An EU long-term research i3 project between Swedish Institute of Computer Science (Dr. Kristina Hook) and Napier University (Prof. David Benyon)
Duration of Project: July 1997 - July 1999
Description of project and further particulars
The issue of how users can navigate their way through large information spaces is crucial to the ever expanding and interlinking of computer systems. Computer users live in a world of information spaces. One of the most critical activities which users need to undertake is to retrieve information from such spaces and thus the problem of how to help the user to navigate, explore and identify the objects of interest is critical to the success of the system.
We shall investigate a new approach to navigation, based on a personalised and social navigational paradigm. Most information retrieval in the ‘real’ world is accomplished through communication between people. We trust certain individuals to possess the information we are looking for. In addition, we expect them to be able to express the information so that it becomes personalised to our needs, understanding and abilities. Often the information seeking is done through talking to several persons, comparing the advice given, reformulating the original need for information, and only sometimes turning to other information sources such as books or on-line databases.
This project seeks to develop our understanding of human activities in information spaces. Specifically we will:
· Create a ‘navigational instrument’: a tool/method that can help designers of systems that include navigational aspects, to choose appropriate metaphors and navigational aids, and to design tools. This instrument will be a computer-based system similar to e.g. the Cognitive Walkthrough method, only it will be directed at identifying and understanding the particular problem of navigation in information spaces. It will build upon a review of differing approaches to issues of navigation in spaces, drawing upon e.g. architectural design, semiotics, sociology, cognitive science (both traditional and modern), geography (both traditional and modern), linguistics, urban studies and spatiology, narrative approaches and other social forms of navigation.
· Design solutions that implement support of social navigation. We shall study social navigation as it happens in existing information spaces. Isolate what makes it appealing to users and then design solutions that support social navigation from this basis.
The project will achieve these aims by using a wide variety of methods of investigation including ethnographically-based approaches to studying navigation in the real world and experimental studies of users navigating in information spaces. We intend to complement such studies with detailed and wide ranging literature reviews and prototype- and/or Wizard-of-Oz based studies of socially-based interaction.
Our results will have an impact on the design of the navigational aspects of information spaces. The aim is to bring social navigation on the agenda as one possible design solution for how to help users.
The result of our work will be communicated through scientific reports, prototype implementations built, described and used, and through the usage, testing and spread of our navigational instrument (for evaluating design of the navigational aspects of systems) to designers.
The issue of how users can navigate their way through large information spaces is crucial to the ever expanding and interlinking of computer systems. Computer users live in a world of information spaces. One of the most critical activities which users need to undertake is to retrieve information from such spaces and thus the problem of how to help the user to navigate, explore and identify the objects of interest is critical to the success of the system.
The existing means for aiding users in finding information is based on cognitively highly demanding information retrieval techniques. Users are expected to formulate their information needs in abstract languages (as with databases or search engines) or to navigate through large information space without loosing track of where they are and were to go next (as in hypermedia systems).
We shall explore existing knowledge of navigation as well as investigate a new approach to navigation, based on a personalised and social navigational paradigm. When exploring the existing knowledge, we shall compile a framework that can serve as a basis for a method that guides the design of and evaluates the effectiveness of navigational aspects of system design.
When investigating the new approach to navigation, we are inspired by what people actually do when they search for information. Most information retrieval in the ‘real’ world is accomplished through communication between people. We trust certain individuals to possess the information we are looking for. In addition, we expect them to be able to express the information so that it becomes personalised to our needs, understanding, and abilities. Often the information seeking is done through talking to several persons, comparing the advice given, reformulating the original need for information, and only sometimes turning to other, ‘objective’, information sources such as books or on-line databases. We aim to explore these characteristics of information search in order to design new tools that can better aid naïve, inexperienced, users to find their way, explore and identify objects in information spaces.
This project seeks to develop our understanding of human activities in information spaces. Specifically we will:
· Create a ‘navigational instrument’: a tool/method that can help designers of systems that include navigational aspects, to choose appropriate metaphors and navigational aids, and to design tools. This instrument will be a computer-based system similar to e.g. the Cognitive Walkthrough method, only it will be directed at identifying and understanding the particular problem of navigation in information spaces. It will build upon a review of differing approaches to issues of navigation in spaces, drawing upon architectural design, semiotics, sociology, cognitive science (both traditional and modern), geography (both traditional and modern), linguistics, urban studies and spatiology, narrative approaches and other social forms of navigation.
· Design solutions that implement support of social navigation. We shall study social navigation as it happens in existing information spaces. Isolate what makes it appealing to users and then design solutions that supports social navigation from this basis.
Our first research challenges is to find a framework that can unite several different perspectives (from architecture, films, navigation in cities, in linguistics, etc.) on navigation. Since our aim is to provide help on how to design tools for navigation, this framework must give us the basis for an instrument by which we can analyse a system and classify/structure/understand the navigational aspects of it.
