INTERACTIVE ADVISORY SYSTEMS

Smart algorithms can be used to process large amounts of data and come up with an "optimal" solution according to the criteria set up by the model of the system. But how should a user interact with such advisory systems? In this paper, we discuss the ability for users to interact with advisory systems at many levels. Thus, rather than only have the ability to use the advisory system to generate a solution, the user has the ability to 1) ask the system to generate alternative solutions based on different criteria or 2) use the system to evaluate a user-generated plan. This provides the user with the benefits of a more flexible advisory tool. Should the "optimal" solution generated by the advisory system be "non-optimal" from the user's perspective (because of knowledge or parameters not taken into account by the system), the user can either change the criteria used by the system to reflect this new knowledge, or tweak the final solution and still take advantage of the system's capabilities to evaluate and compare the solutions.