Abstract:Empathetic response generation endeavours to perceive the interlocutor’s emotional and cognitive states in the dialogue and express proper responses. Previous studies detect the interlocutor’s states by understanding the immediate context of the dialogue. However, these methods are at an elementary/intermediate level of empathetic understanding due to the neglect of the broader context (i.e., the situation) and its associations with the dialogue, leading to inaccurate comprehension of the interlocutor’s states. In this paper, we utilize the EMPATHETIC-DIALOGUES dataset consisting of 25k dialogues, and on this basis, we propose a Situation-Dialogue Association Model (SDAM). SDAM focuses on the broader context, i.e., the situation, and enhances the understanding of empathy from explicit and implicit associations. Regarding explicit associations, we propose a bidirectional filtering encoder. It selects relevant keywords between the situation and dialogue, learning their direct lexical relevance. For implicit associations, we use a knowledge-based hypergraph network grounded to learn convoluted connections between the situation and the dialogue. Moreover, we also introduce a simple fine-tuning approach that combines SDAM with large language models to further strengthen the empathetic understanding capability. Compared to the baseline, SDAM demonstrates superior empathetic ability. In terms of emotion accuracy, fluency, and response diversity (Distinct-1/Distinct-2), SDAM achieves improvements of 12.25 (a 30.47% increase), 0.3 (a 0.85% increase), and 0.86/1.23 (116.22% and 30.67% increases), respectively. Additionally, our variant model based on large language models exhibits better emotion recognition capability without compromising response quality, specifically achieving an improvement of 0.23 (a 0.37% increase) in emotion accuracy.

Optimizing Empathetic Response by Generating and Integrating Emotion Feedback and Topic Discussion.

Improving Empathetic Response Generation by Emotion Recognition and Information Filtration

Generating Empathetic Responses by Injecting Anticipated Emotion

EmpHi: Generating Empathetic Responses with Human-like Intents

EmPO: Emotion Grounding for Empathetic Response Generation through Preference Optimization

Empathetic Response Generation with State Management

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

APTNESS: Incorporating Appraisal Theory and Emotion Support Strategies for Empathetic Response Generation

Improving Empathetic Dialogue Generation by Dynamically Infusing Commonsense Knowledge.

CEM: Commonsense-aware Empathetic Response Generation

Grafting Fine-tuning and Reinforcement Learning for Empathetic Emotion Elicitation in Dialog Generation

Empathetic Response Generation via Emotion Cause Transition Graph

Wish I Can Feel What You Feel: A Neural Approach for Empathetic Response Generation

Empathetic Dialogue Generation via Sensitive Emotion Recognition and Sensible Knowledge Selection

Exploiting Emotion-Semantic Correlations for Empathetic Response Generation

Situation-aware Empathetic Response Generation

Improving Empathetic Response Generation by Recognizing Emotion Cause in Conversations.

Enhancing Empathetic Response Generation by Augmenting LLMs with Small-scale Empathetic Models

Empathy Level Alignment via Reinforcement Learning for Empathetic Response Generation

Improving Empathetic Dialogue Generation with Semantics Decoupling.

EmoElicitor: an Open Domain Response Generation Model with User Emotional Reaction Awareness