Information Tagging
I just came across a very interesting post by Terry Jones via the JOHO blog that I read. Jones is very concerned about the alienation of metadata from content, and that's what his business is apparently trying to remedy. Jones is probably alot smarter than me on this topic, and thus I'm not sure I understand what difference the location makes as long as it is findable and accessible. But here's the problem about metadata that I would like someone to address: how do we know the metadata is right? This may seem contrary to what many people assume, but isn't a great deal of metadata ultimately subjective? If you have a piece about the American Revolution, should it be tagged as politics or history? Of course it could be tagged both ways, but in more complex situations, addding every conceivable tag could become self-defeating.
Imagine this. Using the example of clay tablets from the Jones article, if you give 10 people a stack of 1000 tablets and ask them to create tags for the info pressed into those bricks, you will probably get 10 similar, but different, sets of tags. So which set of tags is right?
Now--just for kicks--pick only one set of the generated tags, which we will call the "right" set. Go back in time, and give each of the same 10 people the stack of 1000 tablets, all tagged according to the one chosen set. Then ask the 10 people to race to find some trivial piece of information. Probably 1 person will find it quickly, and the other nine will find the info, but slightly slower, or in a worst case senario, maybe not at all.
Alternatively, instead of choosing one "right" set of tags, let's assume that they are all "right." So take the ten sets of tagged tablets and mix them into 10 new composite 1000-tablet groups, each with all 10 different tagging rationales represented. Go back in time again and distribute these new stacks and ask the group of proto-librarians for another piece of trivia. This time, every one would have only about a 10 percent chance of finding the requested information with maximum quickness and efficiently.
Many current models of information metadata fall somewhere between these two scenarios. Thus, the creater of the metadata and the user of that data are often distinct.
Finally, let's consider one more paradigm. This time, past out a 1000-tablet set to each of our ten guinea pigs, but after they have all created their tags, ask them to exchange their stacks and allow them to modify, add, and delete eachother's tags. Exchange again and repeat. Continue until every one has contributed to the tags of every stack.
Now go back in time. How effective is this set of community tags? Can digital information technology be designed to maximize this effect? Are there better ways to make metadata more organic? Will reconnecting data and metadata help as Terry seems to suggest?
Read more about:
History , School , Technology , The Internet0 TrackBacks
Listed below are links to blogs that reference this entry: Information Tagging.
TrackBack URL for this entry: http://www.milestravis.org/cgi-bin/mt/mt-tb.cgi/104
Leave a comment