Security business system money laundering risk management and control method and device, and storage medium

Abstract

The invention discloses a money laundering risk management and control method and device for a security business system and a storage medium, and relates to the technical field of transaction risk monitoring. Information of different business systems is integrated into a customer information center according to metadata in customer information of the different business systems; therefore, the risk of the business system can be managed and controlled based on more abundant data. And then each transaction path is analyzed, so that all transaction processes in which each account participates are analyzed, the possibility that each account is a risk account is more accurately judged, and more accurate money laundering risk management and control on a security service system are further realized.

Description

Technical Field

[0001] The present invention relates to the technical field of transaction risk monitoring, and in particular to a money laundering risk control method and device, and a storage medium for a securities business system. Background Art

[0002] Securities business requires continuous money laundering risk control to ensure the security of securities business. In the prior art, the control of risks in securities business systems mainly focuses on the identification of risky transactions and risky accounts, including two methods: manual identification and identification using corresponding algorithms. However, the amount of data generated in the securities trading process is huge, and the money laundering risk control methods in the prior art are difficult to ensure accurate identification of risky accounts or risky transactions in the face of massive data. On the other hand, different business systems are involved in the securities trading process, such as different business systems generated by different trading platforms, different business systems generated by different trading links, and different business systems corresponding to different business products. In the face of transactions between different business systems, the prior art needs to analyze and identify each system separately, which is inefficient. Therefore, a new money laundering risk control method for securities business systems is needed, so that when it is necessary to integrate data from different securities business systems, money laundering risk control for securities business systems can be efficiently and accurately achieved. Summary of the invention

[0003] The present invention provides a securities business system money laundering risk control method and device, and a storage medium, which can integrate different securities business system data to efficiently and accurately implement the money laundering risk control of the securities business system.

[0004] In order to solve the above technical problems, the first aspect of the present invention discloses a money laundering risk management method for a securities business system, the method comprising: For each business system, obtain a customer information set corresponding to the business system, and determine metadata corresponding to the business system based on the customer information set; Building a customer information center based on the metadata corresponding to all the business systems, wherein the customer information center is used to record the relationship between the customer information in all the business systems; Analyze the customer information center using a pre-trained transaction path analysis model to obtain all transaction paths output by the transaction path analysis model; For each of the transaction paths, determining a risk weight for each of the customer information on the transaction path, wherein the risk weight is used to measure the possibility of money laundering transactions occurring in the corresponding customer information; For each piece of customer information in the customer information center, the risk coefficient corresponding to the customer information is determined according to the risk weight of the customer information on all corresponding transaction paths, and the risk coefficient is used to measure the possibility that the corresponding customer information is a risk account.

[0005] As an optional implementation manner, in the first aspect of the present invention, the customer information includes customer attribute information and customer transaction information, and the constructing of a customer information center according to the metadata corresponding to all the business systems includes: Generate an attribute AST abstract syntax tree according to the relationship between the customer attribute information represented in the metadata corresponding to all the business systems; the attribute AST abstract syntax tree is used to record the association relationship between the attributes of the customer information in all the business systems; Generate a transaction AST abstract syntax tree according to the relationship between the customer transaction information represented in the metadata corresponding to all the business systems; the transaction AST abstract syntax tree is used to record the transaction relationship between the customer information in all the business systems; A customer information center is constructed based on the attribute AST abstract syntax tree and the transaction AST abstract syntax tree.

[0006] As an optional implementation, in the first aspect of the present invention, the using of a pre-trained transaction path analysis model to analyze the customer information center to obtain all transaction paths output by the transaction path analysis model includes: The transaction AST abstract syntax tree is analyzed using a pre-trained transaction path analysis model to obtain all transaction syntax branches output by the transaction path analysis model, and all transaction paths are determined based on all the transaction syntax branches.

[0007] As an optional implementation manner, in the first aspect of the present invention, for each of the transaction paths, determining the risk weight of each of the customer information on the transaction path includes: For each of the transaction paths, determine all target customer information in the transaction syntax branch corresponding to the transaction path; determine the correlation degree parameter of each of the target customer information based on the correlation relationship between the corresponding attributes in the attribute AST abstract syntax tree of all the target customer information, and the correlation degree parameter is used to measure the closeness of the correlation relationship between the corresponding target customer information and all other target customers corresponding to the transaction path; based on the correlation degree parameter, determine the risk weight of each of the target customer information corresponding to the transaction path.

[0008] As an optional embodiment, in the first aspect of the present invention, the method further comprises: Determine whether the first transaction cycle is over. If it is determined that the first transaction cycle is over, supplement the relationship between the customer information not determined in the attribute AST abstract syntax tree according to all the customer transaction information in the first transaction cycle to obtain an attribute update AST abstract syntax tree; Determine whether to enter the second trading cycle. If it is determined that the second trading cycle has been entered, perform the following operations: According to the relationship between the customer attribute information represented in the metadata corresponding to all the business systems in the second transaction cycle, the relationship between the customer information not determined in the attribute update AST abstract syntax tree is supplemented to obtain a period attribute AST abstract syntax tree; Generate a periodic transaction AST abstract syntax tree according to the relationship between the customer transaction information represented in the metadata corresponding to all the business systems in the second transaction cycle; Based on the periodic attribute AST abstract syntax tree and the periodic transaction AST abstract syntax tree, a customer information center corresponding to the second transaction period is constructed.

[0009] As an optional implementation manner, in the first aspect of the present invention, constructing a customer information center according to the metadata corresponding to all the business systems includes: According to the relationship identifiers of the metadata corresponding to all the business systems, the mainstream relationship identifiers in all the business systems are determined; wherein the relationship identifier is information used to represent the relationship between data determined based on specific data rules, and the mainstream relationship identifier is the relationship identifier with the highest frequency of use in all the business systems; Generate an initial AST abstract syntax tree according to the mainstream relationship identifier; the initial AST abstract syntax tree is used to record the relationship between the customer information determined according to the mainstream relationship identifier; According to other relationship identifiers except the mainstream relationship identifier in the metadata corresponding to all the business systems, the relationship between the customer information not determined in the initial AST abstract syntax tree is supplemented to obtain a target AST abstract syntax tree, wherein the target AST abstract syntax tree is used to record the relationship between the customer information in all the business systems; A customer information center is constructed based on the target AST abstract syntax tree.

[0010] As an optional implementation, in the first aspect of the present invention, the using of a pre-trained transaction path analysis model to analyze the customer information center to obtain all transaction paths output by the transaction path analysis model includes: Determine the depth of all syntax branches in the target AST abstract syntax tree, traverse and parse each syntax branch of the AST abstract syntax tree in descending order of depth, and obtain a virtual table corresponding to each node on each syntax branch, wherein the virtual table includes one or more of table name information, column information, sub-virtual table, and table condition information; All the virtual tables are input into a pre-trained transaction path analysis model to obtain all transaction paths output by the transaction path analysis model.

