Big data has come to the music business in a huge way. Nearly every aspect of the modern music industry relies on massive amounts of data, machine learning, and analytics to make better decisions faster. In this report, O’Reilly Strata + Hadoop World Chair Alistair Croll takes you on a tour of music science, a relatively new field staffed by data scientists, analytics experts, tastemakers, economists, and even game theorists.
Based on months of research and more than 70 interviews with scientists, founders, and artists, this report provides an overview of this multibillion-dollar venture. You’ll not only see where music science stands today—and where it’s headed—but get a tantalizing glimpse of what other industries will look like in coming years.
- Connected listening: the new supply chain from artist to listener
- The rise of metadata and the sheer volume of data associated with songs
- Recommendations and Pandora’s Music Genome Project
- Many ways to measure music consumption (and music consumers)
- Music science’s Turing problems, such as the problem of predicting hit songs
- How algorithms on listener preference need to learn quickly with fresh data
- What lies on the horizon for the music industry