[Corpora-List] Data release: Graded word meaning judgments

[Katrin Erk]

Dear all,

we are happy to announce the availability of two new datasets that
characterize word meaning in context. Both datasets consists of graded
human judgments.

In the first annotation study, we asked annotators for graded
judgments on the similarity of two usages (two occurrences of the same
target word). The second study asked annotators for graded jugments on
the applicability of WordNet senses (contrasted with the usual "winner
takes all" annotation).

The data from these two annotation studies is now freely available for
download at:

http://comp.ling.utexas.edu/people/katrin_erk/graded_sense_and_usage_annotation

Katrin Erk, Diana McCarthy, Nicholas Gaylord

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