Complementary Views

The graphic below visualizes three complementary views in research on interactive intelligent systems, all of which will be found in TiiS articles:

Focus on Intelligent Technology

At one end of the spectrum lies research that focuses primarily on understanding or improving intelligent technology for interactive systems. It often takes as its point of departure some form of interaction with intelligent systems that has already been realized successfully and aims to enhance it by improving the technology (e.g., by providing faster learning, more accurate recognition, or more reliable prediction).

Focus on Users' Interaction With Intelligent Systems

At the other end of the spectrum, we find research that takes as a starting point some existing intelligent technology and aims to find better ways of interacting with systems that incorporate such technology (e.g., by exploring new interface designs, interaction styles, or usage contexts).

Binocular View

In the middle of the spectrum, we have research that aims to advance understanding of both intelligent technology and users’ interaction with it – for example, by looking for better combinations of intelligent technology and interaction (e.g., a novel algorithm accompanied by a new interaction design that enables people to use it effectively).

Ensuring Relevance to the Journal

Articles anywhere on this continuum can make valuable contributions, and all of them are welcome in TiiS. But there is an important requirement: A manuscript that focuses almost exclusively either on intelligent technology or on interaction must also include at least a brief discussion of the “other” side of the picture, so as to make clear the relevance of the article’s contributions to the general issues addressed by TiiS. (See the subsection “Explaining Relevance” in the Authors’ Introduction for hints about how to formulate such a brief discussion.)  

Common Issues Concerning Interactive Intelligent Systems

Although interactive intelligent systems can take many specific forms, there are some general questions that arise whenever human intelligence comes into contact with machine intelligence, such as the following:

• In what ways can artificial and human intelligence work together effectively?

  For example, where does machine intelligence offer the greatest added value? How can effective mixed-initiative interaction be achieved?

• What challenges to usability and acceptability can the incorporation of intelligence in an interactive system raise, and what strategies are effective in meeting these challenges?

  For example, how can it be ensured that users have adequate forms and levels of control over an autonomously acting system’s behavior and that responsibility is assigned appropriately?

• What types of methodology for research, design, and evaluation are especially suitable for interactive intelligent systems – and how can such methodologies be improved?

Although most individual TiiS articles will focus on specific instantiations of questions like these in the context of particular systems or studies, the whole collection of TiiS articles will advance our general understanding of these issues.