Dr. Frank Ferraro, Assistant Professor, CSEE 
1 pm – 2 pm Friday, November 10, 2017, ITE 325, UMBC
A goal of natural language processing (NLP) is to design machines with human-like communication and language understanding skills. NLP systems able to represent knowledge and synthesize domain-appropriate responses have the potential to improve many tasks and human-facing applications, like virtual assistants such as Google Now or question answering systems like IBM’s Watson.
In this talk, I will present some of my work—past, on-going, and future—in developing knowledge-aware NLP models. I will discuss how to better (1) encode linguistic- and cognitive science-backed meanings within learned word representations, (2) learn high-level representations for document and discourse understanding, and (3) how to generate compelling, human-like stories from sequences of images.
Dr. Frank Ferraro is an assistant professor in the CSEE department at UMBC. His research focuses on natural language processing, computational event semantics, and unlabeled, structured probabilistic modeling over very large corpora. He has published basic and applied research on a number of cross-disciplinary projects, and has papers in areas such as multimodal processing and information extraction, latent-variable syntactic methods and applications, and the induction and evaluation of frames and scripts.