About

I am a Data Scientist at Spotify, working on recommendations within the Personalization team. I conduct research and develop metrics to understand consumption of personalized content, and I set up experiments to evaluate the impact of new features on user satisfaction with Spotify recommendations.

I also hold a PhD from the Music Cognition Lab, part of the Cognitive Science Research Group at Queen Mary University of London. I researched the effects of expectation¹ on chills² experienced in response to music listening with Marcus Pearce and Emmanouil Benetos, using behavioral, physiological, and computational methods. I am interested in music cognition, empirical aesthetics, data science, data vizualisation, and open science.

You can reach me by email, and find me on Twitter, LinkedIn, Github, OSF, and Google Scholar.

Education

• PhD in Cognitive Science at QMUL, Department of Computer Science
• MSc in Music, Mind, and Brain at Goldsmiths, Department of Psychology
• BSc in Osteopathic Studies at British School of Osteopathy

Publications

1. Modelling chills in music [pdf, code]
2. Chills in music: A systematic review [doi, pdf, data]
3. Valence and chills: A corpus analysis [doi, pdf, suppl, code]
4. Reward prediction tells us less than expected about musical pleasure [doi, pdf, code]
5. Information-theoretic measures predict rhythm complexity [doi, pdf, suppl, data]

Conferences

1. SMPC at New York University [pdf, url]
2. SEMPRE at University of Cambridge [pdf, url]
3. DMRN at Queen Mary University of London [pdf, url]
4. EiTIC at Pompeu Fabra University [pdf, url]
5. ESCOM at Royal Northern College of Music [pdf, url]
6. ICTTP at Ionian University [pdf, url]

Press

My work on chills has been covered in Quartz, Big Think, Inc, and IFLScience.

Teaching

Guest lectures for Research Methods and Music Perception and Cognition postgraduate modules at QMUL:

• Music and emotion [url]
• Linear modeling using R [url]
• Inferential statistics using R [url]
• Statistical programming using R [url]
• Typesetting with LaTeX [url]

Workshops for Music Cognition Lab seminars and Media & Arts Technology peer teaching sessions:

• Remote computing at QMUL [url]
• Introduction to ggplot2 [url]
• Introduction to the tidyverse [url]
• Tools for online data collection [url]
• Introduction to statistics [url]

Data journalism

Projects conducted as a Google Fellow for the Guardian Visuals and Data teams:

1. 2019 Queen’s Speech [url]
2. Amazon fires [url]
3. Driving test pass rates [url]
4. GCSE results day [url]
5. A-level results day [url]

---

1. Musical expectation follows a process of probabilistic prediction of the statistical regularities in musical structure, which allows expectations learned through musical enculturation to be violated, delayed, or confirmed when listening to music, resulting in induced emotions [ref] ↩︎

2. Chills are a fleeting, pleasurable bodily sensation, sometimes accompanied by goosebumps, experienced when listening to specific musical passages [ref] ↩︎