Abstract:Objective: Poor lifestyle represents a health risk factor and is the leading cause of morbidity and chronic conditions. The impact of poor lifestyle can be significantly altered by individual's behavioral modification. Although there are abundant lifestyle promotion applications and tools, they are still limited in providing tailored social support that goes beyond their predefined functionalities. In addition, virtual coaching approaches are still unable to handle user emotional needs. Our approach presents a human-virtual agent mediated system that leverages the conversational agent to handle menial caregiver's works by engaging users (e.g., patients) in a conversation with the conversational agent. The dialog used a natural conversation to interact with users, delivered by the conversational agent and handled with a finite state machine automaton. Our research differs from existing approaches that replace a human coach with a fully automated assistant on user support. The methodology allows users to interact with the technology and access health-related interventions. To assist physicians, the conversational agent gives weighting to user's adherence, based on prior defined conditions. Materials and methods: This article describes the design and validation of CoachAI, a conversational agent-assisted health coaching system to support health intervention delivery to individuals or groups. CoachAI instantiates a text-based health care conversational agent system that bridges the remote human coach and the users. Results: We will discuss our approach and highlight the outcome of a 1-month validation study on physical activity, healthy diet, and stress coping. The study validates technology aspects of our human-virtual agent mediated health coaching system. We present the intervention settings and findings from the study. In addition, we present some user-experience validation results gathered during or after the experimentation. Conclusions: The study provided a set of dimensions when building a human-conversational agent powered health intervention tool. The results provided interesting insights when using human-conversational agent mediated approach in health coaching systems. The findings revealed that users who were highly engaged were also more adherent to conversational-agent activities. This research made key contributions to the literature on techniques in providing social, yet tailored health coaching support: (1) identifying habitual patterns to understand user preferences; (2) the role of a conversational agent in delivering health promoting microactivities; (3) building the technology while adhering to individuals' daily messaging routine; and (4) a socio-technical system that fits with the role of conversational agent as an assistive component. Future work: Future improvements will consider building the activity recommender based on users' interaction data and integrating users' dietary pattern and emotional wellbeing into the initial user clustering by leveraging information and communication technology approaches (e.g., machine learning). We will integrate a sentiment analysis capability to gather further data about individuals and report these data to the caregiver.

Developing an AI-based Coaching Chatbot: A Study on Disclosure as Effectiveness Factor in Human-Machine-Coaching

“Don’t Look at Me!”: the Role of Avatars’ Presentation Style and Gaze Direction in Social Chatbot Design

Consciousness: effect of coaching process and specifics through AI usage

Tell me, what are you most afraid of? Exploring the Effects of Agent Representation on Information Disclosure in Human-Chatbot Interaction

Users' Perception on Appropriateness of Robotic Coaching Assistant's Disclosure Behaviors

Digital Confessions: The Willingness to Disclose Intimate Information to a Chatbot and its Impact on Emotional Well-Being

Dialoging Resonance: How Users Perceive, Reciprocate and React to Chatbot's Self-Disclosure in Conversational Recommendations

Inducing Consumers’ Self-Disclosure Through the Fit Between Chatbot’s Interaction Styles and Regulatory Focus

An Empathetic AI Coach for Self-Attachment Therapy

Coaching Copilot: Blended Form of an LLM-Powered Chatbot and a Human Coach to Effectively Support Self-Reflection for Leadership Growth

Assistive Conversational Agent for Health Coaching: A Validation Study

Smooth Talking: Generative Versus Scripted Coaching Chatbot Adoption and Efficacy Comparison

CoachAI: A Conversational Agent Assisted Health Coaching Platform

AI Health Counselor's Self-Disclosure: Its Impact on User Responses Based on Individual and Cultural Variation

AI Chatbots as Multi-Role Pedagogical Agents: Transforming Engagement in CS Education

Effect of AI chatbot emotional disclosure on user satisfaction and reuse intention for mental health counseling: a serial mediation model

Designing Online Virtual Advisors to Encourage Customer Self-disclosure: A Theoretical Model and an Empirical Test

Identity Disclosure and Anthropomorphism in Voice Chatbot Design: A Field Experiment

The HCI Aspects of Public Deployment of Research Chatbots: A User Study, Design Recommendations, and Open Challenges

Artificial Intelligence (AI) Student Assistants in the Classroom: Designing Chatbots to Support Student Success

Evaluation of In-Person Counseling Strategies To Develop Physical Activity Chatbot for Women