Skip to main content Skip to navigation

Narrative-Centered Emotional Reflection: Scaffolding Autonomous Emotional Literacy with AI

Project Overview

The document discusses Reflexion, an innovative AI-powered platform aimed at enhancing emotional self-reflection in educational settings. By incorporating real-time emotion detection, structured reflective prompts, and metaphorical storytelling, Reflexion encourages users to engage deeply with their emotions. Initial studies reveal that this tool significantly improves emotional articulation, cognitive reframing, and psychological resilience among users. As a result, Reflexion holds substantial promise not only in educational environments but also in therapeutic and public health contexts, showcasing the potential of generative AI to foster emotional intelligence and resilience. The findings suggest that leveraging AI in education can create supportive frameworks for students to navigate their emotional landscapes, ultimately enhancing their overall learning experience and well-being.

Key Applications

Reflexion: An AI-powered platform for emotional self-reflection

Context: Educational, therapeutic, and public health contexts for diverse users, including students, educators, and mental health advocates.

Implementation: The platform uses real-time emotion detection, layered reflective prompts, and metaphorical storytelling generated on user input.

Outcomes: Improvements in emotional articulation, cognitive reframing, and perceived psychological resilience among users.

Challenges: Cultural specificity, technical constraints, and the need for empirical validation of long-term effectiveness.

Implementation Barriers

Technical Barrier

The current implementation relies on external APIs and pre-trained models for emotion detection, which can lead to inaccuracies. Future plans include enhancing the backend for real-time, context-aware emotion detection.

Cultural Barrier

Cultural differences in emotional expression and interpretation may affect user engagement and effectiveness. Future work must integrate cross-cultural narrative frameworks and participatory co-design methods involving diverse user groups.

Ethical Barrier

Concerns around emotional nudging or manipulation, as well as privacy and consent regarding emotional data. Solutions include embedding transparent data handling practices and ensuring user-controlled emotional histories.

Project Team

Shou-Tzu Han

Researcher

Contact Information

For information about the paper, please contact the authors.

Authors: Shou-Tzu Han

Source Publication: View Original PaperLink opens in a new window

Project Contact: Dr. Jianhua Yang

LLM Model Version: gpt-4o-mini-2024-07-18

Analysis Provider: Openai

Let us know you agree to cookies