Abstract:Empathy, the ability to understand and respond to others' emotions, is critical in dialogue systems, particularly for applications such as psychological counselling and casual conversation. However, existing approaches often struggle to accurately capture emotional transitions in user utterances and dialogue contexts across multiple levels of granularity. Additionally, the orthogonality of one-hot emotion representations overlooks the subtle relationships between emotion categories. To overcome these limitations, we propose a novel framework: the commonsense inductive relation graph with fine emotional soft labels for empathetic response generation. This framework refines the representation of user utterances by incorporating commonsense knowledge to derive external emotional cues, which are integrated into an adaptive inductive relation graph. This graph explicitly captures and models dynamic emotional trajectories throughout the dialogue. Additionally, we propose a fine-grained soft label construction method that more accurately reflects the nuanced distribution of emotions, replacing the rigid one-hot encoding with a continuous representation that better captures inter-emotion relationships. Our approach achieves strong performance on the benchmark EmpatheticDialogues dataset, with a perplexity score of 27.85, Dist-1 of 1.81, Dist-2 of 6.58, and a dialogue emotion accuracy of 55.65%. These results highlight the effectiveness of our framework in capturing subtle emotions and generating contextually rich, empathetic responses.

Empathetic Dialog Generation with Fine-Grained Intents

A Taxonomy of Empathetic Response Intents in Human Social Conversations

EmpHi: Generating Empathetic Responses with Human-like Intents

Use of a Taxonomy of Empathetic Response Intents to Control and Interpret Empathy in Neural Chatbots

CEM: Commonsense-aware Empathetic Response Generation

Improving Empathetic Response Generation by Emotion Recognition and Information Filtration

EmpCRL: Controllable Empathetic Response Generation via In-Context Commonsense Reasoning and Reinforcement Learning

Generating Empathetic Responses by Injecting Anticipated Emotion

Improving Empathetic Dialogue Generation by Dynamically Infusing Commonsense Knowledge.

Towards Multi-Turn Empathetic Dialogs with Positive Emotion Elicitation

Fine-grained Emotion and Intent Learning in Movie Dialogues

Empathetic Dialogue Generation via Sensitive Emotion Recognition and Sensible Knowledge Selection

Empathetic Response Generation via Emotion Cause Transition Graph

Empathetic Response Generation with State Management

Acknowledgment of Emotional States: Generating Validating Responses for Empathetic Dialogue

Multi-dimensional Evaluation of Empathetic Dialog Responses

Improving Empathetic Response Generation by Recognizing Emotion Cause in Conversations.

CIRG-SL: Commonsense Inductive Relation Graph framework with Soft Labels for Empathetic Response Generation

CAB: Empathetic Dialogue Generation with Cognition, Affection and Behavior

Constructing Emotion Consensus and Utilizing Unpaired Data for Empathetic Dialogue Generation

Perspective-taking and Pragmatics for Generating Empathetic Responses Focused on Emotion Causes