Metric learning + Adaptive Decision Boundary

thuiar / TEXTOIR

TEXTOIR is the first opensource toolkit for text open intent recognition. (ACL 2021)

https://github.com/thuiar/TEXTOIR

MIT License

185 stars 30 forks source link

Metric learning + Adaptive Decision Boundary #46

Closed petrLorenc closed 6 months ago

petrLorenc commented 1 year ago

Hello, is it possible to include also this work - https://github.com/tgargiani/Adaptive-Boundary. I am one of the co-author and it would be interesting to compare with your methods (which were used as root idea for our algorithm) Best regards Petr

HanleiZhang commented 1 year ago

Hi, thanks for your interests. You could integrate it into the TEXTOIR framework and pull request :)

petrLorenc commented 1 year ago

Great, I will probably try it as it seems that our approach published one year ago is pretty similar (metric learning + ADB) to your DA-ADB approach but we obtain slightly higher numbers than numbers you are reporting. So I am interested if there is some noise in our measurements or if we outperform DA-ADB.

I will try to find the time to do the measurements - I will reach you if there will be any problems