Abstract

Abstract: In this talk, I will describe the status of the dialogue research in the NLP community by providing the research background, a survey of available resources, and giving key insights to the application of state-of-the-art dialogue methodology into industry-scale systems. Despite the phenomenal potential of deep learning methods in this sector, very few of these approaches are production-ready. I will walk through these approaches and summarise their pros and cons.

Bio: Pei-Hao (Eddy) Su is a co-founder and Chief Scientist of PolyAI, a London-based startup developing the next generation machine learning platform for building conversational user interfaces. Eddy completed his PhD at the University of Cambridge, working with Professor Steve Young. His research interests centre on applying deep learning, reinforcement learning and Bayesian approaches to dialogue management and reward estimation. He received the best student paper award at ACL 2016 on his dialogue work and has given a tutorial on deep learning for conversational AI at NAACL 2018 . He was a research intern at Facebook AI Research in Summer 2017.