A Review of Quality Assurance Research of Dialogue Systems

Abstract

With the development of machine learning and big data technology, dialogue systems have been applied to many fields, including aerospace, banking and other scenarios that require high accuracy of answer. This has prompted a great deal of research on quality verification and assurance of dialogue systems. As two means to ensure the quality of software, testing and evaluation are rarely comprehensively summarized in current research work. Firstly, the dialogue systems are classified according to different classification standards. Secondly, this paper reviews the existing quality assurance work of dialogue systems from testing and dialogue evaluation, including testing methods, testing tools, evaluation metrics and dialogue quality attributes. Moreover, the issues and needs are discussed aiming at the deficiency in the current work, which can provide references for future research.