I have a question regarding the induction of senses.
Should the system induce the senses only by using the test set? In
other words, are we allowed to use both the training and test set of
the English lexical-sample task, or any other available resource e.g.
the web or BNC
to induce the senses of target words?
Thank you,
Giannis
the idea is that you are not allowed to see sense-tagged data.
As soon as we get the data from task 17, we will release a single corpus 
with no sense tags whatsoever. Sorry about the delay, but it's not in 
our hands.
If you want, you can use any other resource, as long as it has no 
sense-tags.
Thanks for your interest!
best
eneko & aitor
-- 
---------------------http://ji.ehu.es/eneko------------------
Eneko Agirre                          PLEASE NOTE NEW E-MAIL:
Informatika Fakultatea                mailto: e.ag...@ehu.es
649 p.k. - 20.080 Donostia              fax: (+34) 943 015590
Euskal Herria / Basque Country          tel: (+34) 943 015019