We are pleased to announce the release of version 0.14 of WordNet-Similarity.
This  marks the re-inclusion of the vector measure into the package! Note
that there are now two vector-style measures in the package, vector_pairs,
and now vector (which was present in versions previous to 0.13).

You can download this new release from:

http://search.cpan.org/dist/WordNet-Similarity/
http://sourceforge.net/projects/wn-similarity

The distinction between the two vector-style measures is generally simple
to explain. The vector measure builds one large context vector for each
concept that is based on a "super gloss", which is the concatentation
of the gloss of the concept and the glosses of any nearby concepts (as
specified in the vector-relation.dat file).

The vector_pair measure does not concatentate all the glosses together,
but rather builds a vector for each gloss, and then averages those
together to create the overall representation of the context. There are
of course many details about these techniques, but this is the gist of
the difference between vector and vector_pairs.

In addition, this release fixes a few problems that existed in the test
scripts due to old references to the LCSFinder module (which no longer
exists after the 0.13 rewrite).

Please let us know if you have any questions about vector or
vector_pairs, or anything else you might find in 0.14!

Enjoy,
Ted, Sid, and Jason

--
Ted Pedersen
http://www.d.umn.edu/~tpederse


------------------------ Yahoo! Groups Sponsor --------------------~--> 
Give the gift of life to a sick child. 
Support St. Jude Children's Research Hospital's 'Thanks & Giving.'
http://us.click.yahoo.com/h8TXDC/6WnJAA/HwKMAA/x3XolB/TM
--------------------------------------------------------------------~-> 

 
Yahoo! Groups Links

<*> To visit your group on the web, go to:
    http://groups.yahoo.com/group/nlpatumd/

<*> To unsubscribe from this group, send an email to:
    [EMAIL PROTECTED]

<*> Your use of Yahoo! Groups is subject to:
    http://docs.yahoo.com/info/terms/
 


Reply via email to