User Generated Dialogue Systems: uDialogue

This article introduces the idea of user-generated dialogue content and a framework that makes it workable in practice, and describes our experiments aimed at clarifying the conditions that must be met by such systems. One of the appealing points of a spoken interface is that it provides a vivid sense of interactivity that cannot be achieved with a text interface alone. In this study, we will separate dialogue systems into content that can be produced and modified by users, and the systems that drive this content, and we will seek to clarify (1) the requirements of systems that enable the creation of engaging dialogue systems, and (2) the conditions for the sequential generation of engaging dialogue content by users, while at the same time establishing a framework for doing so. For bidirectional digital signage with voice guidance installed in a public space as a dialogue device, we implement an Internet-based content generation environment, and we perform a validation experiment where dialogue content is generated by users. The framework for dialogue content generation proposed in this study is expected to lead to a breakthrough in the spread of spoken content.