Abnormal behavior detection method based on user behavior analysis

Abstract

The invention provides an abnormal behavior detection method based on user behavior analysis. User behavior data are collected by a sliding time window that can be fed back and adjusted, a multi-angle behavior library and a multi-angle behavior analysis model are constructed to analyze the behavior data of a user, the analysis result data group is analyzed and compared with a set acceptable confidence value, and the parameter of the time window of data collection can be fed back and adjusted and the behavior library can be updated according to an analysis and comparison result so as to obtain realize comprehensive analysis and accurate judgment, the analysis result of the user behavior can be obtained by computational complexity as small as possible on the premise of improving the detection accuracy.

Description

technical field [0001] The invention relates to the field of information technology, in particular to a method for detecting abnormal behavior based on user behavior analysis. Background technique [0002] With the increase in the number of Internet users and the diversity of types, it is gradually becoming more and more difficult to analyze user behavior to track potential problems or detect malicious users and malicious behaviors, especially the massive user data. The processing efficiency puts forward higher requirements, making the traditional user data storage and analysis methods no longer competent. With the advent of the era of big data, big data analysis also came into being. Big data analysis refers to the analysis of huge data. Big data analysis can intuitively display data based on data visualization, and data mining allows us to dig deep into the data to mine value. [0003] Malicious behaviors of users are constantly evolving and mutating, and as network tra...

AI Technical Summary

Problems solved by technology

At the same time, the model should have a high accuracy rate, and an excessively high false alarm rate will reduce the user experience and the credibility of the system

[0005] The purpose of the present invention is to solve the problem of serious degradation of the existing machine learning model and the low accuracy of the existing detection method. It proposes a dynamic sliding window that can be adjusted by feedback and statistically analyzes the user's behavior in real time. Dynamic tracking The user's behavior habits, based on the known behavior and the acceptable maximum error probability given by the user, the method of detecting malicious behavior from multiple angles through the analysis module

Images