Human-Centered Data Science

Human-Centered
Data Science

This is a course on human-centered data science, with lectures covering multiple aspects like design methods in data science, including ethical concerns and privacy requirements. You will learn key concepts such as bias, fairness, accountability, transparency, and explainability.

The course was created by the Human-Centered Computing research group of the FU Berlin.

Why Human-Centered?

This approach prioritizes the needs, contexts, and values of people in the design and implementation of data science solutions. It emphasizes ethical considerations, user engagement, and the socio-technical implications of data-driven technologies.

Key Concepts

This course provides a comprehensive overview of human-centered data science, covering key concepts such as bias, fairness, accountability, transparency, and explainability. It explores the socio-technical context in which data is generated and used, and addresses ethical challenges in real-world settings.

Learning Units

The following learning units cover key aspects of designing and practicing human-centered data science. They explore the motivations behind this approach, the socio-technical context in which data is generated and used, and how to address ethical challenges in real-world settings. Topics include different types of bias, the complexity of fairness, and strategies for transparency and accountability. Learners will examine methods for generating explanations, the role of interfaces in making these understandable, and current research shaping the field of human-centered data science.