Developing Backward Chaining Algorithm of Inference Engine in Ternary Grid Expert System

The inference engine is one of main components of expert system that influences the performance of expert system. The task of inference engine is to give answers and reasons to users by inference the knowledge of expert system. Since the idea of ternary grid issued in 2004, there is only several developed method, technique or engine working on ternary grid knowledge model. The in 2010 developed inference engine is less efficient because it works based on iterative process. The in 2011 developed inference engine works statically and quite expensive to compute. In order to improve the previous inference methods, a new inference engine has been developed. It works based on backward chaining process in ternary grid expert system. This paper describes the development of inference engine of expert system that can work in ternary grid knowledge model. The strategy to inference knowledge uses backward chaining with recursive process. The design result is implemented in the form of software. The result of experiment shows that the inference process works properly, dynamically and more efficient to compute in comparison to the previous developed methods.