Automatic analysis method and system for inspection data

Abstract

The invention provides an automatic analysis method and system for inspection data, and the method comprises the steps: obtaining an inspection record text processing neural network through obtaining an inspection record text which is recorded in an inspection terminal in the inspection process of a machine room and is used for carrying out the quasi-anomaly recognition; and carrying out abnormal record identification on the inspection record text to be subjected to abnormal identification based on the inspection record text processing neural network to obtain an abnormal record identification result, and when the abnormal record identification result indicates that an abnormal record is identified, carrying out display marking on the identified abnormal record in the inspection terminal. According to the method, transfer learning is carried out based on the determined multiple target abnormal record recognition results, the transfer effect is better, meanwhile, relevance information among contents in the same inspection record text is measured, the relevance information is fused with information based on representation information and information based on output, and the accuracy of the transfer learning is improved. The recognition accuracy of the guided neural network is higher than that of the guiding neural network, and the abnormal record recognition effect of the inspection record text processing neural network is improved.

Images