Re: What use the AstroOntology

On Tuesday 06 March 2007 11:27, Eric Saunders wrote:
> Two approaches to codifying knowledge have been discussed. In the
> classical philosophy of ontologies

[...]
> The fuzzy, tag-cloud approach of Flickr et al. takes a totally
> different approach.

Today, I released a system that seeks to marry these to "extremes", albeit in a very different context than in astronomy.

My current work, I develop the community site at http://my.opera.com/ where users also tag quite frequently, most notably blog posts and photo albums.

Both tags and ontologies have strengths and weaknesses. Often, tags doesn't have enough meaning outside of the context that it was originally created, even the person who created it might not have thought enough about its semantics. And that's fine, it is an annotation, and it is better than nothing.

There has not been time for real research, but from watching how our users are using tags, I suspected it is not all that useful for the community as a whole, allthough it is very useful for the individual user and small groups.

I think that astronomers may be a little more careful about using the right term for the same concept, and so make tags a little more useful than for the average blogger. Nevertheless, I think a pure-tags approach is likely to suffer from many of the same weaknesses as it does in this community.

To make it useful to a larger group, the community as a whole, and in fact the entire web, I have therefore created a Semantic Web application on the top of the tag system we have. It allows each user to map their tags to (a) large ontologies, for the time being, Wordnet.

I have described some aspects of it in a blog post here: http://my.opera.com/semweb/blog/2007/03/08/marrying-folksonomies-and-taxonomies Please feel free to create an account to try it out.

I believe that this is the way to go: Develop a good ontology for astronomy, but leave the controversial issues to the communities. Then, develop tools that lets astronomers tag resources, but also declare clearly what they mean by those tags.

I suspect that reasoners would rely mainly on the ontologies, but of course, fields undergoing controversial times may also find direct analysis of tag clouds useful.

Best,

Kjetil

-- 
Kjetil Kjernsmo
Programmer / Astrophysicist / Ski-orienteer / Orienteer / Mountaineer
kjetil-at-kjernsmo.net
Homepage: http://www.kjetil.kjernsmo.net/     OpenPGP KeyID: 6A6A0BBC