CN110046910B - Method and equipment for judging the legitimacy of transactions conducted by customers through electronic payment platforms - Google Patents

Abstract

Provided is a method of obtaining a customer group associated with a particular client, comprising obtaining said particular client among a plurality of clients; obtaining client-related data associated with each of the plurality of clients, said client-related data comprising at least Customer relationship data of the relationship between each customer and other customers in the plurality of customers and customer characteristic data of the customer; obtaining predefined extended rule data; and based on the customer-related data and the extended rule The data determines one or more customers related to the specific customer among the plurality of customers, so as to obtain a customer group related to the specific customer. In this way, on the basis of the identified individual suspicious customers, relevant customer groups can be further obtained, thereby increasing the trial efficiency while expanding the coverage of trial clients.

Description

Method and equipment for judging the legitimacy of transactions conducted by customers through electronic payment platforms

technical field

The invention relates to the technical field of the Internet, and in particular to identifying one or more customers associated with a specific customer among multiple customers.

Background technique

With the continuous development of Internet technology, it is applied to various fields, and new technical fields such as Internet finance have been derived. Electronic payment platforms such as Alipay rely on Internet technology to realize fund payment and transfer, which greatly facilitates people's lives.

However, while providing convenience for people's lives, these electronic payment platforms also have hidden dangers. For example, some customers may wish to use the electronic payment platform to achieve certain transactions with illegal purposes. Therefore, how to judge the legitimacy of the transactions conducted by customers through the electronic payment platform is an important issue.

At present, through screening all customers of the payment platform, identifying individual suspicious customers, and then artificially examining these suspicious customers one by one to identify customers with illegal transactions.

Contents of the invention

It is expected to provide a method that can further obtain relevant customer groups on the basis of identified individual suspicious customers, thereby expanding customer coverage and improving trial efficiency.

According to one embodiment, there is provided a method of acquiring a customer group related to a specific customer, comprising acquiring data representing the specific customer among a plurality of customers; acquiring customer-related data related to each customer in the plurality of customers, The customer-related data at least includes customer relationship data representing the relationship between each customer in the plurality of customers and other customers and customer characteristic data of the customer; acquiring predefined extended rule data; and based on the The customer-related data and the extended rule data determine one or more customers related to the specific customer among the plurality of customers, so as to obtain a customer group related to the specific customer.

The current individual suspicious customer identification and trial methods for customers who use electronic payment platforms for transactions do not take into account the relationship between the identified individual suspicious customers and other customers, and only trial individual suspicious customers, which makes the number of trial samples limited. In fact, customers who conduct transactions are not isolated, and they often have certain correlations, especially when conducting transactions for certain illegal purposes, such as money laundering, there is a strong relationship between funds and funds among various customers. /or non-financial relationship. The inventors have recognized the strong correlation between such customers, and with the help of various embodiments of the present invention, on the basis of the obtained individual suspicious customers, according to both the customer-related data and the predefined expansion rules, the strong correlation with them is acquired other customers, push the obtained individual suspicious customers and customers with strong correlation to the examiner as a whole, so that it is possible to conduct a trial on such a strong correlation customer group during the trial process, which can not only improve The efficiency of the trial is also easy to obtain a complete chain of evidence in the trial. Moreover, with the help of this group trial method, it can also expand the coverage of trial clients and increase the defense radius of illegal transaction trials.

According to a further embodiment, determining another one or more customers of the plurality of customers related to each of the one or more customers based on the customer-related data and the expanded rule data; and The other one or more customers are included in the customer group related to the specific customer.

Thus, on the basis of one or more customers in the customer group related to a specific customer acquired for the first time, further customers with strong correlation with the one or more customers can be obtained, and the further acquired customers can also be added Among the customer groups related to the specific customer, the coverage of customers to be tried is further expanded, and the defense radius of the illegal transaction trial is increased. Multiple layers of such extensions can be made if desired.

According to a further embodiment, each customer in the customer group is scored based on the obtained customer-related data of each customer in the customer group related to the specific customer using a machine learning-based scoring model; and, based on The scoring is used to sort the customers in the customer group.

Therefore, using the scoring model based on machine learning to score and then sort can increase the accuracy of scoring, and facilitate the distinction and sorting of customers among the acquired customer groups that have strong correlations with specific customers. Customers with a high possibility of illegal transactions are ranked first.

