Analysis of key indicators based on large-scale drug testing data

The safeguard of drug security is an important prerequisite to guarantee public health and life safety. The pre-warning mechanism is an important means to monitor drug risks quickly and effectively. Based on the data mining technology, oriented to the massive data from each procedure of the drug industry chain, this paper mainly discusses how to pre-find the potential risks of the whole drug industry chain, and provides early-warning guidance via risk indicators. From the view of data modeling and technical processing, the key is how to integrate and analyze multi-source heterogeneous data (across space and time) and use the analysis of key indicators for streaming data. In specific, we first construct a distributed system and carry standardization for complex time-series data from different objects. Secondly, the potential risk of drug data is evaluated by using the key indicator analysis. At last, obtain the abnormal data objects and their fluctuating indicators of different periods. The application of the proposed method can provide potential hints or guidance for the sampling inspection of drugs, which is advantageous to improve the efficiency of the drug safety supervision department.