NoRe: Augmenting Journaling Experience with Generative AI for Music Creation

2025-06-02

AI Summary

This study addresses the lack of affective embodiment support in existing digital journaling tools by investigating how generative AI can enhance emotional engagement and self-reflection through personalized music generation. We propose the first closed-loop “text–music” co-generation system, integrating fine-grained textual emotion analysis, intent recognition, and individual preference modeling to guide LLM-conditioned diffusion-based audio synthesis. The system dynamically generates context-aware background music aligned with diary content and real-time affective states, iteratively refined via user feedback. In a 7-day in-the-wild study with 15 participants, 87% reported significantly improved emotional resonance and memory vividness; the system was used an average of 3.2 times per day. Our core contribution is a reflective, interpretable, and adaptive generative music interaction paradigm grounded in affective computing and human-centered AI design.

Technology Category

• Natural Language Processing: Generation
• Cognitive Modeling & Cognitive Systems: Affective Computing
• Humans and AI: Game Design — Procedural Content Generation & Storytelling

Application Category

• Social Networks and Social Media: Generative AI / large language models and their impact on social systems
• User Modeling, Personalization and Recommendation: Studies of user behavior, including longitudinal effects of personalized systems
• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications

Abstract

Journaling has long been recognized for fostering emotional awareness and self-reflection, and recent advancements in generative AI offer new opportunities to create personalized music that can enhance these practices. In this study, we explore how AI-generated music can augment the journaling experience. Through a formative study, we examined journal writers' writing patterns, purposes, emotional regulation strategies, and the design requirements for the system that augments journaling experience by journal-based AI-generated music. Based on these insights, we developed NoRe, a system that transforms journal entries into personalized music using generative AI. In a seven-day in-the-wild study (N=15), we investigated user engagement and perceived emotional effectiveness through system logs, surveys, and interviews. Our findings suggest that journal-based music generation could support emotional reflection and provide vivid reminiscence of daily experiences. Drawing from these findings, we discuss design implications for tailoring music to journal writers' emotional states and preferences.

Problem

• Enhancing journaling with AI-generated personalized music
• Exploring emotional reflection through journal-based music creation
• Designing AI systems for tailored emotional music experiences

Innovation

• Generative AI transforms journal entries into music
• System tailors music to emotional states
• AI-generated music enhances emotional reflection