According to a further embodiment, characteristic data representative of a plurality of customer characteristics of a set of reference customers is determined based on customer-related data for each of the reference customers; and, using the characteristic data of the reference customers to train the Scoring models for machine learning.

In this way, it is possible to carry out targeted training on the scoring model based on machine learning before ranking the scores.

According to a further embodiment, the characteristic data representing a plurality of customer characteristics of said customer is determined according to the customer-related data of each customer in said customer group; The feature data of the feature scores each customer in the customer group; and based on the score of each customer, ranks the customers in the customer group. This provides a specific example of scoring using the scoring model trained above.

According to a further embodiment, the extended rule data includes data representing a customer group related to the specific customer obtained by performing data mining on the multiple customers based on the customer-related data of each customer in the multiple customers. rule data.

According to a further embodiment, the extended rule data also includes predetermined rule data representing the relevance of each customer in the customer group to the specific customer; and/or predetermined relationship with the acquisition of the customer group purpose-related rule data.

By using the above-mentioned extended rule data of at least two aspects, it is possible to more comprehensively and accurately obtain customer groups related to specific customers.

According to a further embodiment, one or more customer groups among the plurality of customers are identified by matching a predefined template defining corresponding The relationship structure between various customers and/or the corresponding characteristic data of each customer; and the data representing the customer group related to the specific customer among the one or more customer groups is determined to represent the The rule data of the relevant customer groups.

According to a further embodiment, one or more customer groups are determined from the plurality of customers based on customer-related data of the plurality of customers, such that each of each of the one or more customer groups customers each have at least a predetermined number of related customers in the customer group; and determining data representing a customer group of the one or more customer groups related to the particular customer Rule data for customer groups.

The above provides two ways of acquiring the rule data representing the customer group related to the specific customer.

According to another embodiment, there is provided an apparatus for acquiring a customer group related to a particular customer, comprising a memory; and a processor configured to, when running program code from the memory, perform various methods according to the present invention The method described in the examples.

According to another embodiment, there is provided a machine-readable medium storing computer program code which, when executed, causes a computer or a processor to perform the method according to various embodiments of the present invention.

According to another embodiment, there is provided a device for acquiring a customer group related to a specific customer, comprising a first acquiring unit configured to acquire data representing the specific customer among a plurality of customers, and acquire data related to a plurality of customers Customer-related data related to each customer in , said customer-related data includes at least customer relationship data representing the relationship between each customer in said plurality of customers and other customers and customer characteristic data of said customers; 2. an acquisition unit configured to acquire predefined extended rule data; and a determination unit configured to determine, among the plurality of customers, those related to the specific customer based on the customer-related data and the expanded rule data One or more customers to get the customer groups associated with that particular customer.

Description of drawings

FIG. 1 shows a block diagram of a device for acquiring customer groups related to a specific customer according to one embodiment;

Fig. 2 shows a block diagram of a device for obtaining a customer group related to a specific customer according to another embodiment;

Fig. 3 shows a block diagram of a device for obtaining a customer group related to a specific customer according to yet another embodiment;

Fig. 4 shows a flow chart of a method for obtaining customer groups related to a specific customer according to an embodiment;

Fig. 5 shows a flowchart of a method for acquiring customer groups related to a specific customer according to another embodiment.

Various aspects and features of the invention are described with reference to the aforementioned figures. The same or similar reference numerals are generally used to refer to the same parts. The above-mentioned drawings are only schematic and non-limitative. Without departing from the gist of the present invention, the size, shape, label, or appearance of each element in the above-mentioned drawings may be changed, and are not limited to those shown in the accompanying drawings.

Detailed ways

The application of various embodiments of the present invention will be described below with reference to the application of illegal transaction review on customers of electronic payment platforms. It should be understood that the application of various embodiments of the present invention is not limited thereto, and it should be applicable to any On the basis of specific customers, the scope of customers is expanded to obtain customer groups related to specific customers. Therefore, the customers referred to below are not limited to customers who conduct transactions on the electronic payment platform.

Fig. 1 shows a block diagram of a device 10 for obtaining customer groups related to a specific customer according to one embodiment. The device 10 includes a first acquisition unit 11 , a second acquisition unit 12 , a determination unit 13 and an output unit 14 .

The first acquisition unit 11 acquires data representing a specific customer. Electronic payment platforms such as Alipay provide transaction customers with a review of illegal transactions. For example, Alipay currently monitors each transaction of platform transaction customers, and outputs a list of monitored suspicious customers within a predetermined period of time, such as every week, for review. to review. The first obtaining unit 11 can obtain the list of suspicious customers, and store these suspicious customers as specific customers. The particular customer may be one or more of a plurality of customers who transacted within a predetermined period of time.

