CN115292380B

CN115292380B - A data analysis method, device, computer equipment and storage medium

Info

Publication number: CN115292380B
Application number: CN202210598420.9A
Authority: CN
Inventors: 文广明; 田鸥
Original assignee: Ping An Bank Co Ltd
Current assignee: Ping An Bank Co Ltd
Priority date: 2022-05-28
Filing date: 2022-05-28
Publication date: 2025-06-10
Anticipated expiration: 2042-05-28
Also published as: CN115292380A

Abstract

The embodiment of the application provides a data analysis method, a device, computer equipment and a storage medium, which are applied to the field of financial science and technology, wherein the method comprises the steps of acquiring intra-line data, external data and public data of an offshore customer in a target bank; updating a share penetration diagram by using a first risk list included in the intra-row data, the external data and a second risk list included in the public data to obtain an updated share penetration diagram, acquiring a target transaction behavior feature set of the offshore client and the stock control data of the offshore client when the offshore client is determined to be a first type client according to the updated share penetration diagram, and calling an offshore client classification model to classify the offshore client according to the target transaction behavior feature set, the stock control data and the registration operation information of the offshore client included in the public data to obtain the category of the offshore client. By adopting the method and the device, the accuracy of identifying the offshore client category can be improved. The present application relates to blockchain technology, wherein the data can be obtained through a blockchain network.

Description

Data analysis method, device, computer equipment and storage medium

Technical Field

The present application relates to the field of data analysis technologies, and in particular, to a data analysis method, a data analysis device, a computer device, and a storage medium.

Background

The cross-border offshore customers have the natural advantages of convenience in overseas registration, free flow of funds, loose supervision and the like, and meanwhile, the offshore accounts (Off shore Account, OSA) become important tools for illegal behavior dependence such as transferring funds by telecommunication fraud molecules, so that the supervision order and financial stability of financial markets in China are serious.

The risk prevention and control means of the current cross-border offshore clients mainly identify the categories of the offshore clients according to the abnormal object lists provided by banks. However, the inventor finds that as the operation scale of offshore guest groups and banks is continuously expanded, the existing way of identifying the offshore guest class is single, not comprehensive enough and has lower accuracy.

Disclosure of Invention

The embodiment of the application provides a data analysis method, a data analysis device, computer equipment and a storage medium, which can improve the accuracy of identifying the categories of offshore clients.

In one aspect, an embodiment of the present application provides a data analysis method, including:

The method comprises the steps of acquiring intra-line data, external data and public data of an offshore client in a target bank, wherein the external data comprises international service provider data, the international service provider data comprises a first risk list and a stock right penetration map provided by an international service provider, and the public data comprises a second risk list provided by other institutions and registration management information of the offshore client;

Updating the equity penetration graph by using the in-line data, the first risk list and the second risk list to obtain an updated equity penetration graph;

Acquiring a target transaction behavior feature set of the offshore client and stock control data of the offshore client when the offshore client is determined to be a first-class client according to the updated stock right penetration graph;

And calling an offshore client classification model to classify the offshore client according to the target transaction behavior feature set, the indirection data of the offshore client and the registration management information of the offshore client, so as to obtain the category of the offshore client.

In still another aspect, an embodiment of the present application provides a data analysis apparatus, including:

The system comprises an acquisition module, a target bank, an offshore customer acquisition module and a public data acquisition module, wherein the acquisition module is used for acquiring intra-line data, external data and public data of the offshore customer in the target bank, the external data comprises international service provider data, the international service provider data comprises a first risk list and a stock right penetration map provided by an international service provider, and the public data comprises a second risk list provided by other institutions and registration management information of the offshore customer;

the updating module is used for updating the equity penetration graph by utilizing the in-line data, the first risk list and the second risk list to obtain an updated equity penetration graph;

The acquisition module is further used for acquiring a target transaction behavior feature set of the offshore client and stock control data of the offshore client when the offshore client is determined to be a first-class client according to the updated stock right penetration graph;

And the classification module is used for calling an offshore client classification model to classify the offshore client according to the target transaction behavior feature set, the indirection data of the offshore client and the registration management information of the offshore client, so as to obtain the category of the offshore client.

In yet another aspect, an embodiment of the present application provides a computer device, including a processor and a memory, where the processor and the memory are connected to each other, and the memory is configured to store computer program instructions, and the processor is configured to execute the program instructions to implement the data analysis method.

In yet another aspect, embodiments of the present application provide a computer readable storage medium having stored therein computer program instructions for performing the data analysis method when executed by a processor.

In summary, the computer device can acquire intra-row data, external data and public data of the offshore clients in the target bank, update the equity penetration graph by using a first risky list included in the intra-row data, the external data and a second risky list included in the public data to obtain an updated equity penetration graph, acquire the target transaction behavior feature set of the offshore clients and the control data of the offshore clients when the offshore clients are determined to be first-class clients according to the updated equity penetration graph, and call the offshore client classification model to classify the offshore clients according to the target transaction behavior feature set, the control data of the offshore clients and the registration management information of the offshore clients.

Drawings

In order to more clearly illustrate the embodiments of the application or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the application, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

FIG. 1 is a schematic flow chart of a data analysis method according to an embodiment of the present application;

FIG. 2 is a flow chart of a data analysis method according to another embodiment of the present application;

Fig. 3 is a schematic structural diagram of a data analysis device according to an embodiment of the present application;

Fig. 4 is a schematic structural diagram of a computer device according to an embodiment of the present application.

Detailed Description

The technical solutions in the embodiments of the present application will be described below with reference to the accompanying drawings in the embodiments of the present application.

Fig. 1 is a schematic flow chart of a data analysis method according to an embodiment of the application. The method may be applied to a computer device. The computer device may be a user terminal or a server. The user terminal may be a desktop computer or the like. The server may be an independent server, a server cluster or a distributed system formed by a plurality of physical servers, or a cloud server providing cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (Content Delivery Network, CDN), basic cloud computing services such as big data and artificial intelligence platforms, but is not limited thereto. The method may comprise the steps of:

And S101, acquiring intra-line data, external data and public data of the offshore clients in a target bank, wherein the external data comprises international service provider data, the international service provider data comprises a first risk list and a stock right penetration map provided by an international service provider, and the public data comprises a second risk list provided by other institutions and registration management information of the offshore clients.

