SmH, smart help : an application of user models

The development and use of domain models is adapted here to the design of an "intelligent" help and error diagnostic system. A simple implementation of the system, its models, and inference engine are demonstrated. The concept of domain modeling means that information about a certain topic is utilized by a program in such a way that the program bases decisions on the contents of the data. It is also to be expected that this data will be undergoing continuous update, causing the aforementioned program to respond differently even when presented with the same inputs at different times between which changes have occured in the data associated with the domain model. This concept is not new but few systems have been built that utilize this idea in a robust manner to model a users knowledge, interests, and skill. In essence a system that is based on this concept would be identified as being part of the family of AI based, or Expert, systems. The use of modeling does not necessarily dictate that the resulting system is an AI system but it exhibits some of the properties associated with AI. Models can be created that are representative of topics in many areas. They are limited only in that the domain being modeled be expressible in quantitative terms i.e. equations or data structures that can be manipulated by a program.