In addition, the first acquisition unit 11 can also acquire customer-related data of multiple customers, for example, all customers who conduct transactions on the electronic payment platform within the predetermined time period. The above-mentioned specific customers are included in the plurality of customers. The customer-related data includes at least customer relationship data representing a relationship between each of the plurality of customers and other customers and customer characteristic data of each customer. Customer relationship data includes financial and non-financial relationships between any two customers. Funding relationship refers to transactions such as transfers between two customers within the predetermined time period, while non-financial relationship refers to any relationship other than the financial relationship between two customers, such as in the The same-device relationship between two customers within a predetermined period of time, such as a shared mac address, a shared mobile phone address book contact, etc. Customer characteristic data is characteristic data related to an individual customer, for example, the amount of inflow and outflow of funds of the customer within a predetermined period of time, the situation of the counterparty, etc. alerts or have ever been reported for illegal transactions.

The second acquiring unit 12 acquires predefined extension rule data. The extended rule data stipulates how to expand on the basis of the current specific customer to obtain the rules of the customer group related to the specific customer among the multiple customers who conduct transactions, and specifically specifies how to expand on the basis of specific customers Scale the number of clients that need to be vetted. In a preferred embodiment, the extended rule data may include two aspects of rule data. On the one hand, the extended rule data includes predetermined rule data representing a customer group related to a specific customer, which can be pre-determined by performing data mining based on customer-related data on payment platform transaction customers within a predetermined period of time; or can be Use any rule data previously identified that characterizes a customer segment that is relevant to a particular customer. In another aspect, the extended rule data may include rule data representing the relevance of each customer in the customer group to a particular customer and/or purpose of acquiring the customer group. Although it is preferable to use the two aspects of rule data to determine the customer group related to a specific customer, this is not limiting, and one of the rule data can also be used, or other rule data can be introduced if necessary.

Various methods can be used to data mine multiple transacting customers in order to predetermine regular data representing groups of customers related to particular customers. In one embodiment, one or more customer groups among the plurality of customers can be identified by matching a predefined template defining corresponding relationship structure and/or corresponding characteristic data of each customer. In a further embodiment, some suspected customer groups can be identified on the basis of a relationship graph between multiple customers using a template representing, for example, the relationship structure (such as the relationship between funds flow) among various customers of illegal transactions, and then Then, based on its characteristic data, it is further judged whether each customer in the identified customer group should belong to a member of the illegal transaction customer group.

In another embodiment, one or more customer groups can be determined from among the plurality of customers based on customer-related data of the plurality of customers who have transacted within, for example, a predetermined period of time, such that each of the one or more customer groups Each customer in the customer groups has at least a predetermined number of related customers in the customer group.

After one or more customer groups that may, for example, be involved in illegal transactions are determined from multiple customers through the above-mentioned different embodiments, the customer groups related to the specific customer in the one or more customer groups will be represented The data is determined as regular data representing a customer group related to the particular customer. The above-mentioned process of determining rule data representing a customer group related to a specific customer can also be performed by the second acquisition unit 12 on the basis of customer-related data of a plurality of customers conducting transactions. Of course, it is also conceivable that the above rule data is predetermined and only used in the device according to the embodiment of the present invention.

The determining unit 13 not only based on the rule data representing the customer group related to the specific customer but also based on the customer-related data of each customer in the multiple customers who conduct transactions, including customer relationship data and customer characteristic data, to determine the number of customers related to the customer. One or more customers related to a specific customer, so as to obtain the customer group related to the specific customer. The identified one or more customers are included in a customer group associated with the particular customer. For example, if the pre-determined rule data representing a customer group related to a specific customer indicates that a customer using the same mac address and whose outflow of funds within a predetermined period of time is greater than a certain threshold of the outflow of funds of the specific customer is related to the specific customer If you are a member of a customer group, then such a customer can be identified based on customer-related data for a plurality of customers with whom transactions are made.

In addition to the above-mentioned use of rule data representing customer groups related to specific customers to identify relevant customer groups, it is also possible to make further judgments on multiple customers who conduct transactions based on customer-related data to identify customers related to specific customers, for example , can utilize rule data representing the relevance of each customer in the customer group to a particular customer and/or purpose of acquiring the customer group.

