System log anomaly detection method and system based on deep learning technology

The invention relates to the technical field of artificial intelligence, and discloses a system log anomaly detection method and system based on a deep learning technology, and the method comprises the following specific steps: S1, carrying out the template analysis of a training log data set, and obtaining structured log template data; s2, constructing a vectorization model, and inputting the structured log template data into the vectorization model for unsupervised contrast learning training; s3, generating sub-sequence sample data by extracting and associating sessions and events in the data of the training log data set; s4, completing the modeling of a subsequence correlation model according to the vectorized data and the subsequence sample data; s5, training a subsequence correlation model; and S6, detecting log abnormity through the trained sub-sequence correlation model. The problem that the relevance between the log data is understood insufficiently in the prior art is solved, and the method has the a