The microseismic monitoring technology has been widely used on the analysis of rock stability and management of mine safety production owing to its ability in catching the response information of rock mass under mining disturbance.Due to the impact of frequent production activities at the mine site, the microseismic monitoring system would catch so many different types of signals including lots of noise that the response law of rock mass under mining disturbance could not be revealed timely and effectively.This dissertation has analyzed the differences in the characteristics of typical signal waveform parameters according to the microseismic monitoring system built in Ashele Cooper Mine.Moreover, an identification method of rock mass rupture signal based on decision tree classification algorithm has been proposed, and made a comparative analysis of its recognition accuracy in this thesis.Besides the rock rupture signals, the other noise signals mainly contains electrical noise signals, mechanical vibration signals and blasting signals.The results indicate that these signals have different degrees of coincidence in the distribution range of parameters such as duration, rise time, ringing number, rising ringing number, maximum amplitude, and main frequency.Therefore, it is impossible to use a single parameter to effectively identify the rock mass rupture signal and eliminate the influence of noise signals.Additionally, the model basing on the decision tree classification algorithm performs better than SVM in terms of identifying the rock rupture signals, and its recognition accuracy rate has reached 97.8%.The results are of great significance for quickly delineating and early warning of high-risk areas of rock mass damage.