The state of the art in computer technology has advanced to the point where systems for generating driving directions between two points are commonplace. There are several web sites offering street-level driving directions, and several in-car systems available as an option on purchased or rented cars. The availability of digital maps and high-capacity, rapid processors enable this technology, but little attention has been paid to ensuring that the interface is flexible enough to deliver satisfactory routes to users who have different preferences.
Current systems for route advice compute solutions using a shortest-path algorithm to find the minimal-cost route from the origin to the destination. Some systems fix the cost as the estimated travel time, while others let the user choose between the shortest path, the quickest, or the ``most scenic'' one. In all cases, the system then describes the route to the user with little or no recourse if the driver finds the route unsatisfactory. These systems disregard the fact that driving occurs in a rich environment where many factors influence the desirability of a particular route. For example, some drivers may prefer the shortest route as long as it does not have too many turns, or the fastest route as long as it does not go on the highway. The relative importance of these factors varies among individuals, and drivers may not know themselves what they value most in routes.
In this paper, we describe the Adaptive Route Advisor, an adaptive user interface [6] that recommends routes from a source address on a road network to a destination address. We transfer the analogy of a trusted travel agent to our route advice domain. Given a travel task, the travel agent plans an initial solution or two, taking into account the client's preferences if known. After the travel agent presents the initial options, the client may request additional solutions that differ from the initial options along some dimension. For example, if the client is booking a flight, he or she may request a flight with a shorter layover, even if that means an increase in cost. The client and the travel agent continue to work together, generating and evaluating different solutions, until the client is satisfied. During this process, the travel agent can reflect on the choices made by the client and refine his or her model of the client's preferences for the next task. This interaction model is similar to that assumed by the Automated Travel Assistant [7] for airplane flights, except that system involves no personalization and the user must explicitly assign values to preferences.
![\includegraphics[width=5.36in]{ARA-arch.eps}](img1.gif) |
The Adaptive Route Advisor is designed for in-car use. It is a Java application that functions as a resource-light network client, suitable for mobile environments with a wireless communication infrastructure. The remote servers provide resource-intensive functions such as routing and geolocation.1 Although the current version does not yet take advantage of information available from mobile deployment (primarily current and past locations from the Global Positioning System) and the interface is not fully optimized for limited input and output resources common in vehicles, future work will embed the Adaptive Route Advisor in a mobile environment.
The pages that follow describe our approach in more detail and present the results of an experiment in personalizing the user model from rankings of alternative routes. First we present the overall system architecture, including the route generation component, the adaptation method that constructs the user preference model, and the user interface that presents route options to the user and gathers preference feedback. We then report on an experiment adapting a preference model to human subjects and its results. Finally, we outline planned improvements to the agent and consider the Adaptive Route Advisor's relevance to other advisory systems.