Search papers, labs, and topics across Lattice.
This paper introduces a novel dialogue summarization framework that integrates both semantic and emotional dynamics by employing a hierarchical Chain-of-Agents approach. By decomposing dialogues into topic segments and participant-specific utterances, the method generates summaries that reflect not only the content but also the emotional trajectories of the conversation. Evaluation using new emotion trajectory metrics demonstrates that the framework effectively preserves emotional flow while maintaining semantic accuracy, outperforming traditional summarization methods.
Summaries generated from dialogues can now capture emotional dynamics alongside semantic content, enhancing our understanding of conversational nuances.
Existing text summarization research has focused much on monologic information (e.g., newspaper articles, reports) without accounting for the interaction between speakers or authors. In contrast, dialogues are a rich communication channel where multiple participants conduct back and forth exchanges to construct meaning. We propose a dialogue summarization framework that explicitly models both semantic and emotion dynamics using multimodal dialogue inputs, built on an adapted hierarchical Chain-of-Agents approach. We decompose dialogues from two perspectives: (1) topic segments based on the utterances of all participants, and (2) participant-specific utterance segments. These are used to generate corresponding summaries while incorporating automatically inferred emotions. Topic- and participant-level summaries are aggregated into a dialogue summary capturing semantic content and emotion trajectories. To evaluate beyond content accuracy, we introduce emotion trajectory metrics measuring how well summaries preserve emotional flow. Experiments with small language models on multimodal dialogue datasets show that our framework produces summaries with both semantic and emotion content. Further experiments on explicit emotion label availability highlight the efficacy of our proposed methodology and the opportunities in dialogue analysis using language models.
