Sam Miller

A doctoral student at Warwick Business School, Sam Miller's research applies machine learning and imaginative insights to data, like combining the dark web and Wikipedia

Fact file

  • PhD student at Warwick Business School
  • Spent three years as an Economist at the Bank of England analysing financial regulation
  • Studied Economics at the University of Cambridge, where he won the Malthus Prize for high academic achievement

How would you describe your work?

A surge in demand for an illegal drug represents a looming public health threat – think of the US opioid epidemic – but how do you spot such a surge? Current statistics on drug use usually come from annual surveys, which aren’t fit for purpose because they aren’t frequent enough to capture sudden shifts in demand, and people aren’t always truthful in surveys on their substance use.

In our research, we scrape data from the drug markets on the dark web to create a dataset of global drug sales over time. We then use daily Wikipedia page views for each drug (as a proxy for consumer interest in it) to model how demand is changing. Public health authorities could use our model to monitor drug use more accurately, which would allow faster responses to crises like the opioid epidemic.

What’s the most surprising thing to come out of your research?

The serendipitous origin story of the dark web paper. An Oxford academic called Martin Dittus, who I’d never met before, overhead me in the Turing kitchen talking about the drug markets on the dark web. He approached me for a chat. Nine months later and we had finished a paper together!      

What aspect of your work is most exciting you right now?

The massive opportunity for real-world uses of higher-frequency statistics. For example, the first paper I wrote at Turing involved using real-time aircraft radar data to track flights and provide early indication of changes to the aviation sector’s economic performance. This may sound dull to some, but the UK’s Office for National Statistics recently used our work to track the early spread of coronavirus to vulnerable countries. There’s lots of opportunities for applying data science across disciplines.

Where do you see an untapped opportunity for the Turing?

The Turing’s in a unique place to encourage co-operation between academic disciplines. For example, the dark web research was a collaboration between an economist, three computer scientists and a social anthropologist. I think the mix of skills allowed us to produce some great research very quickly, but if Martin hadn’t overheard us in the kitchen that day then it’d probably never have happened. The Turing could do even more to foster such collaborations.

What three words would you use to describe your work, currently?

Big. Fast. Messy

What book should everyone be aware of?

The Gods Themselves by Isaac Asimov. It really changed how I thought about tackling climate change, despite being written before climate science!

And finally… What would have been the most interesting question I could have asked you for the purposes of this Q&A, and what is the answer to that question?

Q: Who provided the domain expertise for the dark web drug markets paper? A: Not me.