If you want to find the most relevant/related sense of a word to a set
of 50 other words, you might want to experiment with
There is a utility program that comes with the package called
wordtoset.pl, and that lets you input a target word and any number of
other words, and you can find the relatedness of that target word to
all those other words.
wordtoset.pl --type WordNet::Similarity::jcn cat dog mouse horse wolf
says that I want to run the Jiang Conrath measure (jcn) to measure the
similarity of cat to dog, mouse, horse, wolf. Below we can see that
the first noun sense of cat scored the best and is most similar to
Does this sound like what you want to do?
Please let us know if you have additional questions!
On Nov 29, 2007 5:02 PM, Vesile Evrim <vesile1@...> wrote:
> As a part of my research, I would like to find the most relevant sense of a
> given word within the certain domain. As an input, I would like to enter a
> word and set of words (to determine the context of a target word) from the
> same domain. It seems that Wordnet::SenseRelate-TargetWord is suitable for
> that purpose and I installed 0.08 version of it to try. However, I am still
> not sure how I will enter my input. I tired the given example to see how the
> program works and I get the following results
> # disamb.pl senseval2.xml
> Loading WordNet::SenseRelate::Tools... done.
> Reading XML data... done.
> art.n art.40003 pop_art#n#1
> art.n art.40006 arts#n#1
> art.n art.40010 art#n#1
> art.n art.40013 art#n#1
> art.n art.40016 art#n#1
> #disamb.pl senseval2.xml | wps2sk.pl
> Loading sense index file... Loading WordNet::SenseRelate::Tools... done!
> Processing index.noun file... done.
> Reading XML data... done!
> Processing index.verb file... done!
> Processing index.adj file... done!
> Processing index.adv file... done!
> art art.40003 pop_art%1:06:00::
> art art.40006 arts%1:09:00::
> art art.40010 art%1:06:00::
> art art.40013 art%1:06:00::
> art art.40016 art%1:06:00::
> Although in the description it is mentioned that wps2sk converts the output
> of the disamb.pl to senseval-2 format it seems that the input file
> serseval2.xml is already in seneseval-2 format.
> I tried to enter another xml, txt formats as an input to disamb.pl and it
> gave an error. Then, I tried to use nameconflate (v0.16) to convert the
> plain text to senseval-2 before passing it to the program but somehow I got
> an error when I enter the words to nameconflate.pl. My question is, what
> shall I do to pass one target word and a document that consists of 50 words
> to the SenseRelate program so that I can learn which sense of the target
> word is relevant.
> Thank & Regards,
> Yahoo! Answers - Get better answers from someone who knows. Try it now.
> SF.Net email is sponsored by: The Future of Linux Business White Paper
> from Novell. From the desktop to the data center, Linux is going
> mainstream. Let it simplify your IT future.
> senserelate-users mailing list