Jiro Nagao
Author directory2026
Surfacing Minor Nuances in Subjective Expressions: Initial Evidence for Improved Output Quality in Collaborative Caption Revision
Takeru Isaka | Jiro Nagao | Iwaki Toshima
Proceedings of the 15th International Joint Conference on Natural Language Processing and the 5th Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)
Takeru Isaka | Jiro Nagao | Iwaki Toshima
Proceedings of the 15th International Joint Conference on Natural Language Processing and the 5th Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)
Creative collaboration benefits when collaborators can absorb the nuances embedded in others’ subjective expressions. We focus on especially difficult-to-notice nuances, which we call **minor nuances**, and provide initial empirical evidence that supporting their communication can improve creative collaboration outcomes. We then propose a commonsense-generation-based method that identifies minor nuances in subjective expressions and presents them as cues for collaborators. We evaluated the proposed method in a collaborative image-caption revision experiment designed to reflect brainstorming. Compared with a baseline in which collaborators freely explained the nuances embedded in their own subjective expressions, our method significantly improved output-quality ratings from human evaluators. These results provide initial evidence suggesting that explicitly sharing minor nuances supports richer creative expression by helping convey fine-grained aspects of senders’ inner states that are difficult to communicate through ordinary explanations.