Sockpuppets? Surely this can't be more than .00000000000001% of the user base?<br><br><div class="gmail_quote">On Dec 19, 2007 7:05 PM, Daniel Arnold <<a href="mailto:arnomane@gmx.de">arnomane@gmx.de</a>> wrote:<br>
<blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">Hello Luca,<br><br>> 1. The trust coloring rightly colored orange (low-trust) some<br>
> unreliable content,<br><br>Yes I was lost in translation. ;-)<br><br>> 2. and the Wikipedia people were quick in reverting it.<br><br>Yes.<br><div class="Ih2E3d"><br>> Note that we also highlight as low trust text that is by anonymous
<br>> contributors. The text will then gain trust as it is revised.<br><br></div>One possible weakness came into my mind after I also read your paper. Your<br>algorithm is perhapes a bit vulnerable to "sock puppets". Imagine person A
<br>with one account and person B with two accounts. Both have a medium<br>reputation value for their accounts. User A edits an article with his account<br>4 times. All 4 subsequent edits are taken together and the article has a
<br>maximum trust value according to the user's reputation. User B makes as well<br>4 edits to an article but switches between his accounts and thus "reviews"<br>his own edits. If I understand your algorithm correctly the sock puppeted
<br>article is trusted more than the other one.<br><br>Quite some time ago I reflected how to avoid incentives for sock puppets in<br>karma systems without even knowing which accounts are sock puppets:<br><a href="http://meta.wikimedia.org/wiki/Meritokratischer_Review" target="_blank">
http://meta.wikimedia.org/wiki/Meritokratischer_Review</a> (sadly in German ;-).<br>The system described there differs from your approach but the idea on how to<br>avoid incentives for sock puppets without even knowing who a sock puppet is
<br>could perhapes adapted to your system.<br><br>The basic idea for a sock puppet proof metric is is that a person has only a<br>limited amount of time for editing (I don't consider bots cause they are<br>easily detectable by humans). A single person needs the same time for
e.g. 4<br>edits (in the following I assume each edit has the same length in bytes)<br>regardless how much accounts are used but two different people with each 2<br>edits only need half of the (imaginary) time (you don't need to measure any
<br>time untits at all).<br><br>So the maximum possible reliability person B can apply to the article with its<br>two accounts (let us say each acount has 2 edits = 4 total edits) has to be<br>the same as the one which is possible with person A's single account (4
<br>edits). So in general two accounts with each X edits should never be able to<br>add more trust to an article than one person with 2*X edits (note: edit count<br>number is only for illustration, you can take another appropriate
<br>contribution unit).<br><div class="Ih2E3d"><br>> About 2, I am very glad that bad edits are quickly reverted; this is the<br>> whole reason Wikipedia has worked up to now.<br>> Still, it might be easier for editors to find content to check via the
<br>> coloring, rather than by staring at diffs.<br><br></div>That's certainly true for articles not on your watchlist (or bad edits that<br>were forgotten and are still the latest version).<br><div class="Ih2E3d">
<br>> - Finding when flagged revisions are out of date (there may be a new<br>> high-trust version later)<br><br></div>Well as I said I'd love to see flagged revisions and your system combined (in<br>a way described by my previous mail). An automated system probably always has
<br>some weaknesses some clever people can abuse but it is very fast, while a<br>hand crafted system depends on the speed of individual persons but is much<br>harder to fool.<br><div class="Ih2E3d"><br>> BTW, as the method is language-independent, we look forward to doing the
<br>> same for wikipedias in other languages.<br><br></div>Good to know. :-)<br><br>Arnomane<br><br>_______________________________________________<br>Wikiquality-l mailing list<br><a href="mailto:Wikiquality-l@lists.wikimedia.org">
Wikiquality-l@lists.wikimedia.org</a><br><a href="http://lists.wikimedia.org/mailman/listinfo/wikiquality-l" target="_blank">http://lists.wikimedia.org/mailman/listinfo/wikiquality-l</a><br><br></blockquote></div><br>