[0011] The second aspect of the present invention discloses a money laundering risk control device for a securities business system, the device comprising: An information acquisition module, used to acquire, for each business system, a set of customer information corresponding to the business system, and determine metadata corresponding to the business system based on the set of customer information; A relationship building module, used to build a customer information center according to the metadata corresponding to all the business systems, and the customer information center is used to record the relationship between the customer information in all the business systems; A transaction analysis module, used to analyze the customer information center using a pre-trained transaction path analysis model to obtain all transaction paths output by the transaction path analysis model; A weight analysis module, for determining, for each of the transaction paths, a risk weight of each of the customer information on the transaction path, wherein the risk weight is used to measure the possibility of money laundering transactions occurring in the corresponding customer information; The risk analysis module is used to determine the risk coefficient corresponding to each piece of customer information in the customer information center according to the risk weight of the customer information on all corresponding transaction paths, and the risk coefficient is used to measure the possibility that the corresponding customer information is a risk account.

[0012] As an optional implementation, in the second aspect of the present invention, the customer information includes customer attribute information and customer transaction information, and the relationship building module builds a customer information center in a specific manner according to the metadata corresponding to all the business systems, including: Generate an attribute AST abstract syntax tree according to the relationship between the customer attribute information represented in the metadata corresponding to all the business systems; the attribute AST abstract syntax tree is used to record the association relationship between the attributes of the customer information in all the business systems; Generate a transaction AST abstract syntax tree according to the relationship between the customer transaction information represented in the metadata corresponding to all the business systems; the transaction AST abstract syntax tree is used to record the transaction relationship between the customer information in all the business systems; A customer information center is constructed based on the attribute AST abstract syntax tree and the transaction AST abstract syntax tree.

[0013] As an optional implementation, in the second aspect of the present invention, the transaction analysis module uses a pre-trained transaction path analysis model to analyze the customer information center, and a specific method of obtaining all transaction paths output by the transaction path analysis model includes: The transaction AST abstract syntax tree is analyzed using a pre-trained transaction path analysis model to obtain all transaction syntax branches output by the transaction path analysis model, and all transaction paths are determined based on all the transaction syntax branches.

[0014] As an optional implementation, in the second aspect of the present invention, the weight analysis module determines, for each transaction path, a specific method of determining the risk weight of each piece of customer information on the transaction path, including: For each of the transaction paths, determine all target customer information in the transaction syntax branch corresponding to the transaction path; determine the correlation degree parameter of each of the target customer information based on the correlation relationship between the corresponding attributes in the attribute AST abstract syntax tree of all the target customer information, and the correlation degree parameter is used to measure the closeness of the correlation relationship between the corresponding target customer information and all other target customers corresponding to the transaction path; based on the correlation degree parameter, determine the risk weight of each of the target customer information corresponding to the transaction path.

[0015] As an optional implementation, in the second aspect of the present invention, the device further includes: The attribute update module is used to determine whether the first transaction cycle is over. If it is determined that the first transaction cycle is over, the relationship between the customer information not determined in the attribute AST abstract syntax tree is supplemented according to all the customer transaction information in the first transaction cycle to obtain an attribute update AST abstract syntax tree; The periodic risk analysis module is used to determine whether to enter the second trading cycle. If it is determined that the second trading cycle has been entered, the following operations are performed: According to the relationship between the customer attribute information represented in the metadata corresponding to all the business systems in the second transaction cycle, the relationship between the customer information not determined in the attribute update AST abstract syntax tree is supplemented to obtain a period attribute AST abstract syntax tree; Generate a periodic transaction AST abstract syntax tree according to the relationship between the customer transaction information represented in the metadata corresponding to all the business systems in the second transaction cycle; Based on the periodic attribute AST abstract syntax tree and the periodic transaction AST abstract syntax tree, a customer information center corresponding to the second transaction period is constructed.

[0016] As an optional implementation, in the second aspect of the present invention, the specific manner in which the relationship building module builds the customer information center according to the metadata corresponding to all the business systems includes: According to the relationship identifiers of the metadata corresponding to all the business systems, the mainstream relationship identifiers in all the business systems are determined; wherein the relationship identifier is information used to represent the relationship between data determined based on specific data rules, and the mainstream relationship identifier is the relationship identifier with the highest frequency of use in all the business systems; Generate an initial AST abstract syntax tree according to the mainstream relationship identifier; the initial AST abstract syntax tree is used to record the relationship between the customer information determined according to the mainstream relationship identifier; According to other relationship identifiers except the mainstream relationship identifier in the metadata corresponding to all the business systems, the relationship between the customer information not determined in the initial AST abstract syntax tree is supplemented to obtain a target AST abstract syntax tree, wherein the target AST abstract syntax tree is used to record the relationship between the customer information in all the business systems; A customer information center is constructed based on the target AST abstract syntax tree.

[0017] As an optional implementation, in the second aspect of the present invention, the transaction analysis module uses a pre-trained transaction path analysis model to analyze the customer information center, and a specific method of obtaining all transaction paths output by the transaction path analysis model includes: Determine the depth of all syntax branches in the target AST abstract syntax tree, traverse and parse each syntax branch of the AST abstract syntax tree in descending order of depth, and obtain a virtual table corresponding to each node on each syntax branch, wherein the virtual table includes one or more of table name information, column information, sub-virtual table, and table condition information; All the virtual tables are input into a pre-trained transaction path analysis model to obtain all transaction paths output by the transaction path analysis model.

[0018] The third aspect of the present invention discloses another securities business system money laundering risk management and control system, the system comprising: A memory storing executable program code; a processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the money laundering risk management method for the securities business system disclosed in the first aspect of the present invention.

[0019] The fourth aspect of the present invention discloses a computer storage medium, which stores computer instructions. When the computer instructions are called, they are used to execute the money laundering risk management method for the securities business system disclosed in the first aspect of the present invention.

