Unlock instant, AI-driven research and patent intelligence for your innovation.

Metho and systems for performing implicit case mining

A hidden case and case technology, applied in the field of hidden case mining and systems, can solve problems such as lack, insufficient policy and model coverage, and achieve the effect of improving coverage.

Pending Publication Date: 2020-02-11
ALIPAY (HANGZHOU) INFORMATION TECH CO LTD
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

In some security scenarios (such as misappropriation and fraud), risk cases usually come from customer complaints, but in some risk scenarios (such as illegal cashing), customers usually do not come to complain, and some cases are only a small amount. Therefore, the number and diversity of black samples will be insufficient. Correspondingly, the coverage of strategies and models will also be insufficient. Therefore, the implementation of hidden case mining is particularly important for the active prevention and control of risks.
[0005] However, there is a lack of solutions in the prior art that can perform hidden case mining

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 more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Metho and systems for performing implicit case mining
  • Metho and systems for performing implicit case mining
  • Metho and systems for performing implicit case mining

Examples

Experimental program
Comparison scheme
Effect test

example 1

[0092] see figure 2 , which shows a flowchart of a method 200 for selecting candidate clusters according to Example 1.

[0093] The method 200 may include: at step 202, determining a preferred variable combination that embodies common characteristics of the hidden cases.

[0094] Specifically, determining the preferred variable combination can be implemented in the following manner.

[0095] First, the known black sample concentration for each cluster may be compared to a threshold concentration to identify one or more high black sample concentration clusters with known black sample concentrations greater than the threshold concentration. The threshold concentration may be set by the developer. For example, developers can choose the optimal threshold concentration empirically or through experiments. For example, the threshold concentration may be chosen to be 90%. Exceeding the threshold concentration indicates that the ratio of the number of known black samples in the cl...

example 2

[0107] see image 3 , which shows a flowchart of a method 300 for selecting candidate clusters according to Example 2.

[0108] In Example 2, similar to Example 1, in step 302, the known black sample concentration of each cluster can be compared with a threshold concentration, so as to determine one or more high black samples whose known black sample concentration is greater than the threshold concentration. Sample concentration clusters. Again, this threshold concentration can be set by the developer. For example, developers can choose the optimal threshold concentration empirically or through experiments. For example, the threshold concentration may be chosen to be 90%. Exceeding the threshold concentration indicates that the ratio of the number of known black samples in the cluster to the total number of samples is greater than 90%.

[0109] At step 304, the number of unknown samples in the one or more high black sample concentration clusters may be counted. For exampl...

example 3

[0115] see Figure 4 , which shows a flowchart of a method 400 for selecting candidate clusters according to Example 3.

[0116] In this example, different from Example 1 and Example 2, after the known black sample concentration is obtained, the known black sample concentration may not be compared with the threshold concentration, but in step 402, the to low to sort the multiple clusters. The multiple clusters are clusters obtained by traversing all variable combinations as described above, for example, the x clusters mentioned above.

[0117] Subsequently, in step 404, multiple clusters with the highest ranking may be selected as candidate clusters. For example, the top 3 clusters of known black sample concentrations can be used as candidate clusters.

[0118] It can be seen that, in this example, instead of the number of unknown samples, the concentration of known black samples is considered. It can be appreciated that the higher the concentration of known black samples ...

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

No PUM Login to View More

Abstract

The invention discloses a method for executing implicit case mining, which comprises the steps of obtaining a sample set, with the sample set comprising known black samples and unknown samples, and the known black samples being samples which have been determined as risk cases; obtaining a variable pool, wherein the variable pool comprises a plurality of variables; performing a clustering algorithmon the sample set using each combination of the plurality of variables to obtain a plurality of clusters; calculating the concentration of a known black sample in each of the plurality of clusters; selecting one or more candidate clusters based on the concentration of the known black sample of each cluster in the plurality of clusters; and identifying unknown samples in the one or more candidateclusters as potential implicit cases. The invention further relates to a corresponding system and a computer readable storage medium.The coverage rate, accuracy and applicability of implicit case mining can be improved.

Description

technical field [0001] One or more embodiments of the present specification relate to methods and systems for performing covert case mining. Background technique [0002] With the popularity of online transaction systems or payment systems, the security issues associated with them have also become common, and the identification, prevention and control of risk cases has become increasingly important. Risk cases include, for example, hacking, fraud, and illegal cashing out. [0003] In current online transaction systems or payment systems, strategies for identifying these risk cases already exist. Through these strategies, many risk cases were identified. In addition, through customer complaints and other methods, some risk cases can also be identified. [0004] However, there are still some hidden cases. Hidden cases can be defined as risky cases that actually exist but are missed by policies and are not complained about, or risk cases that are blocked by policies and do ...

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
no application Login to View More
IPC IPC(8): G06Q20/40G06K9/62G06Q40/04
CPCG06Q20/4016G06Q40/04G06F18/2321
Inventor 陈志招
Owner ALIPAY (HANGZHOU) INFORMATION TECH CO LTD