Cluster-based user abnormal behavior detection method
A detection method and anomaly detection technology, applied in data processing applications, other database retrieval, digital data information retrieval, etc., to achieve the effects of improved accuracy, strong implementability, and simple structure
Active Publication Date: 2021-05-25
万商云集(成都)科技股份有限公司
View PDF19 Cites 4 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0004] However, some people or groups use the loopholes of the event to defraud a large amount of cash red envelopes. Such people will form a group and control multiple accounts by one person, analyze the reward rules of the event merchants and find out their loopholes, and bring many new customers. The illusion of traffic and traffic, and then defraud the event organizers of a large amount of cash red envelopes and other discounts
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment 1
[0032] Such as figure 1 As shown, a cluster-based user abnormal behavior detection method includes the following steps:
[0033] S1. Regularly update and obtain the user behavior data in the database of the business data platform, and make statistics on the user behavior data to obtain the statistics of all the behaviors performed by each user and all the behaviors of other people brought by the user as the inviter. Data; user behavior data includes the unique ID and username of the customer who initiated the action, the unique ID and username of the customer who invited the action, the type and duration of the action;
[0034] S2. Abstract the statistical user behavior data into a graph to obtain a user relationship graph; in the user relationship graph, the directed edges of the graph represent the user's behavior, the source node of the directed edge is the user who initiated the behavior, and the directed edge's purpose The node is the inviter of the behavior, and the weigh...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention provides a cluster-based user abnormal behavior detection method, which comprises the following steps: S1, regularly updating and obtaining user behavior data in a merchant data platform database and carrying out statistics, S2, abstracting the counted user behavior data into a graph to obtain a user relation graph, and S3, inputting the user relation graph into a community division algorithm, performing community division to obtain a plurality of clusters, and performing feature extraction to obtain vectors of user features; S4, inputting the vectors of the cluster and the user characteristics into an anomaly detection model for anomaly detection judgment to obtain an anomaly detection result, wherein the anomaly detection model comprises a plurality of detection channels, each detection channel comprises a plurality of detection layers, each detection channel has different detection algorithms, and the detection algorithms comprise a G-KNN detection algorithm and a GN-LOF detection algorithm; according to the invention, single-point detection is improved into cluster detection, the abnormal behavior characteristics of the user are described by using the cluster characteristics, and the detection efficiency is high.
Description
technical field [0001] The invention relates to the technical field of computer information security, in particular to a method for detecting abnormal behavior of users based on clusters. Background technique [0002] While the network brings convenience to users, it also brings the risk of malicious intrusion. In terms of abnormal user behavior detection based on commercial software, there are actual cases where users use software rules to exploit wool. Utilizing the influence of offline friends’ network, for example, the effect of pushing products or event advertisements in WeChat groups and circles of friends is far greater than the advertisements pushed by the platform. More and more merchants use the personal connections between customers to promote own products. In this process, in order to more effectively motivate customers to help promote products and activities, the merchants who hold the event will give cash red envelopes and coupons to motivate according to the ...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): G06Q30/00G06F16/901G06F16/9035G06F16/904
CPCG06Q30/0185G06F16/9024G06F16/9035G06F16/904Y02D10/00
Inventor 田文洪
Owner 万商云集(成都)科技股份有限公司