[0020] Compared with the prior art, the present invention has the following beneficial effects: For each business system, the present invention obtains the customer information set corresponding to the business system, and determines the metadata corresponding to the business system based on the customer information set; according to the metadata corresponding to all business systems, a customer information center is constructed, and the customer information center is used to record the relationship between customer information in all business systems; the customer information center is analyzed using a pre-trained transaction path analysis model to obtain all transaction paths output by the transaction path analysis model; for each transaction path, the risk weight of each customer information on the transaction path is determined, and the risk weight is used to measure the possibility of money laundering transactions in the corresponding customer information; for each customer information in the customer information center, the risk coefficient corresponding to the customer information is determined according to the risk weight of the customer information on all corresponding transaction paths, and the risk coefficient is used to measure the possibility that the corresponding customer information is a risk account. It can be seen that the present invention can combine the information of different business systems to comprehensively obtain the customer information center, so that the risk of the business system can be controlled based on more substantial data. In addition, the present invention analyzes each transaction path, thereby analyzing all transaction processes involved in each account, and more accurately judges the possibility of each account being a risk account, thereby realizing more accurate money laundering risk control of the securities business system. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0022] Figure 1 It is a flow chart of a method for controlling money laundering risks in a securities business system disclosed in an embodiment of the present invention; Figure 2 It is a flow chart of another method for controlling money laundering risks in a securities business system disclosed in an embodiment of the present invention; Figure 3 It is a structural schematic diagram of a money laundering risk control device for a securities business system disclosed in an embodiment of the present invention; Figure 4 It is a structural schematic diagram of another money laundering risk control device for a securities business system disclosed in an embodiment of the present invention; Figure 5 It is a structural schematic diagram of a money laundering risk management and control system for a securities business system disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0023] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0024] The terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish different objects rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, device, product or end including a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units that are not listed, or may optionally include other steps or units inherent to these processes, methods, products or ends.

[0025] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0026] The present invention discloses a money laundering risk control method and device for a securities business system, and a storage medium. The main inventive concept is: according to the metadata in the customer information of different business systems, the information of different business systems is integrated into a customer information center, and then by analyzing each transaction path, all transaction processes involved in each account are analyzed.

[0027] Embodiment 1 See also Figure 1 , Figure 1 This is a flow chart of a method for controlling money laundering risks in a securities business system disclosed in an embodiment of the present invention. Figure 1The described securities business system money laundering risk control method can be applied to a securities business system money laundering risk control device, and the securities business system money laundering risk control device can be integrated in a cloud server or a local server, which is not limited in the embodiments of the present invention. Figure 1 As shown, the money laundering risk management method of the securities business system may include the following operations: Step 101: For each business system, obtain a customer information set corresponding to the business system, and determine metadata corresponding to the business system based on the customer information set.

[0028] In the embodiment of the present invention, different business systems may be business systems generated at different stages of the securities trading process, business systems generated by different trading platforms, or databases corresponding to different departments in the same enterprise. For example, business systems may be divided according to a list of business products, where each subcategory corresponds to a business system. For another example, business systems may be divided according to a list of business modules, where each system platform corresponds to a business system.

[0029] Each business system stores a collection of customer information. How to integrate customer information distributed in different business systems and establish associations becomes an important prerequisite for determining subsequent money laundering risk management. In the embodiment of the present invention, customer information distributed in different business systems is integrated and associated by analyzing metadata.

[0030] Each database generates corresponding metadata during operation. Metadata is data about data. It provides descriptions, explanations and background information of the data to help people better understand, manage and use the data. Metadata is like a "navigation map" of data, enabling people to quickly find the data they need. For example, in a large database, metadata can be used to understand the names, field meanings, data types and other information of each data table, making it easier to determine which data meets specific query requirements. Among them, metadata has different types or forms of existence based on different metadata processing algorithms.

[0031] In an embodiment of the present invention, optionally, the metadata corresponding to the business system may include an ETL task configuration table corresponding to the target system, and the ETL task configuration table generally includes multiple fields, such as process ID, source file directory, target table, delimiter, character set, etc. These fields together define the path and processing logic of data flow. For example, the process ID is used to uniquely identify each ETL task; the source file directory and the target table specify the original location and loading destination of the data, respectively. ETL tasks support not only simple data handling, but also complex data conversion operations. For example, Flink SQL or other similar tools can be used to achieve rich data conversion and processing, such as JOIN operations, field calculations, etc. These operations can be implemented by specifying corresponding SQL commands or scripts in the ETL configuration table.

[0032] Step 102: Build a customer information center based on the metadata corresponding to all business systems.

[0033] In the embodiment of the present invention, the customer information center is used to record the relationship between customer information in all business systems. Customer information may include customer attribute information plus customer transaction information. Customers may be divided into individual customers, institutional customers, and product customers. The customer attribute information of the above three types of customers is shown in the following table:

[0034] In an embodiment of the present invention, the customer information center may be a customer information model, which includes the topological relationship between all customer information and the attribute information and customer transaction information corresponding to each customer information. In addition, the customer information center may be a graph structure or a tree structure. For example, the customer information center may be a constructed knowledge graph, wherein each node on the knowledge graph represents a customer information, and the relationship between each node identifies the relationship between customer information, including transaction relationship and other relationships. Through the customer information center, it is possible to integrate customer information sets in multiple business systems into the same data system, and record the relationship between all customer information.

[0035] Step 103: Use the pre-trained transaction path analysis model to analyze the customer information center to obtain all transaction paths output by the transaction path analysis model.

[0036] In the embodiment of the present invention, a transaction path refers to a transaction process between a series of customer information having a transaction relationship or other relationship, including the flow of securities resources. In the embodiment of the present invention, a transaction path analysis model is obtained using multiple sets of training data sets, each set of training data sets includes a customer information center and all transaction paths corresponding to the customer information center, and the transaction path analysis model is used to determine all transaction paths according to the customer information center.

[0037] Step 104: For each transaction path, determine the risk weight of each customer information on the transaction path.

[0038] In the embodiment of the present invention, the risk weight is used to measure the possibility of money laundering transactions in the corresponding customer information. For each transaction path, the risk weight of each customer can be determined according to the transaction path. The transaction path refers to the transaction process between a series of customer information with transaction relationships or other relationships, including the flow of securities resources. In the transaction path, customers participate in resource transactions that match them according to different attributes. If a certain customer information in a certain transaction path does not match its customer attribute information, or the matching degree is low, the customer information may be risky. In this embodiment, a larger risk weight is assigned to the customer information.

[0039] Step 105: For each piece of customer information in the customer information center, determine the risk coefficient corresponding to the customer information according to the risk weight of the customer information on all corresponding transaction paths.

[0040] In the embodiment of the present invention, the risk coefficient is used to measure the possibility that the corresponding customer information is a risk account, wherein the risk account may be an account suspected of money laundering. The embodiment of the present invention analyzes the situation of each customer information in all transaction paths and comprehensively analyzes the risk coefficient of the customer information. Optionally, all risk weights of a certain customer information are directly added together to determine the risk coefficient.