Such rule data includes, for example, rule data related to the amount of funds, non-fund relationship, proportion of funds, whether an alert has been received, and whether it has been reported. The rule data involving the magnitude of funds can specify the relationship between the total amount of capital inflows and outflows of customers in the customer group to be determined by the customer and the total amount of fund inflows and outflows of the specific customer; Whether the non-financial relationship indicator between the customer and the specific customer is greater than a certain threshold; the rule data related to the proportion of funds can stipulate that the magnitude of the inflow or outflow of customers in the customer group to be determined accounts for the inflow or outflow of specific customers Whether the proportion of the magnitude is greater than a certain threshold; rule data concerning whether an alert has been received can specify that customers in the customer group to be determined have received alerts about, for example, illegal transactions within a predetermined period of time; rules concerning whether they have ever been reported The data stipulates whether a customer in the customer group to be determined has been reported, for example, for an illegal transaction within a predetermined period of time. The above-mentioned rule data involving the amount of funds, non-fund relationship, and capital ratio belong to the rule data that expresses the correlation with a specific customer, while the above-mentioned rule data related to whether an alert has been received and whether it has been reported belongs to the expression and acquisition of customer groups. The rule data for the relevance of the purpose. Rule data relating to other aspects is also contemplated. In addition, the above rule data can be combined arbitrarily to identify customers related to a specific customer, for example, customers who meet one or more of the above rules can be determined as customers related to the specific customer. Customers related to a specific customer identified from multiple customers using the rule data of the above two aspects can be determined as members of the customer group related to the specific customer.

After the determination unit 13 has determined the members of the customer groups related to the specific customer, the output unit 14 can output the specific customer and the determined customer groups related to it together, so that the examiner can conduct a review on these customers on this basis. Group trial, so as to improve the efficiency of the trial, and facilitate the acquisition of a complete chain of evidence during the trial.

On the other hand, the output unit 14 can also output the extended rules that each customer complies with while outputting the specific customer and the customers in the customer group related to the specific customer, so as to facilitate the trial by the reviewer. For example, if a customer is determined to be part of a customer group associated with a particular customer because of rule data related to the amount of money and whether or not he has received an alert, then for that customer, along with itself, output The above-mentioned rule data related to the amount of funds and whether an alert has been received, or output the representation related to the rule data, such as the relationship between the total amount of funds inflow and outflow of the customer and the total amount of fund inflow and outflow of a specific customer and the alarm data of the customer.

As mentioned above, in the determination unit 13, one or more customers related to the specific customer are determined in the determination unit 13, so as to obtain the customer group related to the specific customer, and the specific customer and its related information are directly output in the output unit 14. Customer groups. However, in order to further expand the coverage of trial customers, in one embodiment, the determination unit 13 can also further expand the relationship with each customer in the one or more customers based on the determined one or more customers. relevant customers to further increase the reach of the identified customer groups. Specifically, the determination unit 13 can determine, based on the customer-related data and the extended rule data of each customer among the multiple customers, another one or the other one or multiple customers, the other one or more customers are included in the customer group related to the specific customer for output by the output unit 14 . In this further extended embodiment, the extended rule data previously obtained by the second obtaining unit 12 can be used directly. It is also contemplated that each customer in the customer group obtained through the further expansion described above is scored and ranked as described below.

Fig. 2 shows a block diagram of a device 20 for obtaining customer groups related to a specific customer according to another embodiment. The main difference between the device 20 shown in FIG. 2 and the device 10 shown in FIG. 1 is that the device 10 shown in FIG. 2 further includes a scoring unit 15 and a sorting unit 16 . The scoring unit 15 receives the specific customer and the determined customer group related to the specific customer from the determination unit 13. In the case of multiple specific customers, in one embodiment, it can receive each specific customer and the determined The customer group related to the specific customer, that is, receiving the customer group corresponding to each specific customer. The scoring unit 15 can use the machine learning based scoring model M to score each customer in each customer group based on the customer-related data of each customer in the customer group related to the particular customer. The scoring model can be trained in advance by means of machine learning. An example of a usable scoring model is a scoring model based on the gradient boosting decision tree algorithm. The sorting unit 16 can sort the customers in each customer group based on the scores of each customer in the customer group. In this case, the output unit 14 outputs each customer group based on the ranking. It is also conceivable to include this particular customer in a group of customers related to it for scoring and ranking.

