CTC system operation log data-oriented analysis method

The invention provides a CTC system-oriented operation log data analysis method, which comprises the following steps: S1, carrying out structured processing on text information of operation log data through a Bert-BiLSTM-CRF model to obtain structured operation log data; s2, automatically generating an incremental log template based on an improved FT-tree algorithm, and performing characteristic frequency statistics, operation sequence generation and association collection on the structured operation log data to form log sequence data; s3, performing abnormal high-risk operation identification and statistics on the log sequence data based on an operation log abnormity/high-risk operation detection model of an improved unsupervised learning K-Means algorithm, and generating an operation log abnormity and high-risk operation identification result; and S4, constructing an abnormal high-risk operation and risk correlation analysis model based on a Bayesian network model, and carrying out correlation analysis on o