Explainable AI meets persuasiveness: Translating reasoning results into behavioral change advice

Area of Science:

• Artificial Intelligence
• Health Informatics
• Behavioral Science

Background:

• Explainable AI (XAI) systems aim to provide understandable justifications for decisions.
• In chronic disease management, patients require adherence to strict dietary rules.
• Effective communication of AI-driven insights is crucial for patient behavior modification.

Purpose of the Study:

• To address challenges in Natural Language Generation for explaining AI reasoning inconsistencies.
• To evaluate the effectiveness of persuasive messages in influencing patient dietary behavior.
• To develop AI systems that support chronic disease management through clear, actionable feedback.

Main Methods:

• Developing Natural Language Generation techniques for explaining AI-driven dietary recommendations.
• Designing persuasive communication strategies for patient feedback.
• Evaluating the impact of persuasive explanations on reducing unhealthy eating behaviors in chronic disease patients.

Main Results:

• Persuasive explanations significantly reduced unhealthy behaviors among users.
• The study demonstrated the feasibility of generating clear and understandable justifications for AI decisions.
• Effective communication of AI reasoning is key to improving patient adherence to dietary guidelines.

Conclusions:

• Explainable AI with persuasive communication can effectively guide patients towards healthier dietary choices.
• This approach holds promise for improving chronic disease management and patient outcomes.
• Further research into NLG and persuasive techniques can enhance patient engagement with AI health systems.