Aug 2026
Constructing embeddings for textual causality via representation regularization and semantic metrics
Compared to end-to-end classification models, this method of constructing an embedding space based on cross-level semantic metric learning enhances the ability to learn textual causality features and significantly reduces the dependency on annotated data.
Jiajun Li, Bo Shen, Guangzhi Lang
· Knowledge and Information Sy... · 0 citations