Folksonomies: How we can improve the tags

The discussion on folksonomies continues. Folksonomies—like controlled vocabularies—are here to stay, and I would like to focus on how we can make them work better.





One thing I have noticed is that I suck at tagging my links in delicious; I draw a blank, and frequently just repeat words from the title. The result is that I have a whopping 220 tags that I have used to tag only one link, and on tag that is used on 47 links, and hardly anything in between. Clay is better, with 94 tags that he has only used once, and one tag used 211 times. Liz has a better distribution, but does show the same trend. (complete data)





This is a problem, because, at least on my part, it is more an artifact of my shortcomings as a tagger, than of the links I am tagging. And the whole purpose tagging and categorization is to divide-and-conquer, so if you categorize everything in the “Other” category, you have achieved nothing.





I suspect that the expertise that we lack, is the one that professionals like Lou Rosenfeld has. But instead of re-hiring the professionals to do controlled vocabularies for us, I believe there are things we can do with the software to empower amateurs to be better taggers.



Some tricks used by the professionals are:





So here are my ideas for making us better taggers:





  • Suggest tags for me. A Google Suggest-style interface will help familiarize people with the universe of existing tags, so they can use an existing tag rather than invent their own. It would also reduce typos and inconsistencies, like “blog” vs. “blogs”, and it might serve as inspiration to get past the obvious tags. The pool of tags suggested from could be a weighted list of my own tags, my friends’ tags, all tags, and tags other people have already used for this link.



  • Find synonyms automatically. In the browsing interface, Flickr is pretty good about showing related tags. Why not show these related tags when I am tagging a photo, making it easy for me to add the ones that apply. They could even do a lookup on WordNet for more synonyms. Since the related tags in the browsing interface feeds off of tags used on the same images on the input side, this would also help make strong ties between tags stronger.



  • Help me use the same tags others use. When doing both the Google Suggest and the synonyms above, show the most used tags in a larger size than less used tags. There is value in people using the same tag for the same thing, and we want to encourage that, without in any way preventing people from choosing a different tag if they want to.



  • Infer hiearchy from the tags. If a large number of the images tagged with “car” are also tagged with “volvo”, and very few images tagged with “volvo” are tagged with something other than car, then it’s reasonable to assume that a Volvo is a special kind of car, thus inferring hierarchy. This might not work in practice, although if people know that this is how things work, that might help. Another idea is to use the idea of related tags. If people are encouraged to use “social software” (2 tags) instead of “socialsoftware”, then when you browse “software”, “social” would show up as a related tag. But if the “social” tag was linked to “software+social” instead of just to “social”, then you would effectively have not a hierarchy, but a network, which is even better.



  • Make it easy to adjust tags on old content. If the above and other ideas work, people’s tagging skills should improve over time. So why not augment the browsing interface so that it’s very easy for me to add or remove tags from my images or links right there, e.g. from a list of suggested tags on the page. A good incentive to retag my content is when I am searching for a link through one tag, e.g. “buenosaires”, but it isn’t there, so I find it under, say, “argentina”, there ought to be a link to let me add the “buenosaires” tag to that link.



This is just a handful of ideas from the let’s-try-it-and-see-if-it-works department. Maybe some of them are too complicated, or produce bad results in practice. Surely there are many other ideas to try. But at least it is a start. And in parallel with improving the tagging experience, we can also do a lot to improve the browsing experience, e.g. with collaborative filtering through a trusted network.

9 comments

Lawrence Krubner
 

The easy part is the "suggest tags for me". I've done that with my own little project, accumulist.com, and I've an open API so that anyone can get the tags from Accumulist and have them auto-suggested. The hard part is hierarchy. How can this be auto-determined? This seems, more than most other things you mention, to scream out for human experts. I did think at one point that hierarchy could be determined from relatedness, but my own experience on accumulist.com showed me how wrong I was. I thought that if you looked at all the tags that were used in conjunction with other tags, ordered by their frequency, you could get a pretty good idea of hierarchy. But on Accumulist.com we've 5 intelligent people trying to tag things in an intelligent manner and yet, to take on example, <a href="http://www.accumulist.com/index.php?whatPage=showRelatedTags&whatTag=xml">for the tag "xml"</a>, we get these as related items: software, rss, apple, ld times, quicktime, web 2.0, api, tags, microsoft, web design, podcasting, google, smil, windows, podcast, webdevelopment, itunes, ajax, mp3, mac, dave winer, amazon, podcasts, itunes My friend Peter's interest in podcasting tilts the results. I think all the other things you mentioned will be solved, but hierarchy remains unsolvable without the help of human experts.
Read more
Read less
  Cancel
Mark Aufflick
 

Tag suggestion:<p> There are a few ways I have seen this done - one was used by nutritiuos (no longer around) which told you what tags had been used by other people who logged the same url.<p> This method is obviously only useful for something already tagged by others.<p> Another I have seen is to ask a user to pick from a list of key tags (given a fixed problem domain, eg. photos), which the system then expands on by saying "Other items in this repositary with the tags you chose have these secondary tags: ...". Not only does this encourage consistancy and thoroughness in tagging, it also makes it much more time efficient (a big problem with tagging existing sets of data).<p> This method is only useful once a decent repositary is tagged to start with and it works best for a fixed content domain.
Read more
Read less
  Cancel
Mark Aufflick
 

Hmm - i've only slightly modified what you already suggested (memo to self: read an article carefully before comenting...)
Read more
Read less
  Cancel
Lars Pind
 

Hi hfx_ben Thanks for pointing that out about the Volvo. I corrected it. /Lars
Read more
Read less
  Cancel
rmom
 

Lars, How would you feel about adopting the scheme developed by librarians for "tagging" books as they place them on the shelves. They have a comprehensive system for identifying information. Why should we reinvent the wheel? Granted that not everyone has access to "Sears List of Subject Headings by the H.W. Wilson Company which is over 800 pages long! I like your idea that some suggested tags could appear as you are thinking of your personal tags or a site, much like it is done in del.icio.us. Robin Martin
Read more
Read less
  Cancel
anthony
 

hey lars i was wondering if you have written any papers on the topic you just discussed because our group might make use of those ideas you said in our thesis or if you can provide any documents relating to social bookmarking. thnx in advance
Read more
Read less
  Cancel
swingers
 

Looking for dates, sex, swingers, <a href=http://swingers.ads6x.com >swingers</a> 3 somes and much more
Read more
Read less
  Cancel
hid kits
 

very very nice infoamations thansk realy
Read more
Read less
  Cancel
hid kits
 

very very nice infoamations thansk realy
Read more
Read less
  Cancel

Leave a comment