On conventional computer systems, human interaction is focused on visual perception via a screen in one direction (computer to human), with the keyboard and pointing device in the other direction. Recent developments have led to the widespread use of devices that allow interaction through multiple modalities, such as sound and touch. In this book, Kühnel asserts that despite the many advantageous aspects of multimodal interactive systems, the evaluation of such systems remains largely unaddressed. She defines these systems as having one or more input or output channel, or, in other words, modalities that engage one or more of the human senses. Drawing from contemporary research, Kühnel proposes a system of classification as a launching point to examine the quantification of quality aspects of multimodal interactive systems. This taxonomy consists of “quality factors related to the user, the system, and the context that have an impact on interaction behavior and thus on perceived quality”; “interaction performance aspects describing user and system performance and behavior”; and “quality aspects related to quality perception and judgment.”

Orthogonal to the challenge of systems that combine multiple modalities is the question of assessing user experience when interacting with such systems. While it is reasonably straightforward to measure certain aspects like number of concrete actions (for example, pressing a key, touching a button, or displaying a page of text or an image), those easily quantifiable aspects are not necessarily the main factors leading to a positive or negative user experience.

Kühnel presents a user’s guide to understanding contemporary research on multimodal systems, focusing specifically on unresolved questions regarding the systematic evaluation of multimodal interactive systems. As the book notes: “This work can thus be seen as a first comprehensive application of the taxonomy of quality aspects of multimodal systems.” Kühnel provides an overview of the relevant terms, concepts, approaches, and related work in the field, highlighting quality and usability as key factors in the evaluation of multimodal interactive systems. She defines interaction parameters based on a literature review, locates parameters identified with her taxonomy of quality aspects, and investigates the strengths and weaknesses of interaction parameters as useful guidelines or tools in multimodal system evaluation. She claims: “In order to benefit from interaction parameters, a clever design to induce goal-oriented behavior and judgment has to be developed, or parameters or other measurements ... are necessary also to assess the hedonic aspect of the interaction.” She further details the INSPIRE_Me system as a system built of several pertinent experiments. In support of her work, Kühnel additionally references a framework for multimodal interactive system evaluation (PARADISE), a procedure for multimodal interaction system evaluation (PROMISE), as well as an embodied conversational agent scale (ECAS).

Kühnel’s highly structured organization allows for easy navigation throughout the book. She initially lays the groundwork for chapter breakdowns, and systematically continues to outline the format and structure of the book from start to finish. Each chapter begins with a short introduction and motivation. This is followed by greater elaboration and detail into the research questions, related experiments, and relative findings or conclusions. Each chapter ends with a summary of answers addressing the major research questions. The core of her work thoroughly examines the following major research questions in successive chapters:

• “How can multimodal interaction behavior be quantified?”
• “How can the output quality of multimodal systems be quantified?”
• “How can the input quality of multimodal systems be quantified?”
• “How does interaction quality, input quality, and output quality contribute to the overall quality of multimodal interactive systems?”

Each major question is broken down into a series of more specific questions, which she chronologically addresses and answers.

Kühnel ends with a big picture summary and overview of the main concepts, criticism of applied methods, general findings, and starting points for future work. Overall, she provides a thoughtful and rigorous examination of meaningful questions in contemporary research on multimodal systems. She also proposes answers and points of departure for further work and development surrounding these questions. Kühnel is cohesive, comprehensive, and meticulously systematic in her approach, though perhaps a little dry for the nontechnical reader. The book offers useful insights to those readers interested in learning about multimodal interactive systems and never-before-seen approaches to the evaluation of these systems. It outlines steps toward measuring certain aspects of the user experience for interactions between humans and computers. Major challenges remain, however, with respect to assessing multimodal interactions and quantifying the interaction experience.