Contents Speech recognition has become a part of our daily lives. Today, we can enjoy a wide array of conversational user interfaces and speech-based interactive services, from speech-based personal assistants in our smartphones and smart speakers, e.g., Amazon Echo, Alexa, Cortana, and Google Home, to chatbots in traditional messaging interfaces. These interfaces are becoming pervasive as reflected by the exponential growth in their adoption. Hence, the expectations are that computer-based conversational systems will soon reach the level of human–human conversation.
Developing human–computer interactions with natural conversational characteristics is, however, a big challenge. Conversational interfaces and interactions are very different from their graphical counterparts. Advances in natural and spoken language processing have given us powerful tools to handle spoken and typed words. Yet, using these tools to create natural conversational interfaces requires further advances in interface design, language technology, and tools for development and testing. In short, natural human conversation is highly complex and, hence, provides a fertile space for multidisciplinary research.
The papers in this special issue explore applications, systems, methodologies, and technologies that relate broad aspects to realize successful conversational user interfaces and interactions.
In the first article, “A conversational robot to conduct therapeutic interventions for dementia,” Dagoberto Cruz and Jesus Favela describe a case of conversational robot interactions. They explore the potential of a robot, developed for therapeutic interventions, to guide the therapy sessions without human intervention and report findings from its deployment in a geriatric residence including automatic generation of a therapy script and strategies to recover from communication breakdowns.
In the second article, “Towards interpersonal assistants: Next-generation conversational agents,” Inseok Hwang, Youngki Lee, Chungkuk Yoo, Chulhong Min, Dongsun Yim, and John Kim present an interpersonal assistant to serve natural human-to-human conversations beyond a personal assistant, unobtrusively and discuss a platform initiative to effectively support the development of interpersonal assistant applications.
In the third article, “Revolution or evolution? Speech interaction and HCI design guidelines,” Christine Murad, Canada Cosmin Munteanu, Canada Benjamin R Cowan, and Ireland Leigh Clark discuss challenges to develop guidelines and heuristics to facilitate the design of conversational, more specifically, voice–user interactions. They critique existing efforts to develop VUI-specific heuristics and to adapt GUI design principles to this space, and argue that the path toward revolutionary new ubiquitous conversational voice interactions must be based on several evolutionary steps that build VUI heuristics off existing GUI design principles.
In the final article, “Accelerating conversational agents built with off-the-shelf modularized services,” Jinho Lee, Inseok Hwang, Thomas S. Hubregtsen, Anne E. Gattiker, and Christopher M. Durham present a novel inference processing architecture to accelerate the responses of conversational agent systems. They develop the concept of speculative inference to overcome the challenge behind common practices in developing conversational agents, i.e., pipelining off-the-shelf modularized services as ready-made building blocks.
We hope that you will enjoy this special issue as much as we have enjoyed preparing it. The research work included here is reflective of our journey toward a richer, more natural and immersive vision of pervasive computing.