Cross-platform malicious code detection method and system

The invention relates to a cross-platform malicious code detection method and system. The method comprises the steps of (1) training a pre-training model through employing benign program samples on a plurality of platforms, so as to capture the structure and semantic correlation in the context of a program instruction, and the structure and semantic generality between program instructions of different platforms, (2) on the pre-training model, constructing a cross-platform malicious code detection model by using a plurality of benign program samples and malicious program samples of a limited scale of platforms, performing parameter fine tuning on the cross-platform malicious code detection model, and migrating knowledge in the pre-training model into the cross-platform malicious code detection model, and (3) detecting unknown program samples on different platforms by using the constructed cross-platform malicious code detection model, and judging whether the unknown program samples are malicious or benign. Pro