Friday, January 19th, 2007, 6:42 pm
Handling SPAM in Large Social Content Management Systems
ANDLING SPAM where collaboration is involved can be a hard task, whereas in non-social systems, erasing SPAM (or content with vandalistic intent) may be easy(ier). There is virtually no interaction with the context in the latter case. With social Web sites and forums this equates to nuking a spammer along with some innocent people in the spammer’s surroundings. In the case of collaborative editing, such as this one which I discovered yesterday, judgment becomes difficult as well. It’s marginal because free of speech, authority, and censorship ought to be weighed.
I would like to throw in some random thoughts: earlier this morning, in Netscape, I had a look at some tag clouds and saw some arcane phrases dominating the cloud. I decided to dive in and I found that these sometimes came from spammers (brute-force tagging) whose account had been suspended. This led me to pondering if by leaving these links in tact spammers have an incentive to return. There’s a true dilemma where SPAM comes to a content management system. It’s much easier when it comes to E-mail, unless excessive marketing and bulk mail which fills up boxes is concerned.
Returning to my main concern, Netscape thrives in high figures. High figures may be good for site vanity and integrity of the whole community (e.g. not purging votes and comments of non-spammers), but what about a bury-like feature (a la Digg) that keeps these submissions of out the reach/sight of search engines (and sometimes human visitors too)? As it stands, spammers are currently being rewarded by going under the ‘radar’ of Web spiders.