Wherein an offshore customer refers to a company that registers outside the country and opens an offshore account within the environment. Wherein the inline data may include historical transaction behavior data. The historical transaction behavior data may include transaction flow data for an offshore account, and the transaction flow data may include account information for both parties to the transaction, transaction time, transaction amount, and the like. In one embodiment, the inline data may also include the debug information and the account opening information. The exhaustion information may include on-site exhaustion information and/or remote exhaustion information. The exhaustion information may include, in particular, basic information of an enterprise. The business base information may include business registration addresses, business office addresses, business employee numbers, and business annual business quotas. The account opening information can comprise indication information of whether to open accounts in different places or not and indication information of whether to open accounts in batches or not.

Wherein the external data may include international facilitator data. The international facilitator data may include a first risk list and equity penetration data. The first risk list may include an international facilitator, such as a list of distrusted abnormal objects (including information of distrusted enterprises, etc., one of the abnormal object lists) provided by an international authority credit rating institution, a degraded observed list (one of the suspicious lists). The equity penetration data may include equity penetration graphs. The equity penetration graph can reflect information such as the equity condition of an enterprise. In one embodiment, the external data may also include customs data. Customs data may include data such as tax refund information for offshore customers.

Wherein the public data may include a second risk list and registration management information for the offshore client. In one embodiment, the second risk list may include an international sanctioned abnormal object list (including enterprise identification information of enterprises sanctioned by the target country, one of the abnormal object lists), an international suspicious list (including one of the suspicious lists relating to legal cases in the target country, such as enterprise identification information of enterprises specifying the legal case or any legal case). The registration operation information may include an operation duration and a registration country. In one embodiment, the public data may include business information of the offshore clients. The registration management information may be obtained from business information. The business information may include information such as business identification information (e.g., business name), business legal information, business registration address, business hold time, business registration funds, business registration age, etc. The legal person to which the embodiments of the present application refer is a legal representative.

And S102, updating the equity penetration graph by using the in-line data, the first risk list and the second risk list to obtain an updated equity penetration graph.

The equity penetration graph may include enterprise identification information of the offshore client, identification information of an individual (such as a name of the individual) having an equity relationship with the offshore client, or enterprise identification information of a first enterprise (refer to an enterprise having an equity relationship with the offshore client), and an equity relationship corresponding to the offshore client. In one embodiment, the equity penetration graph may also include identification information of an individual in a controlling relationship with the first enterprise or enterprise identification information of a second enterprise in a controlling relationship with the first enterprise (referring to an enterprise in a controlling relationship with the first enterprise), a controlling relationship corresponding to the first enterprise, and so on.

In one embodiment, the updated equity penetration graph may also include enterprise identification information for the target enterprise, trade relationships between the offshore customers and the target enterprise, trade attribute information for the offshore customers, and enterprise attribute information for the offshore customers. A target enterprise refers to an enterprise that has a transactional relationship with an offshore customer. A trade relationship may be understood as a resource exchange relationship, such as a funding exchange relationship. The transaction attribute information may include information about a transaction time, a transaction amount, and the like with the target business.

In one embodiment, when the in-line data comprises historical transaction behavior data, perfect information and account opening information, the updated equity penetration graph can be obtained by performing entity identification on the historical transaction behavior data by computer equipment to obtain identification information of a target enterprise in a transaction relationship with the offshore client, performing relationship extraction on the historical transaction behavior data by the computer equipment to obtain a transaction relationship between the offshore client and the target enterprise, performing attribute extraction on the historical transaction behavior data by the computer equipment to obtain transaction attribute information of the offshore client, performing attribute extraction on the perfect information and the account opening information by the computer equipment to obtain first enterprise attribute information of the offshore client, generating a label indicating whether the offshore client is an abnormal object list client or not and indicating whether the offshore client is a suspicious client according to the first risk list and the second risk list to serve as second enterprise attribute information of the offshore client, and updating the equity penetration graph according to the identification information of the target enterprise, the transaction relationship between the offshore client and the target enterprise, the transaction attribute information of the offshore client, the first enterprise attribute information of the offshore client and the second equity penetration graph of the offshore client. In one embodiment, the computer device may generate a tag indicating that the offshore client is an outlier list client upon determining that the international service provider provided outlier list includes enterprise identification information of the offshore client and/or upon determining that the other institution provided outlier list includes enterprise identification information of the offshore client, the offshore client having the tag indicating that the offshore client is an outlier list client. The computer device may also be configured to generate a tag indicating that the offshore client is a suspicious client when it is determined that the suspicious list provided by the international service provider includes enterprise identification information for the offshore client and/or when it is determined that the suspicious list provided by other institutions includes enterprise identification information for the offshore client, the offshore client having the tag indicating that the offshore client is suspicious. The updated stock right penetration graph mentioned above may further include enterprise identification information of the target enterprise, a trade relationship between the offshore client and the target enterprise, trade attribute information of the offshore client, first enterprise attribute information of the offshore client, and second enterprise attribute information of the offshore client. The first enterprise attribute information may include information indicating whether the enterprise is working in a different place, the number of staff, the annual business, information indicating whether to open accounts in a different place, and information indicating whether to open accounts in a batch.

In one embodiment, where the external data further includes the aforementioned customs data and the public data further includes the aforementioned business information, the computer device may further determine the customs data and the business information as third enterprise attribute information for the offshore customer. Correspondingly, the method that the computer equipment updates the equity penetration graph according to the identification information of the target enterprise, the trade relation between the offshore client and the target enterprise, the trade attribute information of the offshore client, the first enterprise attribute information of the offshore client and the second enterprise attribute information of the offshore client can be that the computer equipment updates the equity penetration graph according to the identification information of the target enterprise, the trade relation between the offshore client and the target enterprise, the trade attribute information of the offshore client, the first enterprise attribute information of the offshore client, the second enterprise attribute information of the offshore client and the third enterprise attribute information of the offshore client to obtain the updated equity penetration graph. The updated stock right penetration graph may include enterprise identification information of the target enterprise, a trade relationship between the offshore client and the target enterprise, trade attribute information of the offshore client, first enterprise attribute information of the offshore client, second enterprise attribute information of the offshore client, and third enterprise attribute information of the offshore client.

In one embodiment, when the in-line data comprises historical transaction behavior data, the updated stock right penetration graph can be obtained by performing entity identification on the historical transaction behavior data by computer equipment to obtain identification information of a target enterprise in a transaction relationship with the offshore client, performing relationship extraction on the historical transaction behavior data by the computer equipment to obtain a transaction relationship between the offshore client and the target enterprise, performing attribute extraction on the historical transaction behavior data by the computer equipment to obtain transaction attribute information of the offshore client, generating a label indicating whether the offshore client is an abnormal object list client and whether the offshore client is a suspicious client or not according to the first risk list and the second risk list to serve as second enterprise attribute information of the offshore client, and updating the stock right penetration graph by the computer equipment according to the identification information of the target enterprise, the transaction relationship between the offshore client and the target enterprise, the transaction attribute information of the offshore client and the second enterprise attribute information of the offshore client to obtain the updated stock right penetration graph. The updated stock right penetration graph further comprises enterprise identification information of the target enterprise, trade relation between the offshore client and the target enterprise, trade attribute information of the offshore client and second enterprise attribute information of the offshore client.

