Submitted to: Artificial Intelligence Applications
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/1/1996
Publication Date: N/A
Citation: Interpretive Summary: Effective management of natural ecosystems requires the ability to consider the ongoing functions of many ecosystem components and their interactive responses to management practices. Simulation models are ideally suited to such tasks. However, difficulty of use severely restricts their application. Identifying model interpreting output information are among the factors that make models difficult to use. This paper discusses the development of a user to model interface that is user friendly and greatly simplifies the selection and operation of simulation models for management purposes. A prototype blackboard interface system is successfully demonstrated for simplifying the operation and application of SHAW, a complex model that can be used to simulate soil water and temperature regimes as they relate to the soil microenvironment associated plant and insect dynamics. Simplifying the use of simulation models will enhance their utility as decision making tools with significant benefits to natural resource management. Results of this research indicate that blackboard interfacing of complex simulation models by helping users with such processes as identifying model applications, selecting and organizing model inputs and interpreting outputs.
Technical Abstract: The most difficult tasks that natural resource managers must perform when using simulation models are determining if a model fits their needs, parameterizing inputs, explaining model functions and interpreting model outputs and simulation results. A systematic study of task activities performed by users interfacing with simulation models was conducted by applying a cognitive user-activity model. Because of the cognitive impacts of the user interface for complex simulation models, human engineering of the interface task activities is dominated by cognitive rather than perceptual factors. In this study, a prototype Blackboard Interfacing System (BIS) was developed to investigate the potential of BIS for enhancing user interfaces for simulation models. The user interface of this prototype is graphic and mouse driven with a built in user-cognitive model. Artificial intelligence (AI) techniques were applied to simplify/refine the whole task process that users need to perform when using simulation models. The blackboard approach allowed the integration of many pieces of interface knowledge and the use of the knowledge representation and reasoning strategy selected by users. The BIS system is capable of enhancing the user interface for simulation models.