- From: Phillip Lord <phillip.lord@newcastle.ac.uk>
- Date: Fri, 30 Apr 2010 09:25:28 +0100
- To: Chris Mungall <cjm@berkeleybop.org>
- Cc: Bijan Parsia <bparsia@cs.man.ac.uk>, Owl Dev <public-owl-dev@w3.org>, sonic@tcs.inf.tu-dresden.de
Chris Mungall <cjm@berkeleybop.org> writes: > On Apr 29, 2010, at 5:58 AM, Bijan Parsia wrote: >>> Ideally the implementation would be open source and well-integrated with >>> current tools (e.g. works with the OWLAPI and/or OWLlink). I'd be willing >>> to work a little on the plumbing, but not for closed source tools. >> >> So my first question is about what you need it for and whether true LCS is >> what you need. > > The immediate application is semantic similarity (of which there are many many > measures; the ones I am particularly interested in involve calculation of a > LCS). An approximation of an LCS may be perfectly acceptable. Chris Most of the semantic similarity measures work on a graph rather than a set of axioms; despite their names, they rarely obey semantics of the ontology. For example, when I did this over GO, I just ignored the edge semantics entirely and treated all equally. Given this, the solution seems to me to be the one that Alan ruled out; get to a graph as quickly as possible. Just calculate the LCS over a reasoned hierarchy. Phil
Received on Friday, 30 April 2010 08:26:13 UTC