In one embodiment, when the external data further includes the aforementioned customs data and the public data further includes the aforementioned business information, the computer device may further determine the customs data and the business information as third enterprise attribute information of the offshore client, and accordingly, the computer device updates the equity penetration map according to the identification information of the target enterprise, the trade relationship between the offshore client and the target enterprise, the trade attribute information of the offshore client, and the second enterprise attribute information of the offshore client, and the updated equity penetration map may be obtained by the computer device updating the equity penetration map according to the identification information of the target enterprise, the trade relationship between the offshore client and the target enterprise, the trade attribute information of the offshore client, the second enterprise attribute information of the offshore client, and the third enterprise attribute information of the offshore client, and the updated equity penetration map may further include the third enterprise attribute information of the offshore client. The updated stock right penetration graph further comprises enterprise identification information of the target enterprise, trade relation between the offshore client and the target enterprise, trade attribute information of the offshore client, second enterprise attribute information of the offshore client and second enterprise attribute information of the offshore client.

And S103, when the offshore client is determined to be a first-class client according to the updated share right penetration graph, acquiring a target transaction behavior feature set of the offshore client and stock control data of the offshore client.

In the embodiment of the application, when the offshore client is determined to be the first-class client according to the updated equity penetration graph, the fact that the class to which the offshore client specifically belongs cannot be determined temporarily according to the updated equity penetration graph is indicated, so that the computer equipment can also acquire the target transaction behavior feature set of the offshore client and the stock control data of the offshore client for further determining the class to which the offshore client belongs. Wherein the target transaction behavior feature set is obtained according to the historical transaction data. And controlling the stock data to be obtained according to the updated stock right penetration graph.

In one embodiment, the computer device may determine whether the offshore client is a second type of client based on the updated equity penetration graph, may further determine whether the offshore client is a third type of client based on the updated equity penetration graph, and may further determine that the offshore client is a first type of client when determining that the offshore client is not the second type of client or the third type of client based on the updated equity penetration graph. In one application scenario, the first type of client may be a client to be observed, the second type of client may be an intensive call client, and the third type of client may be a premium type of client. Wherein the offshore customer is determined to be an intensive customer, indicating that the intensive job investigation should be enhanced for the offshore customer. Critical information that a company or individual affects a transaction can be discovered by enhancing the coordination of such offshore customers, thereby circumventing potential risks.

In one embodiment, the computer device may determine whether the offshore client is a second type client by determining that the offshore client is a second type client if the computer device determines that the offshore client is an abnormal object list client based on the updated equity penetration graph, a relationship between the offshore client and an abnormal object list enterprise satisfies a preset condition, a transaction behavior of the offshore client is not in compliance with a specification, or the offshore client is suspicious.

In one embodiment, the computer device may determine whether the offshore client is an abnormal object list client by determining whether the second enterprise attribute information of the offshore client in the updated equity penetration graph includes a tag indicating that the offshore client is an abnormal object list client, and if so, determining that the offshore client is an abnormal object list client.

In one embodiment, the computer device may determine whether the relationship between the offshore client and the abnormal object list enterprise satisfies the preset condition by determining that the enterprise is an abnormal object list enterprise when the enterprise identification information of any enterprise in the updated equity penetration graph appears in an abnormal object list provided by the international server or appears in an abnormal object list provided by other institutions, and determining that the relationship between the offshore client and the abnormal object list enterprise satisfies the preset condition. Because a certain association relationship exists among a plurality of enterprises in the updated equity penetration graph, when the enterprises in the updated equity penetration graph are determined to be the abnormal object list enterprises, the certain association relationship between the offshore clients and the abnormal object list enterprises in the updated equity penetration graph can be indicated, and therefore the relationship between the offshore clients and the abnormal object list enterprises can be determined to meet the preset condition. In one embodiment, the computer device may determine whether the relationship between the offshore client and the abnormal object list enterprise satisfies the preset condition by determining that the target enterprise is an abnormal object list enterprise when the enterprise identification information of the target enterprise in the equity penetration graph appears in an abnormal object list provided by the international service provider or appears in an abnormal object list provided by other institutions, and determining that the relationship between the offshore client and the abnormal object list enterprise satisfies the preset condition. The target enterprise may include an enterprise having a first-level relationship with the offshore client, where the first-level relationship indicates that a node corresponding to the updated equity penetration graph of the target enterprise is directly connected to a node corresponding to the updated equity penetration graph of the offshore client.

In one embodiment, the computer device may invoke a transaction class rule to determine whether the transaction behavior of the offshore customer meets specifications based on the updated equity penetration graph, the transaction class rule may include a determination rule set based on the transaction attribute information of the offshore customer, and/or the computer device may invoke a static information class rule to determine whether the offshore customer is suspicious based on the updated equity penetration graph, the static information class rule may include a determination rule set based on the enterprise attribute information of the offshore customer, and/or the computer device may invoke a suspicious list class rule to determine whether the offshore customer is suspicious based on the updated equity penetration graph.

In one embodiment, the method that the computer equipment invokes the transaction class rule to judge whether the transaction behavior of the offshore client accords with the specification according to the updated stock right penetration graph can be that the computer equipment judges whether the transaction behavior of the offshore client accords with the specification according to the transaction attribute information of the offshore client, and if the transaction behavior of the offshore client does not accord with the specification, the offshore client is determined to be a second class client. The updated equity penetration map includes transaction attribute information for the offshore customer. In one embodiment, the computer device determines whether the transaction behavior of the offshore client meets the specification based on the transaction data information of the offshore client as follows:

① When the transaction flow of the offshore client and the operation scale of the offshore client are determined to be not in accordance with the convention according to the transaction attribute information of the offshore client, the transaction behavior of the offshore client is determined to be not in accordance with the specification. For example, for a first-creation enterprise that just holds for one or two years, if the monthly transaction flow is too large, it indicates that the first-creation enterprise may have transaction risk, so it may be determined that the transaction behavior of the first-creation enterprise is not in compliance with the specification.

② When it is determined that the offshore client has a funds transitive transaction based on the transaction attribute information of the offshore client, it is determined that the transaction behavior of the offshore client is not in compliance with the specification. The transitional fund transaction has the characteristics that the daily transfer-in and transfer-out amounts are basically consistent, and the transaction presents obvious scattered transfer-in.

