Operator performance measures are used for multiple purposes, such as control room design, human system interface (HSI) evaluation, training, and so on. Performance measures are often focused on results; however, especially for a training purpose - at least in a nuclear industry, more detailed descriptions about processes are required. Situation awareness (SA) measurements have directly/indirectly played as a complimentary measure and provided descriptive insights on how to improve performance of operators for the next training. Unfortunately, most of the well-developed SA measurement techniques, such as Situation Awareness Global Assessment Technique (SAGAT) need an expert opinion which sometimes troubles easy spread of measurement's application or usage. A quantitative SA measurement tool named Computational Representation of Situation Awareness with Graphical Expressions (CoRSAGE) is introduced to resolve some of these concerns. CoRSAGE is based on production rules to represent a human operator's cognitive process of problem solving, and Bayesian inference to quantify it. Petri Net concept is also used for graphical expressions of SA flow. Three components - inference transition, volatile/non-volatile memory tokens - were newly developed to achieve required functions. Training data of a Loss of Coolant Accident (LOCA) scenario for an emergency condition and an earthquake scenario for an abnormal condition by real plant operators were used to validate the tool. The validation result showed that CoRSAGE performed a reasonable match to other performance results.