tag:blogger.com,1999:blog-4999686744433629981.post6485366313052833943..comments2024-02-14T22:06:08.851-08:00Comments on Augmented Social Cognition Research Blog from PARC: Mapping the Contents in WikipediaEd H. Chihttp://www.blogger.com/profile/06360447323238002978noreply@blogger.comBlogger6125tag:blogger.com,1999:blog-4999686744433629981.post-65982019406770043842009-04-23T19:22:00.000-07:002009-04-23T19:22:00.000-07:00Lisa T.:
Thanks for your suggestions. Since much ...Lisa T.:<br />Thanks for your suggestions. Since much of the work we do for WikiDashboard is extra-curricular (we have real jobs trying to push these kinds of technologies in the Enterprise), I am not sure how much of this will make it into WikiDashboard eventually, but we'll be trying.Ed H. Chihttps://www.blogger.com/profile/06360447323238002978noreply@blogger.comtag:blogger.com,1999:blog-4999686744433629981.post-50117817372793783362009-04-23T19:09:00.000-07:002009-04-23T19:09:00.000-07:00Sage:
The problem was that some topic categories h...Sage:<br />The problem was that some topic categories had more articles than others. For example, there were a lot more articles on "History and Events" than on "Philosophy and thinking".<br /><br />What we wanted to measure, instead, was given an article, how likely it would be for it to be contentious in that topic area. That's why we normalize the sum of the conflict scores with the number of article-category assignments in that topic (which should be almost the same as the # of articles in that topic).<br /><br />I don't think I was very clear in the post about this, so I will go and change this in the text.Ed H. Chihttps://www.blogger.com/profile/06360447323238002978noreply@blogger.comtag:blogger.com,1999:blog-4999686744433629981.post-51468587263956008232009-04-23T18:50:00.000-07:002009-04-23T18:50:00.000-07:00RDH: In our previous CHI paper, we developed a met...RDH: In our previous CHI paper, we developed a metric for measuring the intensity of conflicts. We called it the Conflict Revision Count (CRC). This was precisely how we apply it, so the graphs here contains some of the weighting you suggested. That is, a conflict with many revisions would be weighted more heavily as a conflict.<br /><br />Here is the reference:<br />http://asc-parc.blogspot.com/2007/05/conflict-and-power-structure-in.html<br /><br />http://www-users.cs.umn.edu/~echi/papers/2007-CHI/2007-Wikipedia-coordination-PARC-CHI2007.pdfEd H. Chihttps://www.blogger.com/profile/06360447323238002978noreply@blogger.comtag:blogger.com,1999:blog-4999686744433629981.post-61107047364236382262009-04-20T02:39:00.000-07:002009-04-20T02:39:00.000-07:00I like this Ed - could you use some sort of sensit...I like this Ed - could you use some sort of sensitivity encoding to show the scale of intensity of conflicts e.g using the grayscales we used all those years ago in the attribute explorer you could look at how many different conflicts individuals were involved in. Not sure maybe you have this in Wikidashboard already. Anyhow I love the fact that philosophy in a tiny number of posts causes the most controversy! Maybe that is why there are so few posts they haven't got round to doing any more yet they are too busy arguing!!Lisa T.https://www.blogger.com/profile/18071757048646304824noreply@blogger.comtag:blogger.com,1999:blog-4999686744433629981.post-68215061166099008302009-04-19T20:30:00.000-07:002009-04-19T20:30:00.000-07:00This is wonderful! I look forward to seeing this ...This is wonderful! I look forward to seeing this in WikiDashboard.<br /><br />One thing I'm confused about is the description of the normalized conflict. It says that "to determine the relative degree of conflict (or “contentiousness”) per article we normalized by the<br />number of article-category assignments in a topic." Does that mean that random Philosophy content is expected to have twice as much conflict as random People content, but that there is 15 times as much conflict total People conflict as Philosophy conflict?<br /><br />That's what I gather from the paper, but you state in the blog post that ""philosophy" and "religion" have generated 28% of the conflicts each."<br /><br />Could you help me understand what "normalized conflict" means?Sagehttps://www.blogger.com/profile/04259090314712198514noreply@blogger.comtag:blogger.com,1999:blog-4999686744433629981.post-19806931970644351672009-04-19T06:16:00.000-07:002009-04-19T06:16:00.000-07:00Interesting.
The next logical step would be to dev...Interesting.<br />The next logical step would be to develop some metric for measuring the intensity of conflicts, then weighing them accordingly. <br /><br />While some topic categories may produce many, small conflicts, other produce few but far more intense ones.<br /><br />For instance, one of the worst edit wars in Wikipedia history was over the proper name for Gdansk/Danzig. Yet geography and places generate only 2% of the conflicts.<br /><br />The metric could take into account: A)The number of individuals involved, B)The conflict's duration, C) How many posts are made, D) and on how many talk pages, E) How far up the Wikipedia chain of command did it go before it was <I>resolved</I>.<br /><br />Counting the number of individuals involved could be seen as tricky due to the use of <I>sock puppets</I> (illicit, multiple accounts owned by the same user, for those who don't speak Wikinese:). But I would argue that in this case it actually helps us measure conflict intensity, since one must be pretty passionate about a subject in order to go to such lengths (along with the risk of being caught and banned) in an attempt to "win".<br /><br />Also, I recommend using number of posts made, rather than how much data space is used, because a few "long winded" posters could skew results and make what is actually a tea pot tempest between a few "Stem winders", look like a major conflict.<br /><br />Thank you for this study and your time.Unknownhttps://www.blogger.com/profile/05153035278996283899noreply@blogger.com