[0041] It can be seen that the implementation of the money laundering risk control method of the securities business system in the embodiment of the present invention can combine the information of different business systems to comprehensively obtain the customer information center, so that the risk of the business system can be controlled based on more substantial data. In addition, the embodiment of the present invention analyzes each transaction path, thereby analyzing all transaction processes involved in each account, and more accurately determines the possibility that each account is a risk account, thereby achieving more accurate money laundering risk control for the securities business system.

[0042] In an optional embodiment, the customer information includes customer attribute information and customer transaction information. According to the metadata corresponding to all business systems, a customer information center is constructed, which may include: Generate an attribute AST abstract syntax tree based on the relationship between customer attribute information represented in the metadata corresponding to all business systems; the attribute AST abstract syntax tree is used to record the association relationship between the attributes of customer information in all business systems; Generate a transaction AST abstract syntax tree based on the relationship between customer transaction information represented in the metadata corresponding to all business systems; the transaction AST abstract syntax tree is used to record the transaction relationship between customer information in all business systems; Build a customer information center based on the attribute AST abstract syntax tree and the transaction AST abstract syntax tree.

[0043] Generally, an AST (Abstract Syntax Tree) is a tree representation of source code, which represents the building blocks of a programming language in the form of a tree. Each node represents a construct in the source code, such as an expression, statement, or declaration. In an embodiment of the present invention, an AST (Abstract Syntax Tree) is used to record the relationship between customer information in all business systems. Each node represents a customer information, and the connection relationship between nodes can be used to represent the relationship between any two customer information. This relationship can be an association relationship between the attributes of the customer information, or a transaction relationship between the customer information. Through the AST (Abstract Syntax Tree), the relationship between customer information in all business systems can be clearly and efficiently measured.

[0044] This optional embodiment constructs two different AST abstract syntax trees from the two perspectives of customer attribute information and customer transaction information, respectively, to describe the relationship between customer information in all business systems in a more fine-grained manner, and finally constructs a customer information center based on the attribute AST abstract syntax tree and the transaction AST abstract syntax tree. Optionally, the customer information center can be directly composed of two syntax trees, the attribute AST abstract syntax tree and the transaction AST abstract syntax tree. When it is necessary to query the relationship between any two customer information, the relationship between the two customer information attributes and the relationship generated by the transaction can be queried from the attribute AST abstract syntax tree and the transaction AST abstract syntax tree respectively.

[0045] It can be seen that this optional embodiment can clearly and efficiently measure the relationship between customer information in all business systems through the AST abstract syntax tree; and construct two different AST abstract syntax trees from the two perspectives of customer attribute information and customer transaction information, respectively, to describe the relationship between customer information in all business systems in a more fine-grained manner.

[0046] In another optional embodiment, using a pre-trained transaction path analysis model to analyze the customer information center to obtain all transaction paths output by the transaction path analysis model may include: Use the pre-trained transaction path analysis model to analyze the transaction AST abstract syntax tree, obtain all transaction syntax branches output by the transaction path analysis model, and determine all transaction paths based on all transaction syntax branches.

[0047] The customer information center is analyzed using a pre-trained transaction path analysis model. However, the customer information center contains two types of relationships: attribute relationships and transaction relationships between customer information. The amount of computation required to analyze the complex relationships using the model is relatively large. To solve the above problem, in this optional embodiment, the object to be analyzed is the AST abstract syntax tree. Since the AST abstract syntax tree can clearly and efficiently measure the relationship between customer information in all business systems, it is more efficient to use the model to analyze the AST abstract syntax tree. On the other hand, this optional embodiment only needs to use a pre-trained transaction path analysis model to analyze the transaction AST abstract syntax tree, without considering the attribute relationship between customer information. Therefore, the amount of computation required in the transaction path determination process is reduced.

[0048] It can be seen that in this optional embodiment, the object to be analyzed is the AST abstract syntax tree, and using the model to analyze the AST abstract syntax tree can more efficiently determine the transaction path; and, it is only necessary to use a pre-trained transaction path analysis model to analyze the transaction AST abstract syntax tree, without considering the attribute relationship between customer information, thereby reducing the amount of computation in the transaction path determination process.

[0049] In yet another optional embodiment, for each transaction path, determining the risk weight of each customer information on the transaction path may include: For each transaction path, determine all target customer information in the transaction syntax branch corresponding to the transaction path; determine the correlation degree parameter of each target customer information based on the correlation relationship between the corresponding attributes of all target customer information in the attribute AST abstract syntax tree; based on the correlation degree parameter, determine the risk weight of each target customer information corresponding to the transaction path.

[0050] For normal securities market transactions, the transaction behavior of customers should correspond to their attributes, such as transactions between intermediaries and their customers, transactions between partners, transactions between customers who are interested in the same product, and transactions between customers with various possible connections. That is, the closer the attribute relationship of customer information, the more frequent the transactions, and the more normal such transactions. However, the existing technology can often only analyze the abnormality of transactions based on the attributes of a single customer, such as based on the customer's regional information, age information, etc., resulting in insufficient comprehensiveness of money laundering risk management.

[0051] To solve the above problems, in an embodiment of the present invention, for each transaction path, all target customer information in the transaction syntax branch corresponding to the transaction path is determined; according to the association relationship between the attributes corresponding to all target customer information in the attribute AST abstract syntax tree, the association degree parameter of each target customer information is determined, and the association degree parameter is used to measure the closeness of the association relationship between the corresponding target customer information and all other target customers corresponding to the transaction path. For example, the current transaction path corresponds to three target customer information A, B, and C. The three target customer information A, B, and C have a transaction relationship with each other in this transaction path. In the attribute AST abstract syntax tree, it is reflected that there is no relationship between A and B, there is less relationship between A and C, and there is more relationship between B and C. Therefore, it can be judged that the association degree parameter of C is larger, the association degree of B is second, and the association degree of A is the smallest.

[0052] In an embodiment of the present invention, the correlation degree parameter is used to measure the closeness of the correlation relationship between the corresponding target customer information and all other target customers corresponding to the transaction path. For example, the correlation degree parameter can be measured according to the number of relationships between the corresponding target customer information and all other target customers corresponding to the transaction path.

[0053] It can be seen that this optional embodiment analyzes each transaction path, determines the correlation degree parameter of each target customer information according to the correlation relationship between the corresponding attributes of all target customer information in the attribute AST abstract syntax tree; based on the correlation degree parameter, determines the risk weight of each target customer information corresponding to the transaction path, which can more efficiently determine the closeness of the relationship between each customer information and other customer information in the transaction path, and at the same time, can more reasonably evaluate the attribute relationship between customer information, reduce the occurrence of omissions in money laundering risk management, and thus improve the comprehensiveness of money laundering risk management.

