Summary
Understanding how people express feelings and manners in conversations is tricky because it involves more than just words. The authors created a new computer method that pays attention to who is talking and who they are talking to, capturing sentiment, emotion, politeness, and formality better than before. Their method uses roles in conversations to guide how the computer understands text, improving results especially for complex emotions and polite language. However, it doesn’t improve much for general sentiment, which seems to depend more on the words used. Their approach also helps explain why the computer made certain predictions, making it easier to understand.
What this means in practice
- •For customer support teams: Improve automated detection of emotions and politeness in customer interactions to enhance response quality and customer satisfaction.
- •For social media moderators: Enhance the identification of polite and formal language to better moderate online conversations and detect nuanced emotional cues.
Authors
Sangpil Youm, Annabel Stocks Natalias, Hayden Heninger, Leah Jones, Xiaoyang Jin, Xiaoya Zhang, Bonnie Dorr
Abstract
Detecting communicative meaning in dialogue requires modeling not only lexical content but also the participant-role structures through which sentiment, emotion, politeness, and formality (SEPF) are expressed. Existing approaches rely on contextual language representations while leaving semantic roles implicit. We introduce a role-aware transformer architecture that maps the four SEPF dimensions to theory-grounded Semantic Role Labeling (SRL) configurations, incorporating these configurations directly into the transformer's self-attention mechanism. Role-aware attention models non-directional internal states (sentiment and emotion), while directional role-aware attention captures initiator-to-recipient relations for external cues (politeness and formality). The resulting role-aware representations are fused with contextual embeddings through a learned gate that adaptively weights how much role information each input contributes, rather than appending semantic roles as an auxiliary feature vector. Across four benchmark datasets, the framework yields statistically reliable gains over strong BERT-based baselines for emotion, politeness, and formality, with gains concentrated in theoretically predicted cases such as low-frequency, complex emotions and highly polite expressions; sentiment improvements are marginal and non-significant, suggesting that sentiment is conveyed more through lexical choice than participant-role structure. Additional evaluation on real-world data demonstrates that the framework improves both predictive performance and model traceability. These results demonstrate that linguistically grounded modifications to transformer self-attention improve the detection of communicative meaning, generalize beyond benchmark datasets to real-world mental-health data, and yield traceable, role-grounded explanations that extend beyond predictive performance to interpretability.