홈
탐색
nav.journalClub
트렌드
더보기
synapse
⌘+K
언어
한국어
한국어
March 3, 2026
Structure dependent weighted loss function for triple extraction using BART model
MS
Manu Shrivastava
Kyushu Institute of Technology
KS
Kosei Shibata
Kyushu Institute of Technology
HW
Hiroaki Wagatsuma
Kyushu Institute of Technology
Key Points
Enhanced triple extraction accuracy using a weighted loss function shows significant improvement in performance.
The improvement metrics include a X% increase in accuracy in comparison to baseline models.
Development of a structure dependent weighted loss function was utilized to optimize the BART model.
This advancement may enable more effective language understanding and could support broader applications in NLP.
Mark Helpful
Like
Save
Bookmark
Relay
Share
Mark Helpful
Like
Save
Bookmark
Relay
Share
Cite This Study
Copy
Shrivastava et al. (Thu,) studied this question.
synapsesocial.com/papers/69a75e56c6e9836116a28d15
https://doi.org/https://doi.org/10.1007/s10044-025-01604-9
Structure dependent weighted loss function for triple extraction using BART model | Synapse