Lecture, three hours. Requisite: one course from Computer Science 31, 32, Program in Computing 10A, 10B with grade of C+ or better, or equivalent. Examination of how large-scale data can be used to systematically measure various aspects of human activities. Review of series of computational and statistical methods which enable scalable analysis and cost reduction. Students learn to interpret and understand research findings and implications from published work. Review of ethical issues in data science, such as privacy and model biases. Investigation of limitations and risks of current methods. Discussion of various ways to improve transparency and accountability of data-driven research. Letter grading.