Obfuscating Gender in Social Media Writing
Researcher, Wellesley College
Friday March 3, 3:15 in Volen 101
The vast availability of textual data on social media has led to an interest in algorithms to automatically predict demographic attributes based on the user’s writing. These methods are valuable for social science research as well as targeted advertising and profiling, but also compromise the privacy of users who may not realize that their personal idiolects can give away their identities. Can we automatically modify a text so that the author is classified as a certain target gender, under limited knowledge of the classifier, while preserving the text’s fluency and meaning? In this talk, I present a model to modify a text, show empirical results with Twitter and Yelp data, and outline future directions.
Sravana Reddy is a researcher at Wellesley College working on natural language processing and its intersections with privacy, the digital humanities, and sociolinguistics. She graduated with a PhD from the University of Chicago, and a bachelors’ from Brandeis University.