In one embodiment, in addition to determining whether the transaction performance of the offshore customer meets the specifications in the manner described above, the computer device may determine whether the transaction performance of the offshore customer meets the specifications by whether the private rotation is frequent (e.g., the frequency of the private rotation is greater than a first predetermined frequency), the revolution is frequent (e.g., the frequency of the revolution is greater than a second predetermined frequency), and/or whether the bulk transaction is frequent (e.g., the frequency of the bulk transaction is greater than a second predetermined frequency), and/or whether the funds are fast-forwarded (e.g., the frequency of the transfer of funds is greater than a third predetermined frequency) and the amount of remaining resources of the offshore account is less than a predetermined value (e.g., the balance of the offshore account is less than a predetermined value).

In one embodiment, the computer device may invoke the static information class rule to determine whether the offshore client is suspicious based on the updated equity penetration graph in such a manner that the computer device determines that the offshore client is suspicious when the first enterprise attribute information of the offshore client includes less than a predetermined number of enterprise employees. And/or the computer device determines that the offshore client is suspicious when the first enterprise attribute information of the offshore client includes an enterprise annual business value that is less than a preset value. And/or the computer device determines that the offshore customer is suspicious when the first enterprise attribute information of the offshore customer includes information indicative of a foreign office. And/or the computer device determines that the offshore customer is suspicious when the first enterprise attribute information for the offshore customer includes information indicative of an off-site account opening. And/or, the computer device determining that the offshore customer is suspicious when the first enterprise attribute information of the offshore customer includes information indicative of a bulk account opening. The updated knowledge-graph includes first enterprise attribute information for the offshore client.

In one embodiment, the computer device may further determine that the offshore client is suspicious when the customs data included in the third enterprise attribute information of the offshore client does not match corresponding transaction data. For example, the tax refund amount and the transaction amount are not matched, such as 1000 ten thousand of the export transaction amount, but the export tax refund amount is only 500 ten thousand, which indicates that the offshore client is suspicious of avoiding tax, so that the offshore client can be determined to be suspicious.

In one embodiment, the computer device may invoke the suspicious list class rules to determine whether the offshore client is suspicious based on the updated equity penetration graph in such a manner that the computer device may determine that the offshore client is an enhanced out-of-tune client when the second enterprise attribute information for the offshore client includes a label indicating that the offshore client is a suspicious client. The updated knowledge-graph includes second enterprise attribute information for the offshore client.

In one embodiment, the computer device may determine whether the offshore client is a third class client by determining that the offshore client is a third class client when the computer device determines that the type of enterprise of the offshore client is the target type based on the updated equity penetration graph. The target type may be a type of enterprise indicating that the enterprise is more reliable, such as a marketing enterprise, a national control stock, state-owned enterprise or a central enterprise.

In one embodiment, the manner in which the computer device obtains the holdings data of the offshore customer may be for the computer device to determine the number of holdings of the offshore customer and/or the number of holdings of the legal representative of the offshore customer based on the updated equity penetration map, and determine the number of holdings of the offshore customer and/or the number of holdings of the legal representative of the offshore customer as the holdings data of the offshore customer.

S104, invoking an offshore customer classification model to classify the offshore customer according to the target transaction behavior feature set, the indirection data of the offshore customer and the registration management information of the offshore customer, and obtaining the class of the offshore customer.

In the embodiment of the application, the computer equipment can input the target transaction behavior feature set, the stock control data and the registration management information into the offshore client classification model for classification processing to obtain the categories of the offshore clients. It should be noted that, the offshore clients mentioned in step S103 belong to the first class of clients and are not the final classification results of the offshore clients, and the classification of the offshore clients obtained by the classification in step S104 is the final classification results of the offshore clients. The classification process mainly relies on the target transaction behavior feature set to classify, and the stock control data and the registration management information play a role in assisting classification. The final classification result of the offshore customer may be one of a first type of customer, a second type of customer, and a third type of customer.

In one embodiment, the computer device may conduct transaction amount management for the offshore customer based on the category to which the offshore customer belongs. For example, when the offshore client is a first type client, the upper limit of the transaction amount of the offshore account of the offshore client can be adjusted to the target value, and when the transaction amount of the offshore account exceeds the target value, the prompting information of insufficient amount is fed back to the offshore client. When the off-shore client is the second-class client, the off-shore client can be prompted to carry out identity authentication or prompt the off-shore client to submit related data to a bank counter for data auditing when the second-class client uses the off-shore account to conduct transactions. In one embodiment, after passing the identity authentication or data auditing of the offshore customer, the offshore customer may be assigned a transaction amount for the offshore customer to conduct a transaction based on the assigned transaction amount. Where the offshore customer is a third type of customer, the restrictions on the transaction amount of the offshore customer's offshore account may be removed. Potential transaction risk can be reduced by the transaction amount management policies described above.

It can be seen that in the embodiment shown in fig. 1, the computer device may acquire intra-row data, external data and public data of the offshore client in the target bank, update the equity penetration graph by using the first risk list included in the intra-row data, the external data and the second risk list included in the public data to obtain the updated equity penetration graph, and acquire the target transaction behavior feature set of the offshore client and the control data of the offshore client when determining that the offshore client is the first type client according to the updated equity penetration graph, and call the offshore client classification model to classify the offshore client according to the target transaction behavior feature set, the control data of the offshore client and the registration management information of the offshore client, so as to obtain the class of the offshore client.

Fig. 2 is a flow chart of a data analysis method according to another embodiment of the application. The method may be applied to the aforementioned computer device. Specifically, the method may comprise the steps of:

s201, historical transaction behavior data of each sample client in a plurality of sample clients, stock control data of each sample client and registration management information of each sample client are obtained.

S202, processing according to the historical transaction behavior data of each sample client to obtain a transaction behavior feature set of each sample client, wherein the transaction behavior feature set comprises a plurality of transaction behavior features.

In the embodiment of the application, the mode that the computer equipment can process the historical transaction behavior data of each sample client to obtain the transaction behavior feature set of each sample client can count the transaction behavior data of each sample client in a plurality of time windows and the transaction behavior data of each frequency window in the plurality of frequency windows according to the historical transaction behavior data of each sample client, the computer equipment processes the transaction behavior data of the sample client in each time window through a first discrete function and a first continuous function to obtain the transaction behavior feature corresponding to the sample client in each time window, the computer equipment processes the transaction behavior data of the sample client in each frequency window through a second discrete function and a second continuous function to obtain the transaction behavior feature corresponding to each frequency window, and the computer equipment constructs the transaction behavior feature set of each sample client, wherein the transaction behavior feature set comprises the transaction behavior feature corresponding to each time window and the transaction behavior feature corresponding to each frequency window. The scheme adopts the characteristic engineering to generate multidimensional transaction behavior characteristics, and enriches the transaction behavior characteristics.

