We designed a novel individualization process for improving personal emotion recognitions in real time. The designed individualization process was performed by a neutralization algorithm of physiological signals, a subjective emotion reflection of a user updated by personal emotion rules in real time. The physiological signals such as PPG(Photoplethysmography), GSR(Galvanic skin reflex), and SKT(Skin temperature) were measured and analyzed to estimate an emotion states of users. Regulating the emotion status using by emotion rules was performed by reflecting subjective evaluations. The agreement of emotion recognition between of individualization and non-individualization method was estimated by 10 undergraduates (5 females, mean age: 22.1 ± .2) of Sangmyung University. During the emotion recognition test, 45 images were randomly presented to each participant five times. In results, the proposed individualization process showed the agreement of 71.67 % which was five times higher than when the process was not applied. Therefore, in this study, we demonstrated that the individualization process was significantly useful for customizing emotion recognitions of personal users in real time. The individualization process will be able to improve satisfactions in various emotion related applications and services in the nearer future.