Fig. 3 shows a block diagram of a device 30 for acquiring a customer group related to a specific customer according to yet another embodiment. The difference between the device 30 shown in FIG. 3 and the device 20 shown in FIG. 2 is mainly that a training unit 17 is further included in the device shown in FIG. 3 . The training unit 17 can determine, according to the customer-related data of each reference customer in a group of reference customers, characteristic data representing a plurality of customer characteristics of the reference customer; and use the characteristic data of the reference customer to train the machine learning-based The scoring model M. The rating of this reference customer is known. The above characteristic data includes but not limited to characteristic data representing, for example, "inflow amount within a predetermined period", "relationship index with mac" and/or "whether it has been reported". It is contemplated that any number and kind of feature data may be employed. The above-mentioned training process can be implemented first before using the device according to the embodiment of the present invention.

In the case of using the scoring model M trained as described above, the determining unit 13 determines feature data representing a plurality of customer features of the customer based on the customer-related data of each customer in the determined customer group. The scoring unit 15 scores each customer in the determined customer group based on feature data representing a plurality of customer features of the customer using a machine learning based scoring model. The sorting unit 16 sorts the customers in the determined customer group based on the score of each customer.

The above described various embodiments of the present invention with reference to the embodiments shown in FIGS. Some of the features are deleted to obtain a new technical solution. For example, the training methods defined by the training unit 17 above can be replaced by other training methods known in the prior art.

The following describes a flow chart of a method 400 for acquiring a customer group related to a specific customer according to an embodiment of the present invention with reference to FIG. 4 .

At 401, data indicating a specific customer among multiple customers is obtained. In one embodiment, the data indicating the specific customer can be received from outside, and the specific customer can be determined in advance from multiple customers conducting transactions.

At 402, customer-related data related to each customer in a plurality of customers is obtained, the customer-related data at least includes customer relationship data representing the relationship between each customer in a plurality of customers and other customers and the customer's customer profile data.

At 403, the predefined expansion rule data is obtained. As mentioned above, in a preferred embodiment, the extended rule data may include two aspects of rule data. In one aspect, the extended rule data includes predetermined rule data representing customer groups related to a particular customer. In another aspect, the extended rule data may include rule data representing the relevance of each customer in the customer group to a particular customer and/or purpose of acquiring the customer group.

At 404, one or more customers among the plurality of customers related to the specific customer are determined based on the customer related data and the extended rule data, so as to obtain a customer group related to the specific customer. In a preferred embodiment, at 404, another one or more customers among the plurality of customers related to each customer in the one or more customers are further determined based on the customer-related data and the extended rule data, so The other one or more customers are included in the customer group related to the specific customer, thereby further expanding the customer group related to the specific customer.

At 405, the particular customer and its associated customer groups are output.

FIG. 5 shows a flow chart 500 of a method for obtaining a customer group related to a specific customer according to an embodiment of the present invention, wherein the processing at 501-503 is the same as the processing at 401-403 in the flow chart 400 shown in FIG. 4 same.

At 504, in addition to the same processing as the above 404, in one embodiment, according to the customer-related data of each customer in the determined customer group, feature data representing multiple customer features of the customer are determined.

At 505, each customer in each determined customer group is scored using a machine learning-based scoring model based on characteristic data representing a plurality of customer characteristics of the customer.

At 506, the customers in the customer group are sorted based on each customer's rating.

At 507, the customer group is entered based on the ranking.

Although various embodiments of the method according to the present invention have been described with reference to the flowcharts shown in FIGS. 4 and 5 . It can be understood that corresponding processes can be added/modified/deleted on the basis of the flow charts of the above embodiments, so as to form new technical solutions to achieve different effects.

In one embodiment, at 403, the predefined extended rule data can be obtained by: matching the predefined template with the customer-related data of multiple customers to identify one or more customer groups among the multiple customers , the predefined template defines the corresponding relationship structure between each customer and/or the corresponding feature data of each customer; the data representing a customer group related to a specific customer in one or more customer groups is determined as representing Rule data of customer groups related to said particular customer.

In another embodiment, at 403, the predefined extended rule data can be acquired in the following manner: based on the customer-related data of multiple customers, one or more customer groups are determined from multiple customers, so that one or more customer groups Each customer in each customer group in the customer group has at least a predetermined number of related customers in the customer group; Rule data for customer groups related to customers.