[0054] In this optional embodiment, further optionally, determining the correlation degree parameter of each target customer information according to the correlation relationship between the attributes corresponding to all target customer information in the attribute AST abstract syntax tree may include: For customer attribute relationships, each customer attribute relationship has a different corresponding impact coefficient. For example, if the attribute relationship between two customer information is: located in the same city, the impact coefficient is 1; if the attribute relationship is: located in the same county, the impact coefficient is 2; if the attribute relationship is: belonging to the same company, the impact coefficient is 5; if the attribute relationship is: Party A and Party B, the impact coefficient is 10. The determination of the correlation degree parameter of target customer information A can be determined based on the number of attribute relationships between A and other target customers and the impact coefficient corresponding to the attribute relationship. In this way, the closeness of the relationship between each customer information and other customer information in the transaction path can be determined more reasonably.

[0055] In yet another optional embodiment, the method may further include: Determine whether the first transaction cycle is over. If it is determined that the first transaction cycle is over, supplement the relationship between the customer information not determined in the attribute AST abstract syntax tree according to all customer transaction information in the first transaction cycle to obtain an attribute updated AST abstract syntax tree; Determine whether to enter the second trading cycle. If it is determined that the second trading cycle has been entered, perform the following operations: According to the relationship between the customer attribute information represented in the metadata corresponding to all business systems in the second transaction cycle, the relationship between the customer information not determined in the attribute update AST abstract syntax tree is supplemented to obtain the period attribute AST abstract syntax tree; Generate a periodic transaction AST abstract syntax tree according to the relationship between the customer transaction information represented in the metadata corresponding to all business systems in the second transaction cycle; Based on the periodic attribute AST abstract syntax tree and the periodic transaction AST abstract syntax tree, a customer information center corresponding to the second transaction period is constructed.

[0056] The second trading cycle is the next trading cycle after the first trading cycle.

[0057] Securities business transactions have a fixed trading cycle, usually one day. The attribute relationship between customer information does not remain unchanged, but needs to be updated once in each trading cycle. In addition, the attribute relationship between customer information also needs to be updated based on the transaction situation in the historical trading cycle. For example, if two customer information that do not have any relationship in themselves continue to have legal transactions in the historical trading cycle, the attribute relationship between the two should be modified to a relationship with business connection, and it can no longer be considered that the two have no attribute relationship.

[0058] Therefore, in this optional embodiment, at the end of each transaction cycle, the relationship between the customer information not determined in the attribute AST abstract syntax tree is supplemented according to all customer transaction information in the first transaction cycle, and the attribute updated AST abstract syntax tree is obtained. In each subsequent transaction cycle, there is no need to rebuild the attribute AST abstract syntax tree. It is only necessary to update the attribute AST abstract syntax tree according to the newly obtained customer attribute information and the transaction information in the historical transaction cycle, which reduces the amount of calculation and improves efficiency.

[0059] In this optional embodiment, for each transaction cycle, the most accurate cycle attribute AST abstract syntax tree can be updated, which can provide more accurate and comprehensive basic data for subsequent money laundering risk management operations. For example, for the operations in the second transaction cycle, it can also include: Generate a periodic transaction AST abstract syntax tree according to the relationship between customer transaction information represented in the metadata corresponding to all business systems in the second transaction period; construct a periodic customer information center based on the attribute AST abstract syntax tree and the transaction AST abstract syntax tree; Use the pre-trained transaction path analysis model to analyze the periodic customer information center and obtain all periodic transaction paths output by the transaction path analysis model; For each cycle transaction path, determine the risk weight of each customer information on the cycle transaction path. The risk weight is used to measure the possibility of money laundering transactions in the corresponding customer information. For each customer information in the periodic customer information center, the periodic risk coefficient corresponding to the customer information is determined according to the risk weight of the customer information on all corresponding periodic transaction paths. The periodic risk coefficient is used to measure the possibility that the corresponding customer information is a risk account in the second transaction cycle.

[0060] It can be seen that in this optional embodiment, there is no need to rebuild the AST abstract syntax tree in each transaction cycle, and only the existing AST abstract syntax tree needs to be updated, which reduces the amount of calculation for money laundering risk management. In addition, for each transaction cycle, the attribute AST abstract syntax tree can be updated according to the newly obtained customer attribute information and the transaction information in the historical transaction cycle, so as to obtain the most accurate cycle attribute AST abstract syntax tree, providing more accurate and comprehensive basic data for subsequent money laundering risk management operations.

[0061] Embodiment 2 See also Figure 2 , Figure 2 This is a flow chart of another method for controlling money laundering risks in a securities business system disclosed in an embodiment of the present invention. The method for controlling money laundering risks in a securities business system may include the following operations: Step 201: For each business system, obtain a customer information set corresponding to the business system, and determine metadata corresponding to the business system based on the customer information set.

[0062] For a detailed description of step 201, please refer to the description of step 101, which will not be repeated in the embodiment of the present invention.

[0063] Step 202: Determine the mainstream relationship identifiers in all business systems based on the relationship identifiers of the metadata corresponding to all business systems.

[0064] In an embodiment of the present invention, a relationship identifier is information used to represent the relationship between data determined based on a specific data rule, and a mainstream relationship identifier is a relationship identifier with the highest frequency of use in all business systems. Among them, the relationship identifier can be a process ID, a source file directory, a target table, a delimiter, a character set, a one-taskid identifier, and an SQL identifier. For example, the target metadata can be a relation table, which is used to store the relationship between upstream and downstream tables between various systems; correspondingly, the one-taskid identifier (unique task identifier) ​​is the key information for associating these tables, and the first source table and the first target table of the target data can be determined through the one-taskid identifier. For another example, an SQL identifier is a text file containing a series of SQL statements, which is used to manage databases and perform database operations; SQL (Structured Query Language) is a standard language for communicating with relational databases, which allows users to query, update, delete and insert data, and manage database objects; through the SQL identifier, the second source table and the second target table of the target data can be determined. In an embodiment of the present invention, the mainstream relationship identifier is the relationship identifier with the highest frequency of use in all business systems. For example, the most frequently used relationship identifier can be determined by analyzing the number of metadata involving each relationship identifier in all metadata.

[0065] Step 203: Generate an initial AST abstract syntax tree according to the mainstream relationship identifier.

[0066] In the embodiment of the present invention, the initial AST abstract syntax tree is used to record the relationship between the customer information determined according to the mainstream relationship identifier.

