WNUT17 - train data & eval script patches

[Leon Derczynski]

Hi,

Two small changes have been made to the training data and eval script in the past few days (thanks to Benjamin Heinzerling for the latter). It is recommended you re-download them to the latest version.

All the best,

Leon

[Hardik Meisheri]

Can you please provide some detailed readme file for training set?

[Leon Derczynski]

Hi Hardik,

Thanks for waiting! You'll be pleased to know that the README is now available too - sorry for the delay. Check this link:

http://noisy-text.github.io/2017/files/README

[Hardik Meisheri]

Thanks leon it will be very challenging to develop a system in such a short time. Let me try what can I do.

Also can you inform us where we can submit results, and is there any leaderboard or anything os such sorts?

[Leon Derczynski]

Hi Hardik,

Yeah, the dev data was late, admittedly. I hope the training data and scoring scripts didn't come too late.

Result submission will be via a Google Form.

Leon

--
You received this message because you are subscribed to the Google Groups "Workshop on Noisy User-generated Text (WNUT)" group.
To unsubscribe from this group and stop receiving emails from it, send an email to wnut+unsubscribe@googlegroups.com.
To post to this group, send email to wn...@googlegroups.com.
Visit this group at https://groups.google.com/group/wnut.
To view this discussion on the web visit https://groups.google.com/d/msgid/wnut/2ae49aa4-fcf0-4f31-ba6b-367b64b3373b%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--

Research Fellow, Department of Computer Science

University of Sheffield, UK

http://www.derczynski.com/

http://twitter.com/leonderczynski