Data leakage prevention analysis method and system based on user behaviors and document contents

Abstract

The invention discloses a data leakage prevention analysis method and system based on user behaviors and document contents. The method includes the following steps: acquiring related data of mail sending behaviors of a user in a scheduled long time period and a scheduled short time period respectively, performing data averaging and normalization, and acquiring a long-term behavior data vector anda short-term behavior data vector of the user respectively; according to a comparison result between the vector distance between the long-term behavior data vector and the short-term behavior data vector of the user and a predetermined vector distance threshold, determining whether the mail sending behaviors of the user have abnormalities; for mails that are sent by the user and have abnormal behaviors, extracting mail content documents, and determining the subject categories of the documents; and selecting text accurate matching strategy rules associated with the categories according to the subject categories of the documents, and determining whether sensitive data exists in the documents. According to the technical scheme of the invention, the accuracy degree of the judgment of sensitivedata leakage events can be significantly improved, and the false alarm rate of performing judgment only through content matching can be effectively reduced.

Description

Technical field [0001] The invention relates to the field of data security, in particular to a data leakage prevention analysis method and system based on user behavior and document content. Background technique [0002] The main function of the enterprise data leakage prevention system is to prevent enterprise employees from sending out sensitive data. Therefore, accurately determining whether the data sent by employees is sensitive is the key to the data leakage prevention system. The traditional method is achieved by means of exact matching, such as the number of hits of keywords or regular expressions, which are often prone to many false positives. Therefore, the data leakage prevention system urgently needs to consider more factors to determine whether the outgoing data behavior of enterprise employees is a security incident. [0003] Comparative Literature 1 [0004] Publication No. 105357217A, Title of Invention: Data theft risk assessment method and system based on user be...

AI Technical Summary

Problems solved by technology

Abnormal operation behavior involves multiple factors such as the operator's emotions and temporary changes in work. Therefore, if it is not considered in combination with other factors, the practicality must not be good.

Images