The Future of Cognitive Strategy-enhanced Persuasive Dialogue Agents: New Perspectives and Trends

Mengqi Chen,Bin Guo,Hao Wang,Haoyu Li,Qian Zhao,Jingqi Liu,Yasan Ding,Yan Pan,Zhiwen Yu

2024-02-07

Abstract:Persuasion, as one of the crucial abilities in human communication, has garnered extensive attention from researchers within the field of intelligent dialogue systems. We humans tend to persuade others to change their viewpoints, attitudes or behaviors through conversations in various scenarios (e.g., persuasion for social good, arguing in online platforms). Developing dialogue agents that can persuade others to accept certain standpoints is essential to achieving truly intelligent and anthropomorphic dialogue system. Benefiting from the substantial progress of Large Language Models (LLMs), dialogue agents have acquired an exceptional capability in context understanding and response generation. However, as a typical and complicated cognitive psychological system, persuasive dialogue agents also require knowledge from the domain of cognitive psychology to attain a level of human-like persuasion. Consequently, the cognitive strategy-enhanced persuasive dialogue agent (defined as CogAgent), which incorporates cognitive strategies to achieve persuasive targets through conversation, has become a predominant research paradigm. To depict the research trends of CogAgent, in this paper, we first present several fundamental cognitive psychology theories and give the formalized definition of three typical cognitive strategies, including the persuasion strategy, the topic path planning strategy, and the argument structure prediction strategy. Then we propose a new system architecture by incorporating the formalized definition to lay the foundation of CogAgent. Representative works are detailed and investigated according to the combined cognitive strategy, followed by the summary of authoritative benchmarks and evaluation metrics. Finally, we summarize our insights on open issues and future directions of CogAgent for upcoming researchers.

Computer Science

What problem does this paper attempt to address?

The main problem this paper attempts to address is the development of dialogue agents capable of effectively persuading users (i.e., Cognitive Strategy Enhanced Persuasive Dialogue Agents, abbreviated as CogAgent). Specifically, the paper focuses on how to integrate strategies from cognitive psychology (such as persuasion strategies, topic path planning strategies, and argument structure prediction strategies) into dialogue agents to achieve a more natural, smooth, and efficient persuasion process.

### Main Problems
1. **Enhancing the Persuasive Ability of Dialogue Agents**: Although existing large language models (LLMs) perform well in understanding context and generating responses, they still fall short in persuading users to accept specific viewpoints or behaviors. The paper aims to enhance the persuasive ability of dialogue agents by introducing theories from cognitive psychology.
2. **Integrating Cognitive Strategies**: The persuasion process is a complex cognitive psychological activity that requires the combination of multiple cognitive strategies. The paper explores how to effectively integrate these strategies into the design of dialogue agents to improve their persuasive effectiveness.
3. **System Architecture and Methodology**: The paper proposes a systematic architecture and methodology for defining and implementing Cognitive Strategy Enhanced Persuasive Dialogue Agents (CogAgent) and summarizes existing datasets and evaluation metrics.
4. **Future Research Directions**: The paper also discusses some open issues and future research trends in the field of CogAgent, including model adaptability and generalizability, multi-party dialogue, and multimodal dialogue.

### Solutions
- **Definition and Formalization of Cognitive Strategies**: The paper defines three main cognitive strategies from cognitive psychology theories: persuasion strategies, topic path planning strategies, and argument structure prediction strategies.
- **System Architecture Design**: A general system architecture is proposed, including modules for dialogue understanding, cognitive strategy mining, cognitive strategy prediction, and dialogue generation.
- **Application Case Analysis**: Detailed introductions are provided on the applications of CogAgent in scenarios such as psychological counseling, bargaining, and social welfare advocacy.

### Summary
Through in-depth research on cognitive strategies and the design of system architecture, this paper provides a theoretical foundation and technical support for developing more effective persuasive dialogue agents. This not only helps to enhance the intelligence level of dialogue agents but also plays an important role in various practical application scenarios.