[Corpora-List] New Dataset Available: Over 200, 000 Word Pair Similarity Judgments

[Jordan Boyd-Graber]

Dear all,

Over the last few years, we've been collecting sense-specific
similarity ratings collected from trained undergraduates and users of
Amazon's Mechanical Turk (we're releasing about 100,000 pairs from
each source, each pair rated by at least three people).  If you're
interested in downloading the data, please visit:

http://wordnet.cs.princeton.edu/downloads.html

Thanks,

Jordan Boyd-Graber
jbg at princeton.edu

_______________________________________________
Corpora mailing list
Corpora at uib.no
http://mailman.uib.no/listinfo/corpora