Refining Text Generation via Direct Preference Optimization for Long-Context Conversational Recommendation
Published in Journal of Natural Language Processing, 2026
Peer-reviewed journal article, published in Journal of Natural Language Processing Vol. 33, No. 2, pp. 957–988 (15 June 2026).
J-STAGE article page · DOI: 10.5715/jnlp.33.957
This work extends our CRS × DPO framework to handle long dialogue contexts, improving recommendation response generation when the dialogue history is too long for a short window.
Recommended citation: Manato Tajiri, Michimasa Inaba. (2026). "Refining Text Generation via Direct Preference Optimization for Long-Context Conversational Recommendation." Journal of Natural Language Processing, Vol. 33, No. 2, pp. 957–988.
Download Paper
