CUNY-NLP Talk Recap: Dr. Elena Filatova’s Look at Sarcasm

The first CUNY-NLP Talk was held this Monday, September 26, 2016 at the CUNY Graduate Center. The CUNY-NLP Talk Series is a student organized series of lectures, which cover various topics of Natural Language Processing research. Invited speakers are usually experienced academics and industry folks, with a mixed audience of computer scientists and (computational) linguists in attendance.

We started off the semester with an interesting (and funny!) talk from Professor Elena Filatova. Dr. Filatova received a BS in Applied Mathematics from Moscow State University, an MS in Computational Linguistics from the University of Southern California, and a PhD in Computer Science from Columbia University. Prior to joining CUNY CityTech, she spent time as a faculty member at Fordham University’s Computer and Information Sciences department. Her research focuses on Natural Language Processing theory and applications and her talk focused on automatic sarcasm detection.


The talk provided a great introduction into sarcasm research, discussing important questions, like “how do we define sarcasm?”, which by the end of the talk is clear to be no simple task. She also raised larger and very current questions like, “should we use feature engineering or deep learning?”  Using evidence from her research on predicting sarcasm in Amazon reviews, Elena showed that this task is feasible, but difficult.  If you are interested in sarcasm, language, computer science, or NLP and want to see Elena’s full talk, her slides are available to view (below). If you interested in Elena’s research and want to learn more, feel free to shoot her an email:

The CUNY-NLP Talks usually take place once a month. Our next speaker will be Alla Rozovskaya from Queens College. For more details, please check out the CUNY-NLP Talk Fall semester schedule. Hope to see you at the next one!