"Listen to Music, Listen to Yourself": Design of a Conversational Agent to Support Self-Awareness While Listening to Music

Title of the Paper

“Listen to Music, Listen to Yourself”: Design of a Conversational Agent to Support Self-Awareness While Listening to Music

Bibliographic Information

• Research Domain: Interaction design of music's impact on self-awareness
• Keywords: Music listening, conversational agent, user engagement, self-determination theory, emotional well-being, human-computer interaction, meditation design, music psychology

Research Background and Problem Statement

• Identified Problems and Challenges:

  1. Music can evoke emotions and cognition, helping individuals understand their feelings and inner thoughts, thereby enhancing self-awareness.
  2. Although music-based technologies can support mental health, few studies focus on integrating conversational agents to enhance self-awareness during music listening.
  3. Existing systems lack guided designs for user self-awareness and expression, falling short in fostering emotional resonance and deep self-expression.

• Importance of the Study:

  1. Self-awareness is closely tied to mental health; strong self-awareness can help individuals better manage emotions and improve overall well-being.
  2. Conversational agents, as an emerging technology, can provide companion-like interactions and have the potential to enhance user engagement and experience.
  3. Exploring design strategies can offer valuable guidance for future music technologies and mental health support tools.

• Motivation and Related Work:

  • Music, as a common daily activity, plays a significant role in emotional regulation and social interactions.
  • The potential of conversational agents in mental health support has been partially validated, but their application in music listening scenarios remains underexplored.
  • Investigating the impact of two key design factors (proactive guidance and social information) can help optimize conversational system designs.

Solution

• Proposed Method or Solution:

  • Designed a novel conversational agent (CA) that supports users' self-awareness and self-expression through dialogue while listening to music.
  • The system integrates two key design factors: proactive guidance (PG) and social information (SI, music commentary).

• Innovative Contributions:

  1. For the first time, interactive conversational experiences are incorporated into music listening scenarios to support users' emotional resonance and self-expression.
  2. A five-day user study provides empirical data on how design factors influence self-awareness, need satisfaction, and user acceptance.

• Implementation Steps and Key Technologies:

  • Constructing the chat interface: The agent proactively asks users about their current mood, recommends music that matches their emotions, and encourages sharing inner feelings.
  • Integrating music commentary: Providing other listeners' comments on the same music to stimulate emotional resonance.
  • Designing experiments: Conducting a 2x2 experimental study to observe the effects of different PG and SI combinations.
  • Measurement metrics: Recording the frequency of emotional resonance, depth and length of self-expression, satisfaction of three basic psychological needs (autonomy, competence, and relatedness), and user acceptance (including perceived usefulness, ease of use, and willingness to use).

Research Findings

• Specific Results:

  1. The system successfully enhanced users' self-awareness: both proactive guidance and social information significantly influenced self-perception and expression capabilities.
  2. Social information (e.g., music commentary) promoted deeper expression, while high proactive guidance increased the frequency of emotional resonance.
  3. Proactive guidance yielded comprehensive effects, but excessive guidance might reduce users' sense of autonomy and perceived usefulness of the system.

• Comparative Advantages Over Existing Solutions:

  • Compared to traditional music recommendation and exploration systems, this study's conversational agent emphasizes cultivating users' self-awareness rather than merely recommending music.
  • By proposing two key design factors and conducting empirical analysis, the study clarifies how to optimize systems to better meet mental health needs.

• Experimental or Evaluation Results:

  1. Task Level: High proactive guidance significantly enhanced self-awareness, while social information significantly increased expression depth.
  2. Technical Level: User acceptance was mediated by the degree of psychological need satisfaction, with high perceived autonomy positively correlated with lower proactive guidance settings.
  3. Life Level: The system slightly improved users' overall mental health scores during the five-day user study.
  4. Music Matching: Music ratings were an important contextual factor influencing emotional resonance and psychological need satisfaction.

• Limitations and Future Directions:

  • Participants were primarily young individuals from China, and cultural and age differences may affect the generalizability of results.
  • Music recommendations were not fully personalized; future improvements could leverage advanced personalized recommendation algorithms (e.g., based on historical data) to optimize user experience.
  • The conversational agent's dialogue design was relatively basic; future enhancements could employ emotionally intelligent language models to improve empathy and interactive feedback.