In one embodiment, the computer device may call an RFM (Recency, frequency, monnetary) algorithm to process the historical transaction behavior data for each sample customer to obtain a set of transaction behavior characteristics for each sample customer. The computer equipment can count transaction behavior data of each sample client in each time window and transaction behavior data of each frequency window in the plurality of time windows according to historical transaction behavior data of each sample client, process the transaction behavior data of the sample client in each time window through a first discrete function and a first continuous function to obtain transaction behavior characteristics of the sample client in each time window, process the transaction behavior data of the sample client in each frequency window through a second discrete function and a second continuous function to obtain transaction behavior characteristics of the sample client in each frequency window, and construct a transaction behavior characteristic set of each sample client, wherein the transaction behavior characteristic set comprises the transaction behavior characteristics corresponding to each time window and the transaction behavior characteristics corresponding to each frequency window.

The plurality of time windows may include a plurality of short-term time windows and a plurality of long-term time windows, where each short-term time window corresponds to a different time length, and each long-term time window corresponds to a different time length. For example, the short-term time windows may be in hours, and the plurality of short-term time windows may include 1 hour (last 1 hour), 24 hours (last 24 hours), 48 hours (last 48 hours), and so on. In one embodiment, the short-term time window is no longer than 48 hours in length. The long-term time window may be in months, and the plurality of long-term time windows may include 1 month (last 1 month), 3 months (last 3 months), 6 months (last 6 months). In one embodiment, the long time window has a time length of no more than 6 months at maximum. The transaction behavior data for each of the plurality of time windows may include a transaction count for the time window, a transaction amount for each transaction, a transaction time for each transaction, and so forth.

The multiple frequency windows may include a first preset frequency, such as 3 times (last 1 time), a second preset frequency, such as 5 times (last 5 times), and in one embodiment, the multiple frequency windows may further include a third preset frequency, such as 10 times (last 10 times), and so on. The transaction behavior data of each of the plurality of frequency windows may include a transaction count of the frequency window, a transaction amount per transaction, a transaction time per transaction, and so on.

The transaction behavior characteristics obtained by adopting the method can comprise the number of days with transactions in a time window, the number of transaction days with account balance smaller than a preset amount in the time window, and the ratio of the number of transaction days with account balance smaller than the preset amount in the time window to the total number of transaction days. For example, the long time window is 1 month, wherein there are 20 days of transactions, and the number of days of transactions in which the account balance is less than the preset amount in the 20 days is 10 days. The transaction days for which the account balance is less than the preset amount in these 20 days are 10 days, and the ratio calculated here is 1/2.

The processing mode of the first continuous function may include maximum value calculation, minimum value calculation, accumulated value calculation, and average value calculation, and the transaction behavior feature obtained by this mode may include a transaction amount maximum value, a transaction amount minimum value, a transaction amount accumulated value, and a transaction amount average value corresponding to the time window. For example, after the transaction data of the time windows including 1 hour, 12 hours, 1 month and 3 months are processed by the first continuous function, the transaction amount maximum value, the transaction amount minimum value, the transaction amount cumulative value and the transaction amount mean value (which are the transaction behavior characteristics of the time window of 1 hour) within 1 hour, the transaction amount maximum value, the transaction amount minimum value, the transaction amount cumulative value and the transaction amount mean value (which are the transaction behavior characteristics of the time window of 12 hours) within 12 hours, the transaction amount maximum value, the transaction amount minimum value, the transaction amount cumulative value and the transaction amount mean value within 1 month (which are the transaction behavior characteristics of the time window of 1 month) within 3 months, and the transaction amount maximum value, the transaction amount minimum value, the transaction amount cumulative value and the transaction amount mean value (which are the transaction behavior characteristics of the time window of 3 months) can be obtained.

The processing mode of the second discrete function may include counting statistics, frequency statistics and ratio calculation, and the transaction behavior characteristic obtained by adopting the mode may include a number of days with transactions in a frequency window, a number of transaction days with account balance less than a preset amount in the frequency window, and a ratio of a number of transaction days with account amount less than the preset amount in the frequency window to a total number of transaction days.

The processing mode of the second continuous function may include maximum value calculation, minimum value calculation, accumulated value calculation, and average value calculation, and the transaction behavior feature obtained by this mode may include a transaction amount maximum value, a transaction amount minimum value, a transaction amount accumulated value, and a transaction amount average value corresponding to the frequency window. For example, the multiple frequency windows include 3 times, 5 times and 10 times, and after the transaction behavior data of the frequency windows are processed by the second continuous function, the transaction amount maximum value, the transaction amount minimum value, the transaction amount accumulated value and the transaction amount mean value (corresponding to 3 times of transaction behavior characteristics) of the last 3 times of transactions can be obtained, and the transaction amount maximum value, the transaction amount accumulated value and the transaction amount mean value (corresponding to 5 times of transaction behavior characteristics) of the last 5 times of transactions are obtained, and the transaction amount maximum value, the transaction amount minimum value, the transaction amount accumulated value and the transaction amount mean value (corresponding to 10 times of transaction behavior characteristics) of the last 3 times of transactions are obtained.

It should be noted that, the feature engineering is the core of the artificial intelligence AI algorithm model, and the quality of the feature engineering directly determines the performance and effect of the algorithm model. In order to find out more potential risk customers and accurately sense risks, the embodiment of the application can construct hundreds of thousands of dimension characteristics in terms of transaction behaviors by using an RFM-based explosive characteristic derivation scheme, accurately delineate the target of each account-moving transaction, screen indexes for distinguishing good and bad samples from the targets, and strictly control the dimension of a modeling variable. Wherein, the partial information of the characteristic derivation scheme can be seen in the following table:

S203, feature screening is carried out on the transaction behavior feature set of each sample client to obtain a sub transaction behavior feature set of each sample client, wherein the sub transaction behavior feature set comprises at least one transaction behavior feature for distinguishing the category to which each sample client belongs.

The transaction behavior feature set of each sample client corresponds to a plurality of feature categories, and the transaction behavior feature set comprises transaction behavior features corresponding to each feature category in the plurality of feature categories. Wherein, a plurality of feature categories corresponding to the transaction behavior feature sets of each sample client are the same. For example, the transaction behavior feature set of sample client 1 corresponds to feature class 1. m is a positive integer. The set of transaction behavioral characteristics of sample client 2 also corresponds to characteristic class 1. The sub-transaction behavior feature set of each sample client corresponds to at least one feature class, and the sub-transaction behavior feature set comprises transaction behavior features corresponding to each feature class in the at least one feature class. Wherein, at least one characteristic category corresponding to the sub-transaction behavior characteristic set of each sample client is the same. For example, the sub-transaction behavioral feature set of sample client 1 corresponds to feature class 1. The sub-transaction behavioral feature set of sample client 2 also corresponds to feature class 1.

