[Wikipedia-l] Categories considered harmful

Sun Jun 20 11:57:04 UTC 2004

Ray Saintonge wrote:
> I checked those links and am not any further ahead.  I think that I can
> understand how thesauri might be relevant; however, I'm puzzled by how
> they have imported a term from metaphysics to serve their purpose.

I think the term "ontology" entered the scene from artificial
intelligence (AI) research, which soaked up some hype from Tim
Berners-Lee's "semantic web" ideas.  There is a pretty good article at
http://en.wikipedia.org/wiki/Ontology_%28computer_science%29
People from information retrieval (IR) or library and information
science (LIS) tend to prefer "thesaurus". In some cases people make a
difference between the two, but I think they are mostly synonyms or
overlapping. WordNet (www.globalwordnet.org and
http://www.cogsci.princeton.edu/~wn/) is probably the biggest
real-world open-source example.

It is possible to divide the world into people who believe in data
structures (those who build WordNet, Dmoz, other thesauri, and the
semantic web), and those who believe in free text (those who build
Wikipedia and the old non-semantic web).  Wikipedia's categories is a
border-crossing.  Some of the people who could/should contribute might
have been turned off by Wikipedia's previous category-free approach.

Sorry for being so vague.  I took computer science at a university
where information retrieval wasn't even taught.

-- 
  Lars Aronsson (lars at aronsson.se)
  Aronsson Datateknik - http://aronsson.se/