"final" official results?

[Ted Pedersen]

Greetings yet again,

As the camera ready deadline approaches, I am wondering if the
"official" results of the task will be modified in any way from what
we currently have?

If they are, I might suggest the following as possible additions :

1) average number of clusters found per system

2) random baseline in supervised case (ideally the same random as used
in the unsupervised evaluation)

3) unsupervised results (fscore, precision, recall) for the full
27,132 instances...

Thanks!
Ted

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

[Eneko Agirre]

Hi ted,

good suggestions!

the official results will stay the same, but we will add the following
(numbers from your message):

1) it's already in the task description paper

2) we prefer to add a note saying that random (with a small number of
clusters) is nearly equivalent to MFS.

3) it's already in the task description paper. The ranking does not vary.

We are checking a few things, we'll be back with more comments about the
supervised evaluation tomorrow.

eneko

--
---------------------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