In one embodiment, the computer device may calculate an information amount (Information Value, IV) value for each of the plurality of feature classes based on the set of transaction behavior features for each sample customer, and rank the plurality of feature classes according to the IV value, thereby obtaining a set of sub-transaction behavior features for each sample customer based on the ranking result. The sub-transaction behavior feature set of the sample client may include transaction behavior features corresponding to each of the K preceding feature types. The manner of ordering may be to order the IV values from front to back, with the order being higher the earlier the IV values are. The value range of the IV value is [0 ], positive infinity). The above procedure can achieve the selection of the modulus-entering variable.

S204, training an initial deep learning model by using the sub-transaction behavior feature set of each sample client, the indirection data of each sample client and the registration management information of each sample client to obtain a trained deep learning model to serve as an offshore client classification model, wherein the target transaction behavior feature set is the sub-transaction behavior feature set of the offshore client.

In the embodiment of the application, the computer equipment can obtain a plurality of training samples according to the sub-transaction behavior feature set of each sample client, the control data of each sample client and the registration management information of each sample client, train an initial deep learning model by utilizing the plurality of training samples, and obtain a trained deep learning model to serve as an offshore client classification model. A training sample may include a set of sub-transaction behavioral characteristics of a sample customer, indirection data for the sample customer, and registration business information for the sample customer. The deep learning model may be a lifting model, such as a lightweight gradient lifting Machine (LIGHT GRADIENT lifting Machine, lightGBM) model.

It should be noted that, the conventional model calculation information gains such as the extreme gradient lifting XGBoost model and the model calculation information gains such as the extreme gradient lifting GBDT model need to scan all training samples to find the optimal dividing point, when Gao Weida data or features are faced, the efficiency and expansibility seriously affect the effect of the model, and the most direct and effective method for solving the problem is to reduce the sample size and the feature size under the condition of not affecting the precision. The LightGBM algorithm is an optimization of the traditional GBDT algorithm, and is added with a Gradient-based One-SIDE SAMPLING, GOSS algorithm and a mutual exclusion feature binding (Exclusive Feature Bundling, EFB) algorithm. The GOSS algorithm calculates information gain through sample sampling and uses a large gradient sample to accelerate calculation speed, and the EFB algorithm further improves model training speed through feature sampling and improves learning efficiency.

The GOSS algorithm processing flow may include the following:

a. sequencing all training samples in descending order according to the absolute value of the gradient of each training sample, and taking the first a training samples as a sample subset A;

b. randomly sampling B training samples from the training samples except the first a training samples as a sample subset B;

c. the information gain is estimated from sample subset a and sample subset B. The formula for estimating the information gain is as follows:

Wherein:

A_l＝{x_i∈A：x_ij≤d},A_r＝{x_i∈A：x_ij>d}

B_l＝{x_i∈B：x_ij≤d},B_r＝{x_i∈B：x_ij>d}

Wherein O represents a training sample of a node to be split in the decision tree, gi is a gradient of the training sample, The number of samples of the left child node of the d node,The number of samples for the right child node of node d,The weight coefficients representing the small gradient samples.

The EFB algorithm processing flow may include the following:

a. Ordering the transaction behavior features according to the number of non-zero values;

b. calculating conflict ratios between different transaction behavior features;

c. And traversing each transaction behavior feature, carrying out feature combination, and calculating conflict ratios among different combined features to minimize the conflict ratios to obtain a trained LightGBM model.

And S205, acquiring intra-line data, external data and public data of the offshore clients in the target bank, wherein the external data comprises international service provider data, the international service provider data comprises a first risk list and a stock right penetration map provided by an international service provider, and the public data comprises a second risk list provided by other institutions and registration management information of the offshore clients.

And S206, updating the equity penetration graph by using the in-line data, the first risk list and the second risk list to obtain an updated equity penetration graph.

S207, when the offshore client is determined to be a first-class client according to the updated share right penetration graph, acquiring a target transaction behavior feature set of the offshore client and stock control data of the offshore client;

S208, invoking an offshore customer classification model to classify the offshore customer according to the target transaction behavior feature set, the indirection data of the offshore customer and the registration management information of the offshore customer, and obtaining the class of the offshore customer.

Step S205 to step S208 may refer to step S101 to step S104 in the embodiment of fig. 1, which are not described herein.

In one embodiment, the manner in which the computer device obtains the target transaction behavior feature set of the offshore client may be that the computer device obtains historical transaction behavior data of the offshore client, and obtains the sub-transaction behavior feature set of the offshore client as the target transaction behavior feature set of the offshore client according to the historical transaction behavior data of the offshore client. Here, since at least one feature type has been determined from a plurality of feature types by means of feature screening, such as the first K feature types from the plurality of feature types, before the initial deep learning model is trained, the sub-transaction behavioral feature set of the offshore client may include the at least one feature category, such as the transaction behavioral feature corresponding to each of the first K feature types. That is, the computer device may not need to perform the processing according to the historical transaction behavior data to obtain the transaction behavior feature set of the offshore client in the above process, and perform the feature screening operation based on the transaction behavior feature set of the offshore client, so that the classification efficiency of the offshore client can be improved.

In one embodiment, the method that the computer equipment acquires the target transaction behavior feature set of the offshore client can acquire historical transaction behavior data of the offshore client for the computer equipment, and obtain the transaction behavior feature set of the offshore client according to the historical transaction behavior data of the offshore client, and the computer equipment performs feature screening on the transaction behavior feature set of the offshore client to obtain the sub-transaction behavior feature set of the offshore client to serve as the target transaction behavior feature set of the offshore client. The method that the computer device processes the historical transaction behavior data of the offshore client to obtain the transaction behavior feature set of the offshore client can refer to the method that the computer device processes the historical transaction behavior data of the sample client to obtain the transaction behavior feature set of the sample client, which is not described herein. The method for performing feature screening on the transaction behavior feature set of the offshore client by the computer device to obtain the sub-transaction behavior feature set of the offshore client can refer to the method for performing feature screening on the transaction behavior feature set of the sample client by the computer device to obtain the sub-transaction behavior feature set of the sample client, which is not described herein.

