[WePS-Semeval07] WePS-3 Data release and schedule adjustment

[Javier Artiles]

Dear WePS participants,

The training data and evaluation script for Task-2 and is currently available in our website. We have uploaded as well Intelius person name frequency data from WePS-1 and WePS-2 names that can be useful for participants in task-1. http://nlp.uned.es/weps/weps-3/data

Since this release represents a delay in the planned schedule we propose to move the start of the evaluation period one week from 1st June to 7th and the submissions due from 15th June to 21st. The other dates in the schedule would remain the same (that is, the release of the evaluation results would still be 15 July).
Please let us know if you have any objections or other proposal.

Note that the papers submission is due the 15 August, in sync with the CLEF Labs working notes submission.

Javier Artiles.

On behalf of the WePS organizers.

--
------------------------------------------------------------
Dr. Javier Artiles Picón
Departamento de Lenguajes y Sistemas Informáticos
ETSI Informática, UNED

Home page: http://nlp.uned.es/~javart

Phone:        +34 91 398 8106
Fax:            +34 91 398 65 35
------------------------------------------------------------

--
You received this message because you are subscribed to the Google Groups "Web People Search Task" group.
To post to this group, send email to web-people-...@googlegroups.com.
To unsubscribe from this group, send email to web-people-search...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/web-people-search-task?hl=en.

[Javier Artiles]

Dear WePS participants,

One of the participants has pointed out that the task-2 training was missing the file with the organizations/query/web_page table.

We have just updated the zip file in the WePS website. I'm also attaching the table to this email, so do you don't have to download again the training data. Notice that this table does not have a "category" column. Training organization names have been obtained from a bigger variety of sources. These sources did not provide a classification as in the names pool used for the trial data set.

A question that has been raised by some participants is whether there will be some labeled examples for the organizations in the test data. The answer is no, we wont provide labeled tweets for the test organization, nor there will be overlap between the names in trial/training and the test data set.

Best regards,

Javier Artiles.

On behalf of the WePS organizesr.