[Most-ai-contest] Fine tuning Answer-Type-Matching-Indictor

[kysu]

Since rough matching information is better than wrong matching information, we might need to reduce the number of entity-types (say, around 6). Also, we can fine tune Answer-Type-Matching-Indictor into 3 types {Exact-Match, Possible-Match, Unmatch}. If two entity-types are frequently confused, then they will be assigned "Possible-Match" (if they are different but in the same confusing-set).

 

KY

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.iis.sinica.edu.tw/pipermail/most-ai-contest/attachments/20200404/9d9559bf/attachment.html>