According to the embodiment of the application, according to the cross-border off-shore transaction mode and compliance requirements, by introducing in-line data, external data and public data and providing a rule model and AI algorithm model combined solution, the types of cross-border off-shore clients can be automatically divided, then different levels of off-shore clients can be subjected to differentiated management, systematic and intelligent interception and management of abnormal services of the cross-border off-shore clients are realized, the transition from manual monitoring to system monitoring of risk prevention is realized, and the risk prevention and control capability of the off-shore clients is comprehensively improved. Compared with the method that the existing technology singly combines data such as an abnormal object list and the like provided by a bank to perform risk investigation on offshore clients, the method and the system for automatically and intelligently determining the categories of the offshore clients by combining multidimensional information are higher than the existing technology in risk investigation efficiency and reliability, and can meet the current risk prevention and control requirements on the offshore clients. In addition, the embodiment of the application enriches the existing risk monitoring dimension by integrating various data sources and analyzing by means of a knowledge graph, and solves the problem that the pain point of a client cannot be comprehensively and penetratingly known at present. In addition, the embodiment of the application can solve the problems of low operation efficiency and poor expansibility caused by using a conventional model by using LightGBM algorithm, and can reduce model errors and improve model accuracy by using a leaf-by-leaf growth (leaf-wise) strategy with depth limitation for multi-thread optimization. In summary, the embodiment of the application comprehensively applies an AI algorithm model and a rule model, enriches the monitoring means of the risk of the cross-border offshore clients, makes up the defects of the existing rule model, and comprehensively improves the identification accuracy of the risk clients.

It can be seen that in the embodiment shown in fig. 2, the computer device may obtain historical transaction behavior data of each sample client, stock control data of each sample client and registration management information of each sample client in the plurality of sample clients, process the historical transaction behavior data of each sample client to obtain a transaction behavior feature set of each sample client, and perform feature screening on the transaction behavior feature set of each sample client to obtain a sub-transaction behavior feature set of each sample client, and train an initial deep learning model by using the sub-transaction behavior feature set of each sample client, the stock control data of each sample client and the registration management information of each sample client to obtain a trained deep learning model as an offshore client classification model, and train the model training efficiency based on the screened feature training model, so as to avoid adverse effects caused by useless features on the prediction effect of the model.

The present application relates to blockchain technology, and various types of data or specified types of data as referred to by embodiments of the present application may be obtained through a blockchain network. The data of each type related to the embodiment of the application obtains the use authorization of the corresponding data body or the data controller.

Fig. 3 is a schematic structural diagram of a data analysis device according to an embodiment of the application. The apparatus may be applied to the aforementioned computer device. Specifically, the data analysis device may include:

The system comprises an acquisition module 301, configured to acquire intra-line data, external data and public data of an offshore customer at a target bank, wherein the external data comprises international service provider data, the international service provider data comprises a first risk list and a stock right penetration map provided by an international service provider, and the public data comprises a second risk list provided by other institutions and registration management information of the offshore customer.

And an updating module 302, configured to update the equity penetration graph by using the in-line data, the first risk list, and the second risk list, and obtain an updated equity penetration graph.

The obtaining module 301 is further configured to obtain a target transaction behavior feature set of the offshore client and live data of the offshore client when the offshore client is determined to be a first type of client according to the updated share right penetration graph.

And the classification module 303 is configured to invoke an offshore client classification model to classify the offshore client according to the target transaction behavior feature set, the indirection data of the offshore client, and the registration management information of the offshore client, so as to obtain a class of the offshore client.

In an alternative embodiment, the in-line data includes historical transaction behavior data, exhaustion information, and account opening information, and the update module 302 is specifically configured to:

entity identification is carried out on the historical transaction behavior data, and identification information of a target enterprise in transaction relation with the offshore client is obtained;

Performing relation extraction on the historical transaction behavior data to obtain a transaction relation between the offshore client and the target enterprise;

Performing attribute extraction on the historical transaction behavior data to obtain transaction attribute information of the offshore clients;

Performing attribute extraction on the adjustment information and the account opening information to obtain first enterprise attribute information of the offshore client;

generating a label indicating whether the offshore client is an abnormal object list client and whether the offshore client is a suspicious client according to the first risk list and the second risk list to serve as second enterprise attribute information of the offshore client;

and updating the equity penetration graph according to the identification information of the target enterprise, the trade relation between the offshore client and the target enterprise, the trade attribute information of the offshore client, the first enterprise attribute information of the offshore client and the second enterprise attribute information of the offshore client, and obtaining the updated equity penetration graph.

In an alternative embodiment, the apparatus may further comprise a determination module 304. A determining module 304, configured to:

if the offshore client is determined to be an abnormal object list client according to the updated stock right penetration graph, the relationship between the offshore client and an abnormal object list enterprise meets preset conditions, the transaction behavior of the offshore client is not in accordance with the specification or the offshore client is suspicious, determining that the offshore client is a second-class client;

If the enterprise type of the offshore client is determined to be the target type according to the updated stock right penetration graph, determining that the offshore client is a third type client;

if the offshore client is not the second type client and is not the third type client, determining that the offshore client is a first type client.

In an alternative embodiment, the apparatus may further comprise a determining module 305. A judging module 305, configured to:

invoking a transaction class rule to judge whether the transaction behavior of the offshore client meets the specification according to the updated stock right penetration graph, wherein the transaction class rule comprises a judgment rule set according to the transaction attribute information of the offshore client, and/or,

Invoking static information class rules to determine whether the offshore client is suspicious based on the updated equity penetration graph, the static information class rules including determination rules set based on enterprise attribute information of the offshore client, and/or,

And calling a suspicious list class rule to judge whether the offshore client is suspicious according to the updated share right penetration graph.

In an alternative embodiment, the apparatus may further comprise a training module 306. Training module 306 for:

Acquiring historical transaction behavior data of each sample client, stock control data of each sample client and registration management information of each sample client in a plurality of sample clients;

Processing according to the historical transaction behavior data of each sample client to obtain a transaction behavior feature set of each sample client, wherein the transaction behavior feature set comprises a plurality of transaction behavior features;

Feature screening is carried out on the transaction behavior feature set of each sample client to obtain a sub transaction behavior feature set of each sample client, wherein the sub transaction behavior feature set comprises at least one transaction behavior feature for distinguishing the category to which each sample client belongs;

And training an initial deep learning model by using the sub-transaction behavior feature set of each sample client, the stock control data of each sample client and the registration management information of each sample client to obtain a trained deep learning model to serve as an offshore client classification model, wherein the target transaction behavior feature set is the sub-transaction behavior feature set of the offshore client.