[0067] In the first embodiment of the present invention, the attribute relationship and transaction relationship between customer information are described respectively by the attribute AST abstract syntax tree and the transaction AST abstract syntax tree. In some cases where the amount of data is not large or the accuracy requirement is not high, it is not necessary to describe them separately by two AST abstract syntax trees, and only a target AST abstract syntax tree needs to be established. Therefore, in some cases, there may be mainstream relationship identifiers in different business systems, that is, the most used relationship identifiers. For example, most of the databases corresponding to different departments in the same enterprise use the A relationship identifier, and only other relationship identifiers are used in special scenarios or departments. Therefore, at this time, the A relationship identifier can be used as the main one, and the customer information relationship recorded by other relationship identifiers can be used to supplement the first relationship identifier. At the same time, there are a large number of repeated relationships in the A relationship identifier and other relationship identifiers. At the same time, the AST abstract syntax tree is constructed according to the A relationship identifier and other relationship identifiers, which brings more repeated calculations and reduces efficiency. The embodiment of the present invention well reduces the occurrence of the above situation.

[0068] Step 204: supplement the relationships between the customer information not determined in the initial AST abstract syntax tree according to other relationship identifiers except the mainstream relationship identifiers in the metadata corresponding to all business systems, and obtain the target AST abstract syntax tree.

[0069] In the embodiment of the present invention, the target AST abstract syntax tree is used to record the relationship between customer information in all business systems; after the initial AST abstract syntax tree is constructed through the mainstream relationship identifier, the main framework of the customer information relationship is built, and many of the relationships are also recorded by other relationship identifiers. At this time, it is only necessary to supplement the customer relationships that are not determined on the initial AST abstract syntax tree according to other relationships other than the mainstream relationship. During the supplementation process, no operation is required for the relationships that have been determined by the mainstream relationship identifier, thereby improving the efficiency of the AST abstract syntax tree construction.

[0070] Step 205: Build a customer information center based on the target AST abstract syntax tree.

[0071] In an embodiment of the present invention, optionally, the customer information center can be directly composed of a target AST abstract syntax tree. When it is necessary to query the relationship between any two customer information, the relationship between the two customer information attributes and the relationship generated by the transaction can be queried in the target AST abstract syntax tree.

[0072] Step 206: Use the pre-trained transaction path analysis model to analyze the customer information center to obtain all transaction paths output by the transaction path analysis model.

[0073] Step 207: For each transaction path, determine the risk weight of each customer information on the transaction path.

[0074] Step 208: For each piece of customer information in the customer information center, determine the risk coefficient corresponding to the customer information according to the risk weight of the customer information on all corresponding transaction paths.

[0075] For the detailed description of steps 206-208, please refer to the description of steps 103-105, which will not be repeated in the embodiment of the present invention.

[0076] It can be seen that the implementation of the money laundering risk control method of the securities business system in the embodiment of the present invention, which describes the relationship between customer information through the AST abstract syntax tree, can more efficiently describe the relationship between customer information and is more convenient for subsequent money laundering risk control operations. In addition, in the embodiment of the present invention, the initial AST abstract syntax tree is first constructed based on the mainstream relationship identifier, and then the initial AST abstract syntax tree is supplemented and improved according to other relationship identifiers other than the mainstream relationship identifier, thereby obtaining the final target AST abstract syntax tree, which improves the efficiency of AST abstract syntax tree construction, and thus can control the risks of the business system based on more efficient data. In addition, the embodiment of the present invention analyzes each transaction path, thereby analyzing all transaction processes involved in each account, and more accurately judges the possibility of each account being a risk account, thereby achieving more accurate money laundering risk control for the securities business system.

[0077] In an optional embodiment, using a pre-trained transaction path analysis model to analyze the customer information center to obtain all transaction paths output by the transaction path analysis model may include: Determine the depth of all syntax branches in the target AST abstract syntax tree, traverse and parse each syntax branch of the AST abstract syntax tree in descending order of depth, and obtain the virtual table corresponding to each node on each syntax branch; Input all virtual tables into the pre-trained transaction path analysis model to obtain all transaction paths output by the transaction path analysis model.

[0078] In this optional embodiment, when parsing the AST abstract syntax tree, the depth-first principle is adopted. Since the syntax branches with large depth record more customer information relationships, the syntax branches with large depth are traversed first to obtain the virtual table corresponding to the main customer information relationship. The syntax tree with smaller depth is traversed later, and only the compensation and refinement operation is performed on the basis that the main virtual table has been determined, thereby improving the efficiency of AST abstract syntax tree parsing. Among them, the virtual table includes one or more of table name information, column information, sub-virtual table and table condition information, and the virtual table is used to record the relationship between the parsed customer information.

[0079] In this optional embodiment, specifically, the AST abstract syntax tree is traversed and parsed in a depth-first manner, and node preprocessing is performed first. At the same time, a stack structure is maintained, and the first traversed node is pushed into the stack. After all child nodes are parsed, the current node is finally parsed. In the AST abstract syntax tree parsing, the concept of virtual table is defined, and each virtual table contains table name information, column information, child virtual table, table condition information, etc. The outermost virtual table is the top virtual table. The information of each subquery is parsed and stored in a virtual table. The columns in the virtual table store the corresponding generated blood parent column list and conditional blood parent column list. Specifically, the generated blood parent column can be a column used to generate data in the select expression; the conditional blood parent column can be a column used in data generation conditions such as where / group by / join on. For example: for the SQL statement "select c1 co1from t1 where c2>1", t1.c1 is the generated blood parent column of col, and t1.c2 is the conditional blood parent column of co1. Then maintain a stack of virtual tables, and the top of the stack saves the parent table of the current virtual table. Each time a subquery is traversed, the current virtual table is pushed into the stack, a new virtual table corresponding to the subquery is created, and the new virtual table is updated to the current virtual table. Each time a subquery is exited, the parent virtual table is taken out of the stack, the current virtual table is added to the parent virtual table as a child table, or other related operations are performed, and then the parent virtual table is updated to the current virtual table. The above method can achieve more efficient parsing of the AST abstract syntax tree.

[0080] The virtual table records the relationship of customer information in the form of a table. Therefore, it will be more efficient and faster to use the model to parse its relationship. For example, the virtual tables can be input into the transaction path analysis model one by one. At the same time, the transaction path analysis model updates all transaction paths to be output according to the virtual tables, and finally obtains all transaction paths.

[0081] It can be seen that this optional embodiment first traverses deeper syntax branches during the process of parsing the AST abstract syntax tree, thereby improving the efficiency of AST abstract syntax tree parsing, thereby improving the efficiency of customer information relationship analysis. In addition, the form of the virtual table provides the transaction path analysis model with more efficient data to be parsed, thereby improving the efficiency of obtaining the transaction path.

