Hi,
Would the extra entities detected within the tweet be penalized i.e if the number of entities in tweet according to your training data are 2, then along with those 2 entities if we detect another 2 more entities not considered according to training data, how would be evaluation in that case. Would the correct linkage of extra entities be penalized.
Could you please tell the evaluation criterion.
Thanking you,
Sandeep Panem
IIIT Hyderabad MS By Research
Hyderabad
India.