This is not a tag line!
Spamming del.icio.us
I am subscribed to the Cocoa tag RSS feed on del.icio.us and I just received what I consider as SPAM: someone posted a bookmark to his/her site with a huge list of seemingly popular tags in effect spamming everyone subscribed to these tags… I’m only linking to the site in question to document (I used the rel="nofollow" attribute to prevent the spammer from benefiting from the link) what I am afraid is the first (though I hope it’s the last) example of del.icio.us spam.
Spamming del.icio.us this way is similar to the old meta trick, which used to be successful in the pre-Google days to ensure high-ranking on search results. It is much, much more efficient due to del.icio.us’ popularity though and is a “clever” use of folksonomies: people subscribed to tags are very likely to check sites marked with such tags. Moreover, RSS feeds ensure wide distribution of the link. I could see spammers efficiently use that technique to make people check their sites out.
The question now is what can be done to prevent such an obvious perversion (or is it?) of social bookmarking? One potential way would be to restrict the number of tags that could be assigned to bookmarks but this is obviously not that useful: spammer would use the 15 (or whatever the tag number limit is) most popular tags to still efficiently propagate their crap. Which would in turn lead to a corruption of the folksonomy: if most popular tags are spammed with irrelevant crap, the tags become less meaningful…
More elaborate solutions could be used by performing content analysis and compare the newly tagged link to the most popular links with the same tag (assuming that these links are relevant, which they should if you are to trust folksonomies) and check that there is some overlap content before allowing the tag to be used. Actually, this wouldn’t even work as illustrated by the cocoa tag itself: cocoa is most known as the substance from which chocolate is issued but on del.icio.us (with its probably slightly geeky crowd) most links tagged with “cocoa” refer to Mac OS X’s programming framework. Though rare, I have seen bookmarks to cocoa-the-substance-related websites in the midst of the flow of cocoa-the-OS-X-API links in the cocoa RSS feed. With the content analysis described above, people wanting to assign the cocoa tag to chocolate-related sites would probably be barred (no pun intended) from using it…
I am not quite sure what can effectively be done to prevent unscrupulous users from spamming others but it sure would be interesting to think about it. In the meant time, it’s a sad day for folksonomies. I am afraid that if you are interested (like me) in subscribing to tags via RSS, you might need to resign yourself to receiving, in a near future, a bunch of irrelevant crap along with legit links… 
According to this post on the del.icio.us discuss mailing list, Joshua is aware of the issue and has code to deal with it in the next upgrade of del.icio.us. Very cool, though it’d be really interesting to know what kind of solution has been implemented.
No related posts.
Related posts brought to you by Yet Another Related Posts Plugin.
| Print article | This entry was posted by Chris on May 4, 2005 at 16:39, and is filed under technology. Follow any responses to this post through RSS 2.0. Both comments and pings are currently closed. |
Comments are closed.
about 4 years ago
I would say that you need to broad your perspective a little bit. del.icio.us is a personal bookmark manager in the first place, and everyone can tag any URL with whatever tags he wants.
If I post some URL with tag Cocoa, and the target site is not related to OS-X, or do not fall into your definition of “cocoa”, that means I am spamming you? That’s ridiculous, I’d say. Maybe my friend has a nick cocoa, and I will tag URL’s which I think he would consider interesting with this tag. I don’t care, if you or someone else subscribed to this tag rss. You seems to want to control a meaning of certain tag, but del.icio.us just do not work this way.
about 4 years ago
I think you misunderstood my point. I’m not trying to dictate how people use their tags. I didn’t think I implied that the “cocoa” tag should *only* be used for bookmarks related to the OS X API. In fact my point was that using a content analysis mechanism based on the most popular bookmarks using a given tag to prevent tag spam would not work for the very reason that tags have different meaning for different people. I use the “cocoa” tag for Apple’s Cocoa-related bookmarks, you use it to tag pages you think your friend nicknamed cocoa would find interesting, while others would use it to tag pages pertaining to chocolate…
The tag spammer I’m refering to was using a large number of popular tags to mark the page (s)he wanted to promote, none of the tags were actually relevant to the actual content of the page. I don’t care about getting the occasional post in my RSS feed that is not related to Apple’s Cocoa: as you say, it is part of the game. However, my RSS aggregator shows for each entry the complete list of tags that were used for that particular entry. The bookmark that ticked me off was using around 50 different tags which were all popular tags when it was obvious (by visiting the linked page, which was the spammer’s intention) that the page didn’t have anything to do with any of the tags the spammer used.
Interestingly, the bookmark in question seems to have been deleted from the del.icio.us database since I haven’t been able to find it again to write this reply…
about 4 years ago
Actually, there it is: http://del.icio.us/dubord. As you can see, pretty spam-like, wouldn’t you say?