Papers with element-wise diversity-aware loss

1 papers
MATO: A Model-Agnostic Training Optimization for Aspect Sentiment Triplet Extraction (2025.naacl-long)

Copied to clipboard

Challenge: Existing models with strong in-house performance may struggle to generalize to diverse expressions.
Approach: They propose a model-agnostic t**raining method to improve ASTE model inference . they propose to compute the violation rate (VR) on each element of one triplet .
Outcome: The proposed method can improve aspect sentiment triplet extraction models consistent with expected results facing triplet element diversity.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations