Re: blog: semantic dissonance in uniprot

2009/3/27 John F. Madden <john.madden@duke.edu>:
> Pat et al.,
>
> It sounds like people sometimes have an irresistible itch to say that "A is
> similar to B", but this statement as such has very little semantic content.
>
> Perhaps it's not really intended as a statement that has a truth value, but
> rather as a record of somebody's feelings.
>
> The semantic web can certainly serve as a repository for recording one's
> feelings, and this might even be useful. (If I were and astrophysicist, for
> example, I might be quite interested in Stephen Hawking's intuitions about
> some problem I was working on.)
>
> So what would you say about an rdf:property called, say,
> "http://www.example.com/intuit#similarTo" that could be used simply to post
> a record that somebody intuited a "similarity" between two things?
>
> It would have little utility for inferencing, unless one were to write a
> custom application (i.e. not OWL) to do so. But it might have utility as a
> semantic web "bookmark" for relationships that could be interesting
> candidates for future formalization.

It would be infinitely better than seeAlso where you have no idea what
the intent is, other than it isn't owl:sameAs or anything else in
between.

Cheers,

Peter