Hi Gil, Good question. I think Amruta's note provides lots of good starting points. To some extent we hope the SenseClusters documentation answers some of the basic questions as to what the package is and what it does, but if it doesn't or is confusing in some respects, please let us know, we are always trying to refine and improve the documentation.
One further comment - SenseClusters doesn't really have any relation to WordNet. It is based on using raw text as input, and clustering together similar contexts. WordNet is a pre-existing resource that organizes concepts according to is-a and other relations. The main difference is that WordNet is manually constructed and maintained, while SenseClusters wants to deal with raw text only. There is one connection, conceptually at least, between WordNet and SenseClusters. The WordNet::Similarity package (which is used in the second article you mention) produces measures of similarity between concepts based on WordNet's information and maybe augmented with some corpus statistics. SenseClusters produces measures of similarity for clusters of similar concepts that might be associated with particular concepts. In short, WordNet::Similarity measures similarity and relatedness between concepts based on WordNet's manually created content, while SenseClusters measures the similarity of contexts. It's possible to extrapolate from SenseClusters output to something like WordNet::Similarity output, but exactly how to do that remains a very interesting question that we are looking at (but don't know much about as yet). I hope this helps a little. Let us know if further questions arise! Cordially, Ted On Fri, 9 Jul 2004, Gil Vidals wrote: > I'd like to learn more about SenseClusters and would like to read more. I've > already read: > > 1) WordNet (the yellow book from MIT) > 2) Ted Pedersen's "Maximizing Semantic Relatedness to Perform Word Sense > Disambiguation" > > What else do you recommend? I don't understand sense clusters and am itching > to learn more. > > > [EMAIL PROTECTED] > Position Research, Inc. > Search engine results by research > tel: (760) 480-8291 fax: (760) 480-8271 > www.PositionResearch.com -- Ted Pedersen http://www.d.umn.edu/~tpederse ------------------------------------------------------- This SF.Net email sponsored by Black Hat Briefings & Training. Attend Black Hat Briefings & Training, Las Vegas July 24-29 - digital self defense, top technical experts, no vendor pitches, unmatched networking opportunities. Visit www.blackhat.com _______________________________________________ senseclusters-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/senseclusters-users
