Taxonomy versus SearchPosted: September 27, 2007
I am being so web 1.0 as I write this post … Whoever is thinking about taxonomy today, what with the emergence of folksonomy as a major mover in the content space. But, this is a question worth asking. And, a question which probably doesnt have a straight-forward answer. Or, if there is one, then I dont know it (anyone out there could give it to me?). There is a lot written out there about the benefits of taxonomy vis vis folksonomy (and both of them have many), but search …
The question I am asking … Is search relevant? For example, when google searches for a document, it doesnt necessarily depend on tags to search. Can we extend this to say that if we can make search more powerful (assuming we can), would that make taxonomy more redundant? Doubtless, taxonomy makes search simpler. But, this brings up another question … How much taxonomy is enough taxonomy.
Penny Crosman, for example, mentions 10 reasons for using taxonomy which make sense to a certain level, but beyond that, I find it hard to agree with them. For example, one of the reasons she mentions is to “make corporate assets more accessible”. However, there is sch a thing as too much taxonomy. It is tempting to create a very rigorous structure, which is too detailed. While this does improve search, it also tends to discourage people from contributing their nuggets to a KM system (the why do they want me to fill out all this information syndrome).
The point here is that there are folks who prefer to use a defined, hierarchical navigation to find a document they want, and then there are folks who detest navigation, as can be seen by the popularity of search. Which means, that the design of such a system has to tread a middle path quite carefully, so as not to lose out on the advantages of taxonomy, while at the same time, not overdoing it.
The question then is … How? I would look at a methodology which is based on the reason for the documentation. This is to ask the question … Why are people contributing documents here. Why are people searching for documents here. I think the answer to the question of where the middle path lies in the answers to questions like these. Making the entire discussion more complex is the fact that its not just about what people do, what are the things which could make it much easier for them to find documents (a la disruptive innovation!).
The other extreme is, of course, no taxonomy, and a powerful search (assuming such a thing exists). This is a surefire way to make sure that people are not able to find what they are looking for. Because, unless the search is powerful, they may not find what they are looking for. To give you an example, I was search for KM on technorati some time back, and I came across quite a few hits for F1 racing, and automobile reviews (km/hr is what technorati was showing me).
The way I see it … Taxonomy should cover the basics … The reasons why people contribute, and why people search … Anything more would, in all probability, be counter-productive.