Making Sense of Cultural Data:
Combining Humanities and Machine Learning Approaches to Analyze News, Film, Television, and Social Media

Course number: NWMEDIA 190 / THEATER 166 / NWMEDIA 290 / THEATER 266
Instructor: Abigail De Kosnik

Making Sense of Cultural Data” is a seminar open to undergraduates and graduate students in Fall 2016. Students will analyze data from news, films, television, and social media, combining humanities and machine learning approaches. Students will learn how to craft high-level humanities queries, and design effective queries for large databases of words and images.

(Undergraduates who wish to enroll in this course must contact the instructor at adekosnik@berkeley.edu and either: a) state that they have already completed Data 8 (CS 8/Info 8/Stats 8), giving the semester they took the course and the grade they received, or b) obtain instructor approval to enroll. To obtain instructor approval, undergraduates should submit a 1-page, single-spaced application to adekosnik@berkeley.edu stating their reasons for wanting to take the course and any background they have in data science or digital humanities — however, students do not need to have any experience in data science or digital humanities to be considered).

This course will invite up to 15 upper-division undergraduates and 15 graduate students to form small (five-person) research teams that formulate humanities-based research questions that they will answer by analyzing large news service databases, films and television series, and the social media platform Twitter.

Students will learn how to craft high-level humanities queries about cultural and social texts, how to design specific and effective queries for large databases of words and images, and how to link those two lines of investigation. Students will collaborate with one another and with tool developers (whom I will invite as guests to the seminar) to conduct their data analyses and produce visualizations. The student teams will then co-author papers that they may consider submitting for conference presentations and/or journal publications. I will teach a curriculum that orients students to relevant humanities and data methods generally, and then will closely advise and guide student teams’ research projects, and facilitate their cooperation with tool developers.