Natural Language Understanding in Conversational AI with Deep Learning
Giới thiệu về sách điện tử này
To this end, the book is structured in seven chapters: Introduction to Natural Language Understanding lays the foundation by tracing the evolution of NLU from early human communication to modern human-computer interactions. Prerequisites and Glossary for Natural Language Understanding then serves as a foundational resource, consolidating essential prerequisites and key terminologies relevant across the book. Single-Turn Natural Language Understanding looks at Single-Turn NLU, focusing on tasks that involve interpreting and processing user inputs in a single interaction, while Multi-Turn Natural Language Understanding moves on systems for extended interactions with users and explores techniques for managing dialogues, using context and integrating external knowledge bases. Next, Evaluating Natural Language Understanding discusses the annotation of datasets and various performance assessment methods, covering different levels of understanding from intent recognition to slot filling and domain classification. Applications and Case Studies in Natural Language Understanding subsequently shows real-world applications of NLU in finance, medicine, and law. Eventually Challenges, Conclusions and Future Directions explores the core obstacles hindering the advancement of NLU, including ambiguity, domain adaptation, data scarcity, and ethical concerns. By understanding these challenges, this chapter highlights the ongoing work needed to advance NLU.
This book mainly targets researchers, PhD students, and professionals who are entering this field and look for a state-of-the-art introduction to NLU applied in conversational systems such as chatbots, large language models, or educational systems.
Giới thiệu tác giả
Caren Han is a senior lecturer at the University of Melbourne, an honorary academic at both the University of Sydney and the University of Edinburgh, and an adjunct professor at POSTECH. She is co-directing the Australia Deep Learning NLP Group. After her PhD in 2017, she received several teaching and research awards, including the Australian Young Achiever Certificate (Teaching Excellence), Teacher of the Year 2020, Supervisor of the Year 2021, Best Research Paper Award in top-tier International Artificial Intelligence Conferences, Early Career Research Award 2023. She currently supervises 23 research students, and her research interests include Natural Language Processing with Deep Learning.
Henry Weld has PhDs in both Computer Science and Mathematics at The University of Sydney and is a member of the Australian Deep Learning NLP Group. His research focuses on Natural Language Understanding, particularly multi-turn NLU, and the use of NLU methodologies in other fields where the data has differing granularity based on aspect.
Yan Li is a PhD student at the University of Sydney and is currently visiting the University of Melbourne. Yan is an NLP researcher in the Australia Deep Learning NLP Group, specialising in long document comprehension and reasoning, multi-turn dialogue systems, and multimodal deep learning. Yan’s work focuses on advancing the capabilities of natural language processing through the integration of multiple data modalities and improving comprehension and reasoning over extensive textual content.
Jean Lee is a researcher and data scientist at the Sydney Informatics Hub, a core research facility of the University of Sydney. Her research areas are Natural Language Processing, Information Retrieval, and Artificial Intelligence applications. Prior to academia, she passed the U.S. Uniform Certified Public Accountancy Examination (a.k.a. AICPA) and worked in management consulting firms including Accenture and KPMG.
Josiah Poon is a senior lecturer in the School of Computer Science at the University of Sydney. He co-founded the Australian Deep Learning NLP Group together with Caren Han. His research focuses on having natural language at the hub but integrating with multimodal learning, explainable AI, as well as integrating neural and symbolic approaches.
