Towards More Robust Natural Language Understanding

Abstract

Natural Language Understanding (NLU) is a branch of Natural Language Processing (NLP) that uses intelligent computer software to understand texts that encode human knowledge. Recent years have witnessed notable progress across various NLU tasks with deep learning techniques, especially with pretrained language models. Besides proposing more advanced model architectures, constructing more reliable and trustworthy datasets also plays a huge role in improving NLU systems, without which it would be impossible to train a decent NLU model. It's worth noting that the human ability of understanding natural language is flexible and robust. On the contrary, most of existing NLU systems fail to achieve desirable performance on out-of-domain data or struggle on handling challenging I feel incredibly fortunate to have Dr. Huan Sun as my advisor, without whom nothing in this thesis is possible. I would like to express my sincere gratitude to her for critiquing my work and my ideas in a constructive way. Her vision and rigorous research attitudes have shaped my thoughts. I am always indebted to her for guiding me all the way, for her contagious energy, for being so supportive and caring about students.I owe a great debt of gratitude to Dr. Marie-Catherine de Marneffe for her countless help. I am super grateful for her many invaluable insights and suggestions on my work. She has been so generous with her time, reading, reviewing and commenting on many of my writings. I am always thankful for her positive encouragement and praise.