[0082] Embodiment 3 See also Figure 3 , Figure 3 This is a schematic diagram of the structure of a money laundering risk control device for a securities business system disclosed in an embodiment of the present invention. Figure 3 As shown, the money laundering risk control device of the securities business system may include: The information acquisition module 301 is used to acquire, for each business system, a set of customer information corresponding to the business system, and determine metadata corresponding to the business system based on the set of customer information; A relationship building module 302 is used to build a customer information center based on metadata corresponding to all business systems. The customer information center is used to record the relationship between customer information in all business systems; The transaction analysis module 303 is used to analyze the customer information center using a pre-trained transaction path analysis model to obtain all transaction paths output by the transaction path analysis model; A weight analysis module 304 is used to determine the risk weight of each customer information on each transaction path, and the risk weight is used to measure the possibility of money laundering transactions occurring in the corresponding customer information; The risk analysis module 305 is used to determine the risk coefficient corresponding to each customer information in the customer information center according to the risk weight of the customer information on all corresponding transaction paths. The risk coefficient is used to measure the possibility that the corresponding customer information is a risk account.

[0083] It can be seen that the money laundering risk control device for the securities business system in the embodiment of the present invention can combine the information of different business systems to comprehensively obtain the customer information center, so that the risk of the business system can be controlled based on more substantial data. In addition, the embodiment of the present invention analyzes each transaction path, thereby analyzing all transaction processes involved in each account, and more accurately determines the possibility that each account is a risk account, thereby achieving more accurate money laundering risk control for the securities business system.

[0084] In an optional embodiment, the customer information includes customer attribute information and customer transaction information. The specific manner in which the relationship building module 302 builds the customer information center according to the metadata corresponding to all business systems may include: Generate an attribute AST abstract syntax tree based on the relationship between customer attribute information represented in the metadata corresponding to all business systems; the attribute AST abstract syntax tree is used to record the association relationship between the attributes of customer information in all business systems; Generate a transaction AST abstract syntax tree based on the relationship between customer transaction information represented in the metadata corresponding to all business systems; the transaction AST abstract syntax tree is used to record the transaction relationship between customer information in all business systems; Build a customer information center based on the attribute AST abstract syntax tree and the transaction AST abstract syntax tree.

[0085] It can be seen that this optional embodiment can clearly and efficiently measure the relationship between customer information in all business systems through the AST abstract syntax tree; and construct two different AST abstract syntax trees from the two perspectives of customer attribute information and customer transaction information, respectively, to describe the relationship between customer information in all business systems in a more fine-grained manner.

[0086] In another optional embodiment, the transaction analysis module 303 uses a pre-trained transaction path analysis model to analyze the customer information center, and the specific method of obtaining all transaction paths output by the transaction path analysis model may include: Use the pre-trained transaction path analysis model to analyze the transaction AST abstract syntax tree, obtain all transaction syntax branches output by the transaction path analysis model, and determine all transaction paths based on all transaction syntax branches.

[0087] It can be seen that in this optional embodiment, the object to be analyzed is the AST abstract syntax tree, and using the model to analyze the AST abstract syntax tree can more efficiently determine the transaction path; and, it is only necessary to use a pre-trained transaction path analysis model to analyze the transaction AST abstract syntax tree, without considering the attribute relationship between customer information, thereby reducing the amount of computation in the transaction path determination process.

[0088] In another optional embodiment, the specific method for the weight analysis module 304 to determine the risk weight of each customer information on each transaction path may include: For each transaction path, determine all target customer information in the transaction syntax branch corresponding to the transaction path; determine the correlation degree parameter of each target customer information based on the correlation relationship between the corresponding attributes in the attribute AST abstract syntax tree of all target customer information, and the correlation degree parameter is used to measure the closeness of the correlation relationship between the corresponding target customer information and all other target customers corresponding to the transaction path; based on the correlation degree parameter, determine the risk weight of each target customer information corresponding to the transaction path.

[0089] It can be seen that this optional embodiment analyzes each transaction path, determines the correlation degree parameter of each target customer information according to the correlation relationship between the corresponding attributes of all target customer information in the attribute AST abstract syntax tree; based on the correlation degree parameter, determines the risk weight of each target customer information corresponding to the transaction path, which can more efficiently determine the closeness of the relationship between each customer information and other customer information in the transaction path, and at the same time, can more reasonably evaluate the attribute relationship between customer information, reduce the occurrence of omissions in money laundering risk management, and thus improve the comprehensiveness of money laundering risk management.

[0090] In yet another optional embodiment, Figure 4 As shown, the device may also include: The attribute update module 306 is used to determine whether the first transaction cycle is over. If it is determined that the first transaction cycle is over, the relationship between the customer information not determined in the attribute AST abstract syntax tree is supplemented according to all customer transaction information in the first transaction cycle to obtain an attribute updated AST abstract syntax tree; The periodic risk analysis module 307 is used to determine whether to enter the second trading cycle. If it is determined that the second trading cycle has been entered, the following operations are performed: According to the relationship between the customer attribute information represented in the metadata corresponding to all business systems in the second transaction cycle, the relationship between the customer information not determined in the attribute update AST abstract syntax tree is supplemented to obtain the period attribute AST abstract syntax tree; Generate a periodic transaction AST abstract syntax tree according to the relationship between the customer transaction information represented in the metadata corresponding to all business systems in the second transaction cycle; Based on the periodic attribute AST abstract syntax tree and the periodic transaction AST abstract syntax tree, a customer information center corresponding to the second transaction period is constructed.

[0091] It can be seen that in this optional embodiment, there is no need to rebuild the AST abstract syntax tree in each transaction cycle, and only the existing AST abstract syntax tree needs to be updated, which reduces the amount of calculation for money laundering risk management. In addition, for each transaction cycle, the attribute AST abstract syntax tree can be updated according to the newly obtained customer attribute information and the transaction information in the historical transaction cycle, so as to obtain the most accurate cycle attribute AST abstract syntax tree, providing more accurate and comprehensive basic data for subsequent money laundering risk management operations.

[0092] In another optional embodiment, the specific manner in which the relationship building module 302 builds the customer information center according to the metadata corresponding to all business systems may include: According to the relationship identifiers of the metadata corresponding to all business systems, the mainstream relationship identifiers in all business systems are determined; wherein the relationship identifier is information used to represent the relationship between data determined based on specific data rules, and the mainstream relationship identifier is the relationship identifier with the highest frequency of use in all business systems; Generate an initial AST abstract syntax tree according to the mainstream relationship identifier; the initial AST abstract syntax tree is used to record the relationship between the customer information determined according to the mainstream relationship identifier; According to other relationship identifiers except mainstream relationship identifiers in the metadata corresponding to all business systems, the relationship between customer information not determined in the initial AST abstract syntax tree is supplemented to obtain a target AST abstract syntax tree, which is used to record the relationship between customer information in all business systems; Build a customer information center based on the target AST abstract syntax tree.

