Human Computer Interaction
Music, Affect, Attention and Personality: A Responsible Multimodal AI Study
Investigate how music-related audio features, self-reported affect, attention measures and personality traits interact, using transparent multimodal AI and non-clinical, consent-based research methods.
- 0
- Progress
- 2
- Researchers
About this project
This proposed study will explore relationships among music, affect, attention, memory and personality through a responsible multimodal AI framework. It will begin with a preregistered scoping and feasibility phase, then design consent-based data collection or use appropriately licensed public datasets combining audio features, listening context and validated non-clinical self-report measures. The team will compare interpretable statistical baselines with multimodal and graph-learning approaches, examine cultural variation and bias, and test whether explanations are meaningful to participants and researchers. This is a new research direction, not a claim of completed music, clinical or neuroscience research, and it will not diagnose mental-health or neurological conditions.
Objectives
- Develop a preregistered ethical framework for multimodal music and cognition research
- Map evidence linking musical features, affect, attention, memory and personality
- Design a consent-based pilot study or licensed-dataset benchmark
- Compare interpretable statistical, multimodal and graph-learning models
- Evaluate cultural variation, fairness, robustness and explanation usefulness
- Publish methods, reusable code and clearly bounded findings
Research priorities
- Music information and affective computing
- Attention, memory and personality
- Multimodal and graph-based learning
- Human-centred explainability
- Cross-cultural validity and fairness
- Consent, privacy and non-clinical research boundaries
Sustainable Development Goals
- SDG 3 · Good health and well-being
- SDG 4 · Quality education
- SDG 9 · Industry, innovation and infrastructure
- SDG 10 · Reduced inequalities
Activities
- Preregistered scoping review and research protocol
- Ethics, consent and measurement design
- Dataset licensing review or pilot data collection
- Audio, contextual and self-report feature engineering
- Baseline, multimodal and graph-model experiments
- Bias, robustness and explanation evaluation
- Open materials and manuscript preparation
How to take part
Researchers can contribute expertise in music information retrieval, psychology, cognitive science, HCI, multimodal machine learning, graph learning, statistics, ethics, qualitative evaluation or scholarly review. Clinical interpretation is outside scope unless separately led and approved by qualified professionals.
Skills required
- Python
- audio analysis
- multimodal learning
- graph learning
- experimental design
- statistics
- HCI
- research ethics
- scholarly writing
Roles needed
- Multimodal ML researcher
- Music information researcher
- Cognitive science researcher
- HCI researcher
- Statistician
- Responsible AI researcher
Who should apply
Open to postgraduate students and researchers with relevant computing, music, psychology, cognitive science, statistics, HCI or ethics expertise. Applicants must respect the non-clinical scope and consent and privacy requirements.
Expected contribution
Collaborators should follow the preregistered protocol, use ethically sourced data, avoid diagnostic or therapeutic claims, document methods and contributions, participate in regular meetings, and maintain reproducible version-controlled work.
