System and method of training behavior labeling model

A system of training behavior labeling model, comprising:

• an input unit, receiving a labeled data set, wherein the labeled data set comprises a training data set and a verification data set, and each data of the training data set and each data of the verification data set respectively comprise first labeling information;
• a storage unit, storing a plurality of learning modules;
• a processing unit, connected to the input unit and the storage unit and respectively inputting each data of the training data set to the learning modules to establish a plurality of labeling models based on different supervised training algorithms,
• wherein the processing unit further obtains a plurality of second labeling information corresponding to each data of the verification data set respectively according to the labeling models and respectively generates a behavior labeling result corresponding to each data of the verification data set according to the second labeling information corresponding to each data of the verification data set,
• wherein the processing unit further obtains a labeling change value according to the behavior labeling result and the first labeling information corresponding to each data of the verification data set and determines whether the labeling change value is greater than a change threshold, and in response to that the labeling change value is greater than the change threshold, the processing unit updates the first labeling information corresponding to each data of the verification data set according to the behavior labeling results, exchanges the training data set and the verification data set and reestablishes the labeling models according to the exchanged training data set.