Search papers, labs, and topics across Lattice.
This paper introduces PragAlign, a reply assistance system that differentiates between context reading and selective clarification to improve appropriateness judgments in multilingual settings. Evaluated by native speakers of Chinese and Japanese, PragAlign outperformed both Direct and Rule methods in the Chinese context, while achieving the highest top-rank rate in Japanese evaluations. The findings reveal both shared and language-specific patterns in appropriateness judgments, providing valuable insights for enhancing linguistic and cultural understanding in AI-assisted communication.
PragAlign significantly outperforms traditional methods in multilingual reply assistance, revealing critical insights into cultural and linguistic appropriateness.
Reply assistance in multilingual settings requires linguistic competence and culturally situated judgments of appropriateness. We present PragAlign, which separates context reading from selective clarification, and evaluate it alongside Direct and Rule. Nine native Chinese speakers judged Chinese materials, while three native Japanese speakers judged matched Japanese versions. In the Chinese evaluation, PragAlign received significantly better ranks than both baselines. In the Japanese evaluation, Direct had the lowest mean rank, PragAlign had the highest top-rank rate, and the omnibus difference was not significant. The groups selected the same top condition in 5 of 10 scenarios, including four shared PragAlign selections. The results identify shared and language-specific judgment patterns and inform reply assistance designed to support linguistic and cultural understanding.