In another embodiment, each customer in the customer group can be scored at 505 using a machine learning-based scoring model based on customer-related data for each customer in the customer group associated with the particular customer; and at 506. Based on the scoring, sort the customers in the customer group.

In yet another embodiment, the scoring model can be trained before step 505 according to the above process is performed, specifically, according to the customer-related data of each reference customer in a set of reference customers, the number of points representing the reference customer is determined. feature data of a customer feature; and train the scoring model using the feature data of the reference customer.

It can be understood that the functions of each unit in the device for obtaining a customer group related to a specific customer and the flow of the method in each embodiment of the present invention can be implemented by computer programs/software. These software can be loaded into the working memory of the data processor and, when run, are used to perform the methods according to the various embodiments of the present invention.

Exemplary embodiments of the invention cover both creating/using the computer program/software of the invention from the outset and converting an existing program/software to use the computer program/software of the invention by means of an update.

According to a further embodiment of the present invention there is provided a machine (e.g. computer) readable medium, such as a CD-ROM, wherein said readable medium has stored thereon computer program code which, when executed, instructs A computer or processor executes methods according to various embodiments of the invention. The machine-readable medium is, for example, an optical storage medium or a solid-state medium supplied with or as part of other hardware.

The computer program for carrying out the methods according to embodiments of the invention may also be distributed in other forms, for example via the Internet or other wired or wireless telecommunication systems. The computer program may also be provided on a network such as the World Wide Web and can be downloaded from such a network into a working computer of the data processor.

It can also be understood that each unit in the device for acquiring a customer group related to a specific customer and the flow of the method in each embodiment of the present invention can also be implemented by hardware or a combination of hardware and software.

In one embodiment, a system for obtaining a customer group related to a particular customer can be implemented by a memory and a processor. The memory can store computer program codes for executing the method procedures according to the various embodiments of the present invention; when executing the program codes from the memory, the processor executes the procedures according to the various embodiments of the present invention.

It has to be pointed out that embodiments of the invention have been described with reference to different subject matters. In particular, some embodiments are described with reference to method type claims whereas other embodiments are described with reference to apparatus type claims. However, those skilled in the art will appreciate from the above and the following description that, unless otherwise specified, any combination of features related to different subject matter, in addition to any combination of features belonging to one type of subject matter, is also considered to be disclosed by the present application. up. Also, all features can be combined, providing a synergistic effect greater than the simple sum of the features.

The present invention has been described above with reference to specific embodiments. Those skilled in the art should understand that the technical solutions of the present invention can be implemented in various ways without departing from the spirit and basic features of the present invention. The specific embodiments are illustrative only, not limiting. In addition, these embodiments can be combined arbitrarily to achieve the object of the present invention. The protection scope of the present invention is defined by the appended claims.

The word "comprising" in the description and claims does not exclude the presence of other elements or steps. The functions of the various elements described in the specification or described in the claims can also be divided or combined, and realized by corresponding multiple elements or a single element.

Claims (20)

