FiSMiness: A Finite State Machine Based Paradigm for Emotional Support Conversations

2025-04-16

AI Summary

Emotion Support Conversation (ESC) systems suffer from low long-term user satisfaction, primarily due to the absence of explicit modeling of dialogue state evolution in existing LLM-based approaches. To address this, we propose the FSM-LLM协同 paradigm—introducing Finite State Machines (FSMs) into ESC for the first time. This framework enables a single large language model to autonomously perform end-to-end closed-loop processing, including state recognition, emotion inference, strategy selection, and response generation. Leveraging multi-stage state transitions and LLM self-guided reasoning, it eliminates reliance on external modules or static prompts while enabling dynamic strategy adaptation. Evaluated across multiple ESC benchmarks, our method significantly outperforms baselines—including direct inference, chain-of-thought prompting, self-refinement, fine-tuning, and external tool augmentation—achieving superior long-term support efficacy even with smaller model parameters. This work establishes a novel, state-driven paradigm for trustworthy, controllable ESC systems.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Conversational AI/Dialog Systems
• Multiagent Systems: Multiagent Learning

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information

Abstract

Emotional support conversation (ESC) aims to alleviate the emotional distress of individuals through effective conversations. Although large language models (LLMs) have obtained remarkable progress on ESC, most of these studies might not define the diagram from the state model perspective, therefore providing a suboptimal solution for long-term satisfaction. To address such an issue, we leverage the Finite State Machine (FSM) on LLMs, and propose a framework called FiSMiness. Our framework allows a single LLM to bootstrap the planning during ESC, and self-reason the seeker's emotion, support strategy and the final response upon each conversational turn. Substantial experiments on ESC datasets suggest that FiSMiness outperforms many baselines, including direct inference, self-refine, chain of thought, finetuning, and external-assisted methods, even those with many more parameters.

Problem

• Defining emotional support conversation from state model perspective
• Improving long-term satisfaction in emotional support conversations
• Enhancing single LLM performance in emotional support planning

Innovation

• FSM-based framework for emotional support conversations
• Single LLM bootstraps planning and self-reasoning
• Outperforms baselines with fewer parameters