University of Massachusetts at Amherst
Su Lin Blodgett is a PhD candidate in computer science at UMass Amherst working in the Statistical Social Language Analysis Lab, advised by Brendan O’Connor. She is interested in developing fair, robust statistical text analysis tools to answer social science questions, with a particular interest in how research in computational sociolinguistics and in fair natural language processing systems can support each other. She has received a National Science Foundation Graduate Research Fellowship. Previously, she received her B.A. in mathematics from Wellesley College.
My research is focused on computational sociolinguistics, the emerging intersection between natural language processing (NLP) and sociolinguistics; I aim to develop computational techniques for using large-scale social media datasets to study language variation and change. In particular, I work on identifying and characterizing regional patterns of dialectal variation on social media, as well on characterizing linguistic attitudes–how people think and talk about language varieties–on social media. One significant challenge of computational analysis of social media is that the very NLP systems we want to use to study this linguistic variation often fail in this setting. As accurate sociolinguistic analysis requires unbiased computational analysis, I additionally focus on developing fair NLP systems, and on using insights derived from our sociolinguistic analysis to improve our computational systems.