Search papers, labs, and topics across Lattice.
This paper introduces AtmosERC, a novel graph-based framework for Emotion Recognition in Conversation (ERC) that emphasizes the dialogue-level affective atmosphere, a latent signal crucial for predicting emotions in dialogues. By utilizing a relation-aware graph extractor, AtmosERC effectively filters and integrates diverse contextual information, resulting in improved emotion prediction while maintaining lightweight processing. Experimental results across four ERC benchmarks demonstrate that AtmosERC not only enhances the performance of lightweight ERC but also serves as an effective plug-in for LLM-based systems, yielding more stable predictions amidst emotional fluctuations.
Emotion recognition can be dramatically improved by focusing on the latent dialogue-level affective atmosphere, leading to more accurate predictions even in emotionally unstable contexts.
Emotion Recognition in Conversation (ERC) aims to predict utterance-level emotions in dialogues and has largely advanced through context-centric modeling. However, global context is a heterogeneous signal, and not all contextual information is equally relevant to emotion prediction. This paper focuses on the affect-oriented component of this signal, termed dialogue-level affective atmosphere, which captures a latent tendency commonly reflected in conversational emotion patterns. To estimate and exploit this tendency, we propose AtmosERC, a graph-based ERC framework that models each dialogue as a conversational graph over utterances and speakers. A relation-aware graph extractor filters and fuses heterogeneous graph signals to produce dialogue-level and speaker-conditioned affective priors. The resulting compact prior guides lightweight sequential emotion prediction and can also be verbalized into prompt-level cues for LLM-based ERC without modifying backbone models. Experiments on four ERC benchmarks show that AtmosERC improves lightweight ERC, enhances LLM-based ERC as a plug-in cue, and yields more stable predictions under local emotional deviations.
