The skilled machinists evaluate exactly the cutting state. Therefore, the study on the modeling machinist's thinking process is progressed recently. In this paper, a rational way for the system monitoring the cutting state in turning is discussed. The proposed system consists of the fuzzy integral model which evaluates the individual cutting trouble (e.g. tool life, chatter, chip controllability) and the neural network which applies to the model depending on machining conditions. The fuzzy integral model is systematized by the machinist's evaluation to individual trouble. The neural network is structured by the measure of information that compared the evaluation of machinist with fuzzy integral model. The validity of this system was confirmed by the experiments under various cutting conditions in this study.