Nielsen [NIELSEN, 2000] reports a study into Wap usability that was conducted in late 2000. Since Nielsen is acknowledged as the leading specialist in the field of Wap ergonomics and given that the study in question was extremely thorough, this report is our exclusive reference point. This study reveals that, in its current state, Wap utilisation is not satisfactory. During their information-seeking tasks , users are regularly confronted with a range of problems that occur at various levels and relate, for example, to connection, navigation or information retrieval. These problems are due to a large extent to the fact that Wap is still a very new technology and it is to be expected that the connection and download problems will be solved within a few years . However, the problems relating to information searches and retrieval will not be entirely resolved by technological advances. These are design problems which will have to be solved by designers and engineers . The constraints due to the size of the screen and the limitations in the way these tools can be handled (small keypad) would seem to make it clear that the data available via Wap and the navigation options that can be used to obtain this data should be specific to this type of tool. However, it would appear that this is not always the case.
Wap navigation is based on the same principles as Internet navigation. The idea is that users choose options from a series of menus until they find what they are looking for.
Each of the Wap networks attempts to provide all the information that is likely to be of interest to its customers. To do this, it offers all the services that users might need in the form of lists and makes these available directly via the gateway. The limited screen size makes it impossible to display all the options clearly and precisely in a single screen. These are therefore grouped into simple categories in a portal (news, entertainment, etc.). Once the required option has been found, a sublist with a new set of options is displayed. Given the large number of available options, users have to pass through many menus and submenus before they access the desired information. However, they often do not get that far after being prematurely disoriented in their search. Users have also pointed out another problem: some of the options or sites that are proposed are actually links to non-existent addresses or sites that are currently being built. Wap therefore offers its users data which is not actually available.
Apart from the data which is directly accessible via the options, users may also need to perform specific searches and they are able to use search engines to do this. However, these functions are not easy to find and few users access them. This is another consequence of the size of the screen since even if these search engines are available, they are not indicated clearly. To give an example, there are six good search engines available on Wap, yet during the study only five out of 20 users found them. After finding the search engines, the users had considerable difficulty formulating their requests . The small size of the keypad makes it difficult to enter data and increases the risk of error.
After searching for information, users have to process it. In Wap, the data is distributed over several screens that users can browse through using a scroll bar. This confronts users with new difficulties. The first is the use of these scroll bars which they consider to be hard to manipulate. The second is reading the screen display. The ability to read small screens differs from user to user . Some are prepared to read large amounts of text while others are not. However, for 70% of them the small size of the screen is a consideration which dissuades them from using Wap.
Finally, the users were generally fairly critical of the quality of the information they obtained, finding most of it to be unsatisfactory. It should be noted that these results were obtained in a context in which the range of available services was still relatively restricted (December 1999) and was offered via a mobile telephone. The study conducted by Salembier and co-workers [SALEMBIER, this volume] seems to indicate that PDAs, on the one hand, and multimodal techniques (voice, tactile, gestural, embodied), on the other, may provide users with a more satisfactory solution.
Nielsen's study brings to light two positive points: the use of mobile telephones for Wap access is learnt easily and users can easily remember the various functions of the buttons . However, overall, Wap is affected by serious problems of usability, in particular when we consider the criteria of efficiency, the management and prevention of errors, and user satisfaction. Users are confronted with many difficulties, lose their connections and the few instructions that do exist do not enable them to take sufficient control of the system and make it truly effective. The navigation and option labelling often make the user/machine interface totally uncommunicative . Searching for information is a painstaking task which all too often yields unsatisfactory results.
It might be imagined that tools that allow subjects to customise their portals and the various download sites would represent a considerable aid to users in allowing them to access the options they require much more swiftly. However, this is not enough. These tools must also be capable of helping users perform their information-seeking tasks in a much more effective way. We can well imagine that interfaces that adapt automatically to the user's operating habits and query types (adaptive interfaces) would be extremely useful here. Going further, intelligent agents could allow users to simply describe what they are looking for and then let the system perform the search for them. Such tools would probably make Wap simpler to use and lead to more satisfactory search results. However, Wap does not enable users to perform all the information searches that are available on the Internet and adaptive interfaces do not provide the necessary performance for all types of information search. In effect, the integration of adaptive interfaces for Internet-based information-seeking tasks makes it necessary to define the type of task for which they are to be adapted . The problem concerning the Internet here is the same as the one we have already formulated for Wap: what are the types of information-seeking task for which these communications objects (mobile, PDA, laptop PC) or these protocols (Web, Wap) are suitable? Before we can answer these questions we must be able to provide a description of the information-seeking tasks.
As a first stage, be intend to study (a) the implementation of the goal or "the objective characteristics of the information-seeking task", and (b) the effects of these characteristics on users' activities. We believe that some of these characteristics will correspond to cases in which an adaptive interface is useful and, indeed, to cases in which a particular communications object makes it possible to search for information effectively.