In an alternative embodiment, training module 306 is specifically configured to:

And processing according to the historical transaction behavior data of each sample client to obtain a transaction behavior feature set of each sample client, wherein the transaction behavior feature set comprises the following steps:

Counting transaction behavior data of each sample client in each time window and transaction behavior data of each frequency window in a plurality of frequency windows according to the historical transaction behavior data of each sample client;

Processing transaction behavior data of the sample client in each time window through a first discrete function and a first continuous function to obtain transaction behavior characteristics of the sample client corresponding to each time window;

Processing the transaction behavior data of the sample client in each frequency window through a second discrete function and a second continuous function to obtain the transaction behavior characteristics of the sample client corresponding to each frequency window;

and constructing a transaction behavior feature set of each sample client, wherein the transaction behavior feature set comprises transaction behavior features corresponding to each time window of the sample client and transaction behavior features corresponding to each frequency window of the sample client.

In an alternative embodiment, the obtaining module 301 is further specifically configured to determine, according to the updated share right penetration graph, a number of companies that are in possession of the offshore client and/or a number of companies that are in possession of the offshore client by legal representatives, and determine, as the in possession data of the offshore client, the number of companies that are in possession of the offshore client and/or the number of companies that are in possession of the offshore client by legal representatives.

It can be seen that in the embodiment shown in fig. 3, the data analysis device may acquire intra-row data, external data and public data of the offshore client in the target bank, update the equity penetration map by using the first risk list included in the intra-row data, the external data and the second risk list included in the public data to obtain an updated equity penetration map, and acquire the target transaction behavior feature set of the offshore client and the control data of the offshore client when determining that the offshore client is the first type client according to the updated equity penetration map, and call the offshore client classification model to classify the offshore client according to the target transaction behavior feature set, the control data of the offshore client and the registration management information of the offshore client, so as to obtain the class of the offshore client.

Fig. 4 is a schematic structural diagram of a computer device according to an embodiment of the present application. The computer device described in this embodiment may include one or more processors 1000 and memory 2000. The processor 1000 and the memory 2000 may be connected by a bus.

The Processor 1000 may be a central processing module (Central Processing Unit, CPU), which may also be other general purpose processors, digital signal processors (DIGITAL SIGNAL Processor, DSP), application SPECIFIC INTEGRATED Circuit (ASIC), off-the-shelf Programmable gate array (Field-Programmable GATE ARRAY, FPGA) or other Programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or the like. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.

The memory 2000 may be a high-speed RAM memory or a non-volatile memory (non-volatile memory), such as a disk memory. Wherein the memory 2000 is used for storing a computer program comprising program instructions configured to invoke the processor 1000 to perform the steps of:

In one embodiment, the in-line data includes historical transaction behavior data, debug information, and account opening information, and when updating the equity penetration map using the in-line data, the first risk list, and the second risk list to obtain an updated equity penetration map, the processor 1000 is configured to invoke the program instructions to specifically perform the following steps:

In one embodiment, the processor 1000 is configured to invoke the program instructions, and further performs the steps of:

In one embodiment, when the transaction behavior feature set of each sample client is obtained from the historical transaction behavior data processing of each sample client, the processor 1000 is configured to invoke the program instructions, specifically to perform the following steps:

In one embodiment, in acquiring live data of the offshore client, processor 1000 is configured to invoke the program instructions, specifically performing the steps of:

determining the number of the companies of the offshore clients and/or the number of the companies of legal representatives of the offshore clients according to the updated equity penetration graph;

the number of establishments of the offshore customer and/or the number of establishments of the legal representative of the offshore customer is determined as the establishments data for the offshore customer.

In a specific implementation, the processor 1000 described in the embodiment of the present application may perform the implementation described in the embodiment of fig. 1 or the embodiment of fig. 2, or may perform the implementation described in the embodiment of the present application, which is not described herein again.

The functional modules in the embodiments of the present application may be integrated into one processing module, or each module may exist alone physically, or two or more modules may be integrated into one module. The integrated modules may be implemented in the form of sampling hardware or in the form of sampling software functional modules.

Those skilled in the art will appreciate that implementing all or part of the above-described methods in accordance with the embodiments may be accomplished by way of a computer program stored on a computer readable storage medium, which when executed may comprise the steps of the embodiments of the methods described above. Wherein the computer readable storage medium may be volatile or nonvolatile. For example, the computer storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM), or the like. The computer readable storage medium may mainly include a storage program area which may store an operating system, an application program required for at least one function, etc., and a storage data area which may store data created according to the use of the blockchain node, etc.

The blockchain is a novel application mode of computer technologies such as distributed data storage, point-to-point transmission, consensus mechanism, encryption algorithm and the like. The blockchain (Blockchain), essentially a de-centralized database, is a string of data blocks that are generated in association using cryptographic methods, each of which contains information from a batch of network transactions for verifying the validity (anti-counterfeit) of its information and generating the next block. The blockchain may include a blockchain underlying platform, a platform product services layer, an application services layer, and the like.

The above disclosure is only a preferred embodiment of the present application, and it should be understood that the scope of the application is not limited thereto, but all or part of the procedures for implementing the above embodiments can be modified by one skilled in the art according to the scope of the appended claims.

Claims

1. A method of data analysis, comprising:

Invoking an offshore customer classification model to classify the offshore customer according to the target transaction behavior feature set, the indirection data of the offshore customer and the registration management information of the offshore customer to obtain a class of the offshore customer;

The in-line data includes historical transaction behavior data, adjustment information and account opening information, the method comprises updating the equity penetration graph by using the in-line data, the first risk list and the second risk list to obtain an updated equity penetration graph, and the method comprises the following steps:

2. The method according to claim 1, wherein the method further comprises:

3. The method according to claim 2, wherein the method further comprises:

4. The method according to claim 1, wherein the method further comprises:

5. The method of claim 4, wherein said processing said historical transaction behavior data for each sample customer to obtain a set of transaction behavior characteristics for each sample customer comprises:

6. The method of claim 1, wherein the obtaining of the live data of the offshore customer comprises:

7. A data analysis device, comprising:

The classification module is used for calling an offshore client classification model to classify the offshore client according to the target transaction behavior feature set, the indirection data of the offshore client and the registration management information of the offshore client to obtain the category of the offshore client;

The in-line data comprise historical transaction behavior data, adjustment information and account opening information, and the updating module is specifically used for:

8. A computer device comprising a processor and a memory, the processor and the memory being interconnected, wherein the memory is adapted to store computer program instructions, the processor being configured to execute the program instructions to implement the method of any of claims 1-6.

9. A computer readable storage medium, characterized in that the computer readable storage medium has stored therein computer program instructions for performing the method according to any of claims 1-6 when being executed by a processor.