gram model) to detect flaws...

Re:Wikipedia Deploys AI to Expand Its Ranks of Human Editors by milesstevenson on Thursday December 03, 2015 @12:54PM (#10671188)
What is this kind of thing written in, I wonder. How cool :)
Re:Wikipedia Deploys AI to Expand Its Ranks of Human Editors by boxy310 on Thursday December 03, 2015 @01:13PM (#10671375)
>Using a set of open source machine learning algorithms known as SciKit Learn—code freely available to the world at large—the service seeks to automatically identify blatant vandalism and separate it from well-intentioned changes.

It's a Python library, available for free download [1]. I've got a book or two on it I've been meaning to read.

[1] http://scikit-learn.org/stable/

Re:Wikipedia Deploys AI to Expand Its Ranks of Human Editors by ngram on Thursday December 03, 2015 @10:26AM (#10670029)
All Watched Over by Machines of Loving Grace :-)