In this talk, I will discuss my group’s recent work on using logically-implied textual entailment knowledge to improve a variety of downstream natural language generation tasks. First, we employ a multi-task learning setup to combine a directed premise-to-entailment generation task with the given downstream generation task such as multimodal video captioning (where the caption entails the video) and automatic document summarization (where the summary entails the document), achieving significant improvements over the state-of-the-art on multiple datasets and metrics. Next, we optimize for entailment classification scores as sentence-level metric rewards in a reinforcement learning style setup (via annealed policy gradient methods). Our novel reward function corrects the standard phrase-matching metric rewards to only allow for logically-implied partial matches and avoid contradictions, hence substantially improving the generation results.

Dr. Mohit Bansal is an assistant professor in the Computer Science department at University of North Carolina (UNC) Chapel Hill. Prior to this, he was a research assistant professor (3-year endowed position) at TTI-Chicago. He received his PhD from UC Berkeley in 2013 (where he was advised by Dan Klein) and his BTech from the IIT Kanpur in 2008. His research interests are in statistical natural language processing and machine learning, with a particular interest in multimodal, grounded, and embodied semantics (i.e., language with vision and speech, for robotics), human-like language generation and Q&A/dialogue, and interpretable and structured deep learning. He is a recipient of the 2016 and 2014 Google Faculty Research Awards, 2016 Bloomberg Data Science Award, 2014 IBM Faculty Award, and 2014 ACL Best Paper Award Honorable Mention. Webpage: http://www.cs.unc.edu/~mbansal/