Human Computer Interaction
ERCIM News No.46, July 2001 [contents]

Navigating Information Spaces

by David Benyon

New HCI challenges arise from the emergence of information spaces and the related concept of navigation. Research at Napier University focuses on this topic.

Evaluation is a fundamental part of human-computer interaction (HCI). Good HCI practice tells designers to evaluate: evaluate requirements, evaluate designs, evaluate prototypes. The purpose of evaluation is to improve the usability of a software system; that is to make it easy to use, easy to learn, effective and enjoyable. But what is usability and what makes one device easier to use than another? Traditional HCI theory has produced a number of evaluation techniques and guidelines. These are based on some basic psychological assumptions which date back to the sixties.

Broadly speaking it is assumed that people have goals that they are trying to achieve and the reason that they are using a software system is to enable them to achieve those goals. People have to translate their goals into actions and then evaluate the response of the system to determine if their goals have been achieved. People have, therefore, to bridge the ‘gulf of execution’ (translating goals into actions) and the ‘gulf of evaluation’ (perceiving and interpreting what has happened). A practical method of evaluation, the Cognitive Walkthrough method, is based on these ideas. There are other approaches to evaluation which involve coming up with a set of good design ‘heuristics’. Don Norman spurned the cognitive psychology and Jacob Nielsen gave HCI the most famous set of heuristics. Norman & Nielsen now travel the world advising on usability.

These ideas of usability and evaluation developed during the hay-day of HCI, the decade from the mid eighties. These were the days before the Web, when people went to work and sat in front of computers. These were the days before everyone used Microsoft Office, when programmers developed applications specifically for their users. The twenty-first century is not going to be like that (at least not the first part!) and it is a moot point whether the concepts and methods of usability developed during the eighties will transfer. For example, the design of Web sites, virtual environments, distributed computing environments and novel settings such as households all throw up usability challenges that traditional methods seem ill-equipped to deal with. Navigation in Web spaces and other hypermedia spaces is a key feature, yet counts for only one of Nielsen’s heuristics. ‘Know your user’ is another heuristic; yet how can you know the people who visit a web site? Navigation is even more of an issue in virtual environments, where there are many more degrees of freedom and where users more literally move to where they want to go. Ideas of tasks and goals do not transfer easily to environments such as the household where entertainment and relaxation are key, and style and fashion are significant factors. With distributed computing environments people are literally surrounded by devices that are communicating autonomously with each other. People will be wearing computers and exerting their identity. These large open ‘information spaces’ have quite different usability requirements and hence require different evaluation mechanisms. One very important aspect of these systems is that we need to consider accessibility in addition to usability.

The traditional view of HCI sees the person outside the computer, looking onto a world of information. An alternative is to see people as inside their information space. This in turn leads to the idea that people are navigating through this space. An information space is made up from information artefacts which are the devices we use to help us undertake activities in the ‘real world’.

The concept of an information ‘space’ leads naturally to consider ideas of navigation. In the real world we navigate quite happily. We also know a lot about how to design spaces such as museums, galleries, and so on to assist navigation and to make navigation a pleasurable experience. Research looking at how to take knowledge of signage, maps and so on to evaluate and inform the design of information spaces is on-going.

More than that we recognise that in the real world we do not just follow signs to get around — we ask people, follow people, make recommendations to our friends. We are also looking at issues in ‘social navigation’. Direct social navigation through discussions and indirect social navigation through awareness of others and of history.

One approach to dealing with the difficulties of navigation in information spaces is to provide more personalised help and assistance through the use of intelligent agents and recommender systems to help navigation. Other ways of helping people find their way is through signs, maps and landmarks. The new environments that we have to design for demand new approaches to the ideas of usability and of how systems are evaluated. The ‘navigation of information space’ view of humans and computers in the twenty-first century provides a new perspective on usability: design for people to live in their information spaces.

Please contact:
David Benyon — Napier University Edinburgh, United Kingdom
Tel: +44 131 455 4294