The idea of socially-based navigation is new. The only other mention of it is by Andreas Dieberger (1997) who uses it slightly differently, but partly aiming at the same phenomenon, namely that people do not search for information in an ‘objective’ manner. Information is often sought based what other persons that we trust have found interesting, valuable, reliable etc.. That is, information retrieval is ‘subjective’. Furthermore, our perception of whether the information found is indeed interesting, valuable, reliable, etc. is also ‘subjective’. A lot of the time people 'ask around' to find information, and the replies they get are ‘personalised’ to fit with their needs, knowledge, the context they are in, or other circumstances.
So the second research challenge for PERSONA is not only to isolate why social navigation is used in the ‘real world’ and what it consists in, but to take our understanding of it to a design scenario and apply it. How can we make a system support users to navigate in a socially-based manner? How can the search be enhanced and become more ‘subjective’?
Our last research challenge is then to show that this form of navigation is in fact providing more mileage or better searches for users than ordinary navigational tools (such as maps, overviews, dialogue histories, etc.). An interesting aspect would be to see whether individual differences are of less importance when in socially-based navigation than it is in more spatially demanding systems. In the latter, we know that the differences are large: the slowest subject takes about 20 times longer to find information than the fastest.
The project will reach its objectives through implementing tools for navigation. It will use a wide variety of methods of investigation including ethnographically-based approaches to studying navigation in the real world and experimental studies of users navigating in information spaces. We intend to complement such studies with detailed and wide ranging literature reviews and prototype- and/or Wizard-of-Oz based studies of socially-based interaction. The use of different methods is described in each task of the workpackages below.
The two objectives of PERSONA, the navigational instrument and the social navigation design, are both measured against the same set of criteria:
· The two tools must be able to relate our understanding of human navigational behaviour to design of tools in various different systems (hypertext, VR, etc.).
· The way to measure the success of one design of will be both in terms of task completion time, but also in terms of satisfaction, understanding of the space, number of errors, and other measures relating to their subjective satisfaction of the navigation through the space. It is not always the time spent in the system that will be the best measure of the success of it.
· The two tools should show how they relate to the framework outlined in the beginning of the project since it provides us with a classification of the typical tasks users are involved in when navigating information spaces: e.g. exploration, wayfinding and identification of objects. It shall be made clear when one design can be preferred over another, and, in particular, when social navigation comes into play.
These criteria will be considered when we design our empirical testing of the two tools.
The first yardstick for the navigational instrument comes early on in the project where we based on our knowledge of design of navigational systems will produce a first outline of the instrument. The instrument will then be evaluated and refined by using it in a number of settings (such as hypertext systems, VR-systems, DB-design, etc.). The instrument should be effective in distinguishing between systems and in helping designers to understand and provide appropriate navigational support usable - in that it is not too cumbersome or long-winded and it should be useful for designers.
We will document the effectiveness and usability of the final version of the instrument and measure these by using questionnaires and other evaluation techniques to establish the usefulness of the instrument.
The social navigation tool(s) has basically two yardsticks. The first comes when the design document is produce in WP2, task 2. The design will be based on the explorative study on how and when people employ a socially-based approach to finding information and navigating in information spaces. This study together with the framework from WP1 will provide the basis for our design.
The second yardstick for the social navigation design comes at the end of the project where the evaluation of its usefulness is done. Since social navigation is a new concept, the evaluation of the final results will look somewhat different than the evaluation of the navigational instrument. We aim to complete a comparative study where a more social paradigm for interaction is compared to a traditional, spatially-demanding, design.
To reach the objectives, and in order to work efficiently, we have divided the research issues into two main workpackages:
WP1 Development of Navigational Instrument
WP2 Design of Support for Social Navigation (Social Navigation)
The two workpackages correspond to the two main objectives of the project. Each is divided into a set of tasks that correspond to a prestudy / implementation / evaluation cycle. They focus on a dedicated problem with relations and dependencies clearly identified.
Each workpackage has its own independent main result that will be demonstrated separately. They will both deliver reports, software and evaluation studies that relate to their main objective. Both partners in the PERSONA project will be involved in both of the workpackages even if we have divided the responsibility so that Napier takes overall responsibility for the first, and SICS for the second workpackage. This close cooperation between Napier and SICS ensures that our research activities will be coherent and the main results of the two workpackages will feed into each other.
In both workpackages the PERSONA project will gain much mileage from seeking cooperation with the other I3 projects. We have initiated such contacts and aim for close cooperation with (to start with) Co-NEXUS, eRENA and eSCAPE. This will also reduce our workload on the implementation side. Rather than building the whole environment for navigation, we plan to build our results on top of other I3 project systems.