February 8th, 2024 at 17:30 on Zoom

Meeting ID: 283 744 3344

Password: 2354290

Conversational systems are gateways to knowledge and community. They influence our social interactions. As important and influential as they are, they are not designed with minority communities in mind and they may not reflect our societal values. In this talk, we will be going over how our dialogue research on American Sign Language, and African-American Vernacular English can improve the dialogue experience for all. We will also talk about the ethical implications of such multimodal dialogue systems and provide suggestions for the future.

Mert Inan is a doctoral student in the Khoury College of Computer Sciences at Northeastern University, based in Boston. He is advised by Malihe Alikhani. Mert has a Master’s degree from Computational Biology Department in School of Computer Science and Mellon College of Science of Carnegie Mellon University. He did research with Prof. Tai-Sing Lee on Computational Neuroscience. His Bachelor’s degree in Computer Science is from Bilkent University, Turkey.

Mert’s doctoral studies focuses on natural language processing, computational linguistics, and multimodal conversational AI. He is particularly interested in interdisciplinary projects with real-world impact; for example, he has worked on systems that combine cognitive science with conversational artificial intelligence systems, such as sign language processing and multimodal evaluation systems that combine visual and linguistic information. He hopes to make these conversational systems more human-like and cognitively aware, and to design dialogue systems for use by the deaf and hard of hearing community. Mert also served as TA for many courses including Theory of Computation, Foundations of Computer Science, Natural Language Processing, Computer Graphics, where he can gain additional exposure to practices that he employs in his research. 

Mert’s work has received several recognitions, including the Amazon Alexa Prize TaskBot Challenge Grant after his team came in third in Amazon’s TaskBot challenge. He has also published in ACL, NAACL, EMNLP, SIGDIAL, COLING and CogSci. Going forward, Mert is looking for new interdisciplinary applications for his computational linguistics know-how, with an eye toward improving quality of life. In his free time, Mert loves cooking and baking international gourmet cuisine.

Personal Website: https://merterm.github.io/

Katherine Atwell is a doctoral student in the Khoury College of Computer Sciences at Northeastern University, based in Boston. She is advised by Malihe Alikhani.

Atwell started her doctorate in natural language processing (NLP) at the University of Pittsburgh in 2020, then followed her lab and advisor to Northeastern in 2023. Since joining, she has enjoyed Khoury College’s atmosphere and opportunities for interdisciplinary research. She is especially interested in solving NLP ethics problems, including analyzing bias in language models and tackling the problem of automatic social media moderation. Atwell is fascinated by the topics’ interdisciplinary nature, the way they incorporate formal linguistics theories, and their potential for social good. 

Before joining Khoury College, Atwell worked on HCI and user experience projects as part of the Research Experience for Undergraduates program at Iowa State University. She also interned in software engineering at Google for two summers, placed third in the Alexa Prize TaskBot Challenge, and won a Best Paper Award at UAI 2022.  

Outside of work, Atwell enjoys bouldering, singing, biking, being outdoors, playing the piano, and video games. 

• Organized by Robin Downey  (STS Instructor) and  Zeynep Dellal (Computer Science Student)