Emovectors: assessing emotional content in jazz improvisations for creativity evaluation

2025-12-09

AI Summary

This study addresses the automated assessment of creativity in jazz improvisation, investigating whether emotional expressivity serves as a valid proxy metric. We propose the “emovector” framework: it extracts audio embeddings from a pretrained neural model and maps them onto an interpretable, comparable emotional vector space grounded in a psychology-informed music emotion taxonomy; emotional richness is then quantified via cross-fragment statistical aggregation. Empirical evaluation demonstrates that the emovector metric exhibits significant correlation with human expert judgments of improvisational creativity and robustly discriminates between high- and low-creativity passages. To our knowledge, this work constitutes the first psychology-anchored, computationally tractable representation of musical emotion—establishing a novel, scalable, and interpretable paradigm for evaluating creativity in generative music systems, particularly within large language model–driven audio generation contexts.

Technology Category

• Cognitive Modeling & Cognitive Systems: Computational Creativity
• Humans and AI: Emotional Intelligence
• Natural Language Processing: Generation

Application Category

• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• User Modeling, Personalization and Recommendation: Metrics for user behavior and evaluating success
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

Music improvisation is fascinating to study, being essentially a live demonstration of a creative process. In jazz, musicians often improvise across predefined chord progressions (leadsheets). How do we assess the creativity of jazz improvisations? And can we capture this in automated metrics for creativity for current LLM-based generative systems? Demonstration of emotional involvement is closely linked with creativity in improvisation. Analysing musical audio, can we detect emotional involvement? This study hypothesises that if an improvisation contains more evidence of emotion-laden content, it is more likely to be recognised as creative. An embeddings-based method is proposed for capturing the emotional content in musical improvisations, using a psychologically-grounded classification of musical characteristics associated with emotions. Resulting'emovectors'are analysed to test the above hypothesis, comparing across multiple improvisations. Capturing emotional content in this quantifiable way can contribute towards new metrics for creativity evaluation that can be applied at scale.

Problem

• Assessing creativity in jazz improvisations via emotional content
• Developing automated metrics for creativity in generative AI systems
• Quantifying emotional involvement in music for scalable evaluation

Innovation

• Emotional content analysis using embeddings-based method
• Psychologically-grounded classification of musical characteristics
• Quantifiable emovectors for scalable creativity evaluation metrics