Intelligent Knowledge-Based Near-Real-Time Materials Accountancy

An intelligent knowledge-based systems approach to the analysis of near-real time accountancy data is proposed. One of the keys to building a sucessful system is in representing knowledge in an appropriate manner, another is in the merging of the information from statistical tests with that obtained from experience. A multi-layered inference mechanism incorporating the statistical test, a plant simulation and a number of heuristic searches are recommended.