I created around 50 Data samples on Patent Dataset containing 3 entities as place, author, patent_num.
Data
Output I am getting
Using: "iob_tagging_type":"box_and_within_box_level"
author entity is getting extracted in most of the test samples but not the other entities and for few of the test sample there is no prediction at all.
All the entities are equal in number.(for every document these entities exist)
Queries:
1) How can I improve the result?
2) Is it because of feeding low sample dataset?
3) What is the acceptable no. of data samples to yield the better result?
I created around 50 Data samples on Patent Dataset containing 3 entities as
place
,author
,patent_num
. DataOutput I am getting
Using: "iob_tagging_type":"box_and_within_box_level"
author
entity is getting extracted in most of the test samples but not the other entities and for few of the test sample there is no prediction at all.All the entities are equal in number.(for every document these entities exist)
Queries: 1) How can I improve the result? 2) Is it because of feeding low sample dataset? 3) What is the acceptable no. of data samples to yield the better result?