Campus network fraud early warning method and system based on DPI, and readable medium

The invention relates to a DPI-based campus network fraud early warning method and system, and a readable medium, and the early warning method comprises the steps: regularly collecting a campus deep packet inspection DPI log, extracting a domain name field in the DPI log, and carrying out the comparison and filtering processing, and obtaining a to-be-analyzed domain name list; carrying out website content collection and fraud-related analysis on the basis of a website corresponding to the to-be-analyzed domain name list, judging whether the website is a fraud-related website or not, and updating a domain name corresponding to the fraud-related website to a fraud-related domain name library; collecting a campus DPI log in real time, extracting a domain name field in the DPI log to perform collision analysis with the fraud-related domain name library, and outputting a risk log; performing double analysis on the risk logs, removing logs corresponding to non-fraud-related websites, and outputting a fraud-related