Applications of data science for responsible gambling: a scoping review

Technological innovations in the gambling industry have revolutionized the availability, storage, and use-cases of data. How this data may be leveraged for responsible gambling has emerged as a popular field of inquiry. We conducted a scoping review following PRISMA guidelines to understand the current state of data science applications for responsible gambling by exploring the aims, study designs, and methods used by researchers. Thirty-seven studies were included in the final review that spanned three categories: (1) cluster analysis (n = 14), (2) supervised machine learning with behavioral tracking data (n = 17), and (3) other data science applications (n = 6). Over half of the studies were published between 2018 and 2021. Existing research focuses on the development of responsible gambling tools centered around customer profiling and risk-detection. Our analysis of the records revealed limitations in terms of generalizability and reproducibility, as well as a considerable lack of peer-reviewed work. The current evidence suggests that the utility and adoption of data science in practice remains largely unexplored. Future work may focus on additional data science techniques with novel datasets and in situ research.