1. A method for judging the legitimacy of transactions conducted by customers through electronic payment platforms, including obtaining data representing an individual suspicious customer among a plurality of customers; obtaining customer-related data related to each of the plurality of customers, the customer-related data including at least customer relationship data representing relationships between each of the plurality of customers and other customers and the customer's customer profile data; Get predefined extension rule data; and Determine one or more customers related to the individual suspicious customer among the plurality of customers based on the customer-related data and the extended rule data, so as to obtain a customer group related to the individual suspicious customer, the customer group For the trial of illegal transactions. 2. The method of claim 1, further comprising determining an additional one or more customers of the plurality of customers related to each of the one or more customers based on the customer-related data and the extended rule data; and The additional one or more customers are included in a customer population associated with the individual suspect customer. 3. The method of claim 1, further comprising using a machine learning based scoring model to score each customer in the customer population based on the acquired customer-related data related to each customer in the customer population in relation to the individual suspicious customer; and Based on the scores, the customers in the customer group are ranked. 4. The method of claim 3, further comprising determining characteristic data representative of a plurality of customer characteristics of a set of reference customers based on customer-related data for each of the reference customers; and The machine learning-based scoring model is trained using characteristic data of the reference customer. 5. The method of claim 4, further comprising determining characteristic data representing a plurality of customer characteristics of said customer based on customer-related data of each customer in said customer group; scoring each customer in the customer population based on the characteristic data representing a plurality of customer characteristics of the customer using a machine learning based scoring model; and Based on the score of each customer, the customers in the customer group are sorted. 6. The method according to any one of claims 1-5, wherein said expanding rule data comprises performing data mining on said plurality of clients based on client-related data of each client in said plurality of clients The obtained rule data represents a customer group related to the individual suspicious customer. 7. The method of claim 6, wherein the extended rule data further includes Pre-determined rule data representing the relevance of each customer in said group of customers to said individual suspect customer; and/or Pre-determined rule data related to the purpose of acquiring the customer group. 8. The method of claim 6, further comprising identifying one or more customer groups among the plurality of customers by matching a predefined template defining relationships between corresponding respective customers with customer-related data of the plurality of customers structure and/or corresponding profile data of each customer; and Determining data representing a customer group related to the individual suspicious customer among the one or more customer groups as rule data representing a customer group related to the individual suspicious customer. 9. The method of claim 6, further comprising Determining one or more customer groups from the plurality of customers based on customer-related data of the plurality of customers, such that each customer in each of the one or more customer groups is in the customer group each have at least a predetermined number of related customers; Determining data representing a customer group related to the individual suspicious customer among the one or more customer groups as rule data representing a customer group related to the individual suspicious customer. 10. A device for judging the legitimacy of transactions conducted by customers through electronic payment platforms, including memory; and A processor configured to, when running program code from said memory, perform the method according to any one of claims 1-9. 11. A machine-readable medium storing computer program code which, when executed, causes a computer or processor to perform the method according to any one of claims 1-9. 12. A device for judging the legitimacy of transactions conducted by customers through electronic payment platforms, including A first acquisition unit configured to acquire data representing individual suspicious customers among a plurality of customers, and acquire customer-related data related to each of the plurality of customers, the customer-related data including at least customer relationship data of the relationship between each of the customers and other customers and customer characteristic data of said customers; a second obtaining unit configured to obtain predefined expansion rule data; and A determination unit configured to determine one or more customers related to the individual suspicious customer among the plurality of customers based on the customer-related data and the extended rule data, so as to obtain information related to the individual suspicious customer Customer groups, the customer groups are used for trial of illegal transactions. 13. The device according to claim 12, wherein the determining unit is further configured to determining an additional one or more customers of the plurality of customers related to each of the one or more customers based on the customer-related data and the extended rule data; and The additional one or more customers are included in a customer population associated with the individual suspect customer. 14. The apparatus of claim 12, further comprising a scoring unit configured to score each customer in the group of customers based on the acquired customer-related data for each customer in the group of customers related to the individual suspicious customer using a machine learning based scoring model; and A sorting unit configured to sort the customers in the customer group based on the scores. 15. The apparatus of claim 14, further comprising a training unit configured to determine, from the customer-related data of each reference customer in a set of reference customers, characteristic data representative of a plurality of customer characteristics of the reference customer; and use the characteristic data of the reference customer to train the Scoring models for machine learning. 16. The device of claim 15, wherein The determination unit is further configured to determine characteristic data representing a plurality of customer characteristics of the customer according to customer-related data of each customer in the customer group; The scoring unit is further configured to score each customer in the customer population based on the characteristic data representing a plurality of customer characteristics of the customer using a machine learning based scoring model; and The sorting unit is further configured to sort the customers in the customer group based on the score of each customer. 17. The apparatus according to any one of claims 12-16, wherein said expanding rule data comprises data mining of said plurality of customers based on customer-related data of each of said plurality of customers The obtained rule data represents a customer group related to the individual suspicious customer. 18. The device of claim 17, wherein the extended rule data further includes Pre-determined rule data representing the relevance of each customer in said group of customers to said individual suspect customer; and/or Pre-determined rule data related to the purpose of acquiring the customer group. 19. The device according to claim 17, wherein the second acquisition unit is further configured to identifying one or more customer groups among the plurality of customers by matching a predefined template defining relationships between corresponding respective customers with customer-related data of the plurality of customers structure and/or corresponding individual customer profile data; and Determining data representing a customer group related to the individual suspicious customer among the one or more customer groups as rule data representing a customer group related to the individual suspicious customer. 20. The device according to claim 17, wherein the second acquisition unit is further configured to Determining one or more customer groups from the plurality of customers based on customer-related data of the plurality of customers, such that each customer in each of the one or more customer groups is in the customer group Each has at least a predetermined number of relevant customers; determining data representing a customer group related to the individual suspicious customer in the one or more customer groups as rule data representing a customer group related to the individual suspicious customer .

Images