Deep Learning for Dialogue Systems: Chit-Chat and Beyond

The emergence of dialogue systems that supports extensive topics and chit-chat conversations in natural language processing (NLP), information retrieval (IR), and machine learning (ML) is surveyed. The survey explores techniques to build chit-chat systems and chit-chat components in completing IR tasks, including deep neural models, task-oriented systems, and data resources and evaluation methods. It also analyzes the main challenges, possible new exploration directions, and rising trends in building human-like systems. The survey aims to bridge the IR and NLP communities and provide insights for researchers, engineers, and anyone interested in advancing chit-chat systems.

Format: Paperback / softback
Length: 186 pages
Publication date: 30 June 2022
Publisher: now publishers Inc

The rapid advancement of deep neural models and the abundance of data resources have given rise to dialogue systems capable of engaging in extensive topics and engaging in chit-chat conversations in the fields of natural language processing (NLP), information retrieval (IR), and machine learning (ML). In order to facilitate the development of both retrieval-based chit-chat systems and IR tasks that leverage these systems, the authors conducted a comprehensive survey that explores these systems from two perspectives. The first perspective focuses on the techniques used to build chit-chat systems, while the second perspective examines the role of chit-chat components in completing IR tasks.

The primary contributions of this survey lie in its comprehensive coverage of deep neural models, its connection of recently resurgent chit-chat systems with task-oriented systems, its introduction of various solutions for challenges from different perspectives, including data-side and model-side solutions, and its utilization of extra resources. The survey also presents data resources and evaluation methods for building retrieval-based and generation-based chit-chat systems. Furthermore, the authors analyze the main challenges, potential new exploration directions, and rising trends in the field, which will provide valuable insights into building human-like systems.

The survey is designed to bridge the gap between researchers in IR and the NLP community, enabling them to advance chit-chat systems and support a wider range of IR tasks. It will be of interest to IR or NLP researchers who wish to study chit-chat from different perspectives, IR researchers who require the assistance of chit-chat systems to complete their tasks, engineers with hands-on experience in building these systems to leverage advanced chit-chat modeling techniques, or anyone who wants to stay up-to-date with the latest developments in chit-chat systems or learn how to build them with deep neural architectures.

Weight: 270g
Dimension: 234 x 156 (mm)
ISBN-13: 9781638280224