Results on PubMed 200k RCT

[Franck-Dernoncourt]

Did you try running your code on PubMed 200k RCT? I'm curious to see how your model compares with the LSTM+CRF baseline PubMed 200k RCT contains (F1-score = 91.6 on test set).

[jind11]

I have another architecture that works on PubMed 200k RCT dataset, which obtains 93.9 F1 score. This work is still under review at EMNLP.

[Franck-Dernoncourt]

Thanks, very nice, looking forward to reading it!

…

On 23 July 2018 at 00:16, Di Jin ***@***.***> wrote: I have another architecture that works on PubMed 200k RCT dataset, which obtains 93.9 F1 score. This work is still under review at EMNLP. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#1 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAA748sbMpChVc8VOA7U1jHKoWCLRVKjks5uJXhXgaJpZM4VaFow> .