Paulina Pankowska receives grant for research on algorithm benchmarking in the social sciences

For the project ‘Social Science Benchmark - Validating Predictive Accuracy for Social Systems’, Paulina Pankowska received a grant to work on the topic of algorithm benchmarking in the social sciences, together with Adriënne Mendrik (Eyra) and Daniel Oberski (Utrecht University). The grant was awarded by ODISSEI - Open Data Infrastructure for Social Science and Economic Innovations.

04/19/2021 | 11:05 AM

PaulinaBenchmarking is about creating a system of reference, in which various models that address the same task can be compared, in order to be able to analyse the strengths and weaknesses of various approaches and gain new insights. For computer scientists, benchmarks provide a standardized way to validate predictive models, which has proven to be able to detect major breakthroughs, like deep learning. However, social scientists mostly focus on causality and explanation, and so whether social systems can be predicted is to them debatable. It could be argued though that causality and explanation, as well as prediction are different sides of the same coin. As such, they can supplement and strengthen each other and provide a fuller, more insightful picture of social phenomena.

As a part of this project they will design and set-up a benchmark for social sciences that addresses a real world problem that social scientists are working on and preferably draws from various data sources.