A Method for Identifying Telecom Fraud Risks in Large-Scale Data Streams
A technology of risk identification and data flow, applied in telephone communication, electrical components, wireless communication, etc., can solve problems such as telephone fraud with inconspicuous concealment features
Active Publication Date: 2021-06-15
NAT COMP NETWORK & INFORMATION SECURITY MANAGEMENT CENT
View PDF6 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0005] Among the existing telecommunications fraud detection methods, technologies such as voice analysis can only identify fraudulent behaviors with relatively obvious fraudulent characteristics. here comes the big challenge
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 Construction
[0030] The present invention will be further described in detail below in conjunction with accompanying drawings and examples.
[0031] The invention discloses a telecommunications fraud risk identification method in a large-scale data flow, based on heterogeneous calling and called user behavior information, to realize high-risk telecommunications called user risk modeling. Specifically, the method includes: first, initially screening the original called set: screening out a certain number of high-risk calling and called user records from the original call record database according to the rules; and constructing a fraudulent called index database; then, preprocess the call record data of high-risk called parties, and extract salient features as the input of the subsequent risk model; adopt the two-level cascade classification model, according to the call record data and historical feature data of the calling end, the The maximum value of the attack risk among the L callers wh...
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 discloses a method for identifying telecommunication fraud risks in large-scale data streams, and belongs to the fields of data mining, machine learning, business intelligence and the like. Screen high-risk called users and calling users from the call record database as analysis objects, build a fraudulent called index database, and extract the salient features corresponding to each index; use a two-level cascade classification model to obtain the attack of each called number Risk value; retain the maximum value of the latest sliding window abnormal score of each high-risk called number as the respective sequence risk value; construct a bipartite graph of the communication relationship between the called number and the calling number, and calculate the value of each called number Contact risk value: With the help of the logistic regression model, the attack risk, sequence risk and contact risk of each called number are fused to obtain the respective comprehensive risk value of each called number. The final comprehensive risk value of the present invention has higher stability and interpretability, and realizes higher classification and detection efficiency.
Description
technical field [0001] The invention relates to a method for identifying telecommunication fraud risks in large-scale data streams, and belongs to the fields of data mining, machine learning, business intelligence and the like. Background technique [0002] Telecommunications fraud detection is an important problem in the domestic and foreign telecommunications industry. It threatens people's property safety and brings huge troubles to the normal operation of telecommunications. Therefore, operators and government departments are trying to adopt various means and technologies to detect and identify telecommunications fraud. [0003] How to use the current advanced data mining technology to assist the regulatory authorities to identify fraudulent transactions and identify potential risks has become an important challenge at present, which involves background knowledge such as classification and clustering in the field of data mining and anomaly detection. In practical proble...
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 Patents(China)
IPC IPC(8): H04W12/12H04M3/22
CPCH04M3/2218H04W12/12H04M2203/6027
Inventor 万辛林浩刘冠男吴孟其安茂波李鹏高圣翔黄远林格平
Owner NAT COMP NETWORK & INFORMATION SECURITY MANAGEMENT CENT