We are pleased to announce the release of WordNet::SenseRelate::WordToSet, version 0.03. This is a Perl module that allows you to find out which sense of a given word is most related to a set of words. Unlike TargetWord or AllWords, the order of the words is not considered, and in many respects it's a much simpler operation.
For example, you can find out which sense of "game" is most related to the set of words "turkey deer boar", and then which sense of "game" is most related to "monopoly chess checkers". You could do that via the command line interface as follows: wordtoset.pl game turkey deer boar wordtoset game checkers chess monopoly By default similarity/relatedness is measured by WordNet::Similarity::lesk, but you can specify any of the measures in the command above to get a different measure: wordtoset game turkey deer boar --type WordNet::Similarity::jcn WordToSet has not had a new release since June 2005, and version 0.03 is meant to get the module "caught up" to our other modules with respect to standards of documentation and testing. I've updated the documentation somewhat, made the code examples found in the synopsis sections such that they can be cut, paste and run, and added some additional test cases. Please give WordToSet a try if you haven't already. If you already use WordNet-Similarity or one of the other SenseRelate packages, then you already have everything you need, so it's especially easy in that case to download and install. WordToSet is very simple to use from the command line as you can see from the above, and nearly as easy to write programs with (look at the synopsis examples to see how). You can find WordToSet at CPAN or sourceforge : http://senserelate.sourceforge.net for links, or go directly to it at CPAN at : http://search.cpan.org/dist/WordNet-SenseRelate-WordToSet/ Please let us know of any questions or concerns. Enjoy, Ted -- Ted Pedersen http://www.d.umn.edu/~tpederse
