Understanding Unanticipated, Social Consequences of Popular Algorithms
Mentor: Dr. Shion Guha
Description: Data-driven approaches have become common in technology and social computing to analyze and infer social behavior of people in large scale networks. However, data-driven algorithms are not impartial much as the popular narrative might imply. In this project, we will examine the unanticipated effects of popular data-driven algorithms on publicly available social and networked data. By doing this, we will learn how the biases and assumptions of data scientists, algorithm designers and data-driven decision makers affect the overall behavior of different algorithms leading to unforeseen, future consequences that affect the lives and behavior of the people being analyzed and examined. We will make inferences about society and technology at large from our results and prepare a poster submission for an appropriate conference (CSCW’18 or GROUP’18). The student will receive full credit and first authorship on this poster upon successful completion of the project.
Preferred qualifications: Some programming experience (formal courses, moocs, self taught etc.). Interest in data science and social behavior. Some previous experience in statistics/machine learning is useful but not required.