Risk behavior recognition method, storage medium, device and system

Abstract

The invention discloses a risk behavior recognition method, a storage medium, a device and a system, and relates to the field of big data risk control. The method comprises the steps of extracting, according to behavior log data, each specific behavior of each user in a set user set, and obtaining a specific behavior sequence for each user; using pre-defined strong rules to find high-risk suspected users in the set user set, recording the suspected users as a first user set, and recording the remaining users in the set user set as a second user set; setting specific behavior pairs, and calculating a first transition probability and a second transition probability; for the same specific behavior pair, judging whether the difference between the first transition probability and the second transition probability is within a preset range, and if not, recording the specific behavior pair as a risk specific behavior pair; and calculating the information entropy of each risk specific behaviorpair group, and judging whether the user is a suspected user according to the size of the information entropy. The risk behavior recognition method in the invention can find malicious users whose cheating characteristics are not obvious.

Description

technical field [0001] The invention relates to the field of big data risk control, in particular to a risk behavior identification method, storage medium, equipment and system. Background technique [0002] There are some malicious users on the live broadcast platform. These users usually do some behaviors of falsely following or gaining popularity. Finding such malicious users will help maintain the order of the live broadcast platform and ensure the long-term healthy development of the platform. [0003] For the identification of such malicious users, the usual method is to use strong rules, that is, for example (1) one account is logged in on multiple devices; (2) the user's behavior log is abnormal, such as the user sent a barrage but did not click on the barrage button etc. However, strong rules can only find out those malicious users with obvious characteristics, and can do nothing for other users with less obvious characteristics. Moreover, the cheating methods of ...

AI Technical Summary

Problems solved by technology

However, strong rules can only find out those malicious users with obvious characteristics, and can do nothing for other users with less obvious characteristics

Moreover, the cheating methods of malicious users have become more complicated at present. By disguising the behavior of malicious users and normal users, the behavior of malicious users is becoming more and more similar to that of normal users, and the strong rule recognition effect is not good.

Images