Democratizing Machine Learning for Interdisciplinary Scholars: Reflections on the NLP+CSS Tutorial Series

Katherine A. Keith, Ian Stewart

Abstract: Many scientific fields—including biology, health, education, and the social sciences—use machine learning (ML) to analyze data at an unprecedented scale. However, ML researchers who develop advanced methods rarely provide tutorials showing how to apply these methods. We attempt to democratize the use of ML methods by making them accessible to a broader set of reserachers and practitioners. To that end, we organized a year-long, free, online tutorial series targeted at teaching advanced natural language processing (NLP) methods to computational social science (CSS) scholars. Two organizers worked with fifteen subject matter experts to develop tutorials with hands-on Python code for a range of methods and use cases, from data pre-processing to analyzing temporal language changes. Although live participation was more limited than expected, surveys of participants showed an increase in their perceived knowledge by almost one point on a 7-point Likert scale. Furthermore, participants asked thoughtful questions during tutorials and engaged readily with the content afterwards, as demonstrated by approximately 30K total views of posted tutorial recordings. We distill five principles for democratizing other ML+X tutorials, and we hope that future organizers continue to lower barriers to developing ML skills for researchers of all fields.