[0093] It can be seen that this optional embodiment first constructs an initial AST abstract syntax tree based on mainstream relationship identifiers, and then supplements and improves the initial AST abstract syntax tree based on other relationship identifiers except the mainstream relationship identifiers, so as to obtain the final target AST abstract syntax tree, which improves the efficiency of AST abstract syntax tree construction and enables the risk of business systems to be managed based on more efficient data.

[0094] In another optional embodiment, the transaction analysis module 303 uses a pre-trained transaction path analysis model to analyze the customer information center, and the specific method of obtaining all transaction paths output by the transaction path analysis model may include: Determine the depth of all syntax branches in the target AST abstract syntax tree, traverse and parse each syntax branch of the AST abstract syntax tree in descending order of depth, and obtain a virtual table corresponding to each node on each syntax branch, where the virtual table includes one or more of table name information, column information, sub-virtual table, and table condition information; Input all virtual tables into the pre-trained transaction path analysis model to obtain all transaction paths output by the transaction path analysis model.

[0095] It can be seen that this optional embodiment first traverses deeper syntax branches during the process of parsing the AST abstract syntax tree, thereby improving the efficiency of AST abstract syntax tree parsing, thereby improving the efficiency of customer information relationship analysis. In addition, the form of the virtual table provides the transaction path analysis model with more efficient data to be parsed, thereby improving the efficiency of obtaining the transaction path.

[0096] Embodiment 4 See also Figure 5 , Figure 5 Schematic diagram of the structure of a money laundering risk control system for a securities business system disclosed in an embodiment of the present invention. Figure 5 As shown, the money laundering risk management and control system of the securities business system may include: A memory 401 storing executable program codes; a processor 402 coupled to the memory 401; The processor 402 calls the executable program code stored in the memory 401 to execute the steps of the money laundering risk control method of the securities business system described in the first embodiment of the present invention.

[0097] Embodiment 5 An embodiment of the present invention discloses a computer storage medium, which stores computer instructions. When the computer instructions are called, they are used to execute the steps in the money laundering risk control method of the securities business system described in the first or second embodiment of the present invention.

[0098] Embodiment 6 An embodiment of the present invention discloses a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program, and the computer program is operable to enable a computer to execute the steps of the money laundering risk control method for a securities business system described in Embodiment 1 or Embodiment 2.

[0099] The device embodiments described above are only illustrative, wherein the modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules, i.e., they may be located in one place, or they may be distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Those of ordinary skill in the art may understand and implement it without creative work.

[0100] Through the specific description of the above embodiments, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on such an understanding, the above technical solution can be essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, and the storage medium includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable programmable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0101] Finally, it should be noted that the money laundering risk control method and device for a securities business system and the storage medium disclosed in the embodiments of the present invention are only preferred embodiments of the present invention, which are only used to illustrate the technical solution of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, it should be understood by those skilled in the art that the technical solutions described in the aforementioned embodiments can still be modified, or some of the technical features therein can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for controlling money laundering risks in a securities business system, characterized in that: The method comprises: For each business system, obtain a customer information set corresponding to the business system, and determine metadata corresponding to the business system based on the customer information set; Building a customer information center based on the metadata corresponding to all the business systems, wherein the customer information center is used to record the relationship between the customer information in all the business systems; Analyze the customer information center using a pre-trained transaction path analysis model to obtain all transaction paths output by the transaction path analysis model; For each of the transaction paths, determining a risk weight for each of the customer information on the transaction path, wherein the risk weight is used to measure the possibility of money laundering transactions occurring in the corresponding customer information; For each piece of customer information in the customer information center, the risk coefficient corresponding to the customer information is determined according to the risk weight of the customer information on all corresponding transaction paths, and the risk coefficient is used to measure the possibility that the corresponding customer information is a risk account.

2. The method for controlling money laundering risks in a securities business system according to claim 1, characterized in that: The customer information includes customer attribute information and customer transaction information. The customer information center is constructed according to the metadata corresponding to all the business systems, including: Generate an attribute AST abstract syntax tree according to the relationship between the customer attribute information represented in the metadata corresponding to all the business systems; the attribute AST abstract syntax tree is used to record the association relationship between the attributes of the customer information in all the business systems; Generate a transaction AST abstract syntax tree according to the relationship between the customer transaction information represented in the metadata corresponding to all the business systems; the transaction AST abstract syntax tree is used to record the transaction relationship between the customer information in all the business systems; A customer information center is constructed based on the attribute AST abstract syntax tree and the transaction AST abstract syntax tree.

3. The method for controlling money laundering risks in a securities business system according to claim 2, characterized in that: The using of the pre-trained transaction path analysis model to analyze the customer information center to obtain all transaction paths output by the transaction path analysis model includes: The transaction AST abstract syntax tree is analyzed using a pre-trained transaction path analysis model to obtain all transaction syntax branches output by the transaction path analysis model, and all transaction paths are determined based on all the transaction syntax branches.

4. The method for controlling money laundering risks in a securities business system according to claim 3, characterized in that: The step of determining, for each of the transaction paths, the risk weight of each piece of customer information on the transaction path includes: For each of the transaction paths, determine all target customer information in the transaction syntax branch corresponding to the transaction path; determine the correlation degree parameter of each of the target customer information based on the correlation relationship between the corresponding attributes in the attribute AST abstract syntax tree of all the target customer information, and the correlation degree parameter is used to measure the closeness of the correlation relationship between the corresponding target customer information and all other target customers corresponding to the transaction path; based on the correlation degree parameter, determine the risk weight of each of the target customer information corresponding to the transaction path.

5. The method for controlling money laundering risks in a securities business system according to claim 3, characterized in that: The method further comprises: Determine whether the first transaction cycle is over. If it is determined that the first transaction cycle is over, supplement the relationship between the customer information not determined in the attribute AST abstract syntax tree according to all the customer transaction information in the first transaction cycle to obtain an attribute update AST abstract syntax tree; Determine whether to enter the second trading cycle. If it is determined that the second trading cycle has been entered, perform the following operations: According to the relationship between the customer attribute information represented in the metadata corresponding to all the business systems in the second transaction cycle, the relationship between the customer information not determined in the attribute update AST abstract syntax tree is supplemented to obtain a period attribute AST abstract syntax tree; Generate a periodic transaction AST abstract syntax tree according to the relationship between the customer transaction information represented in the metadata corresponding to all the business systems in the second transaction cycle; Based on the periodic attribute AST abstract syntax tree and the periodic transaction AST abstract syntax tree, a customer information center corresponding to the second transaction period is constructed.

6. The method for controlling money laundering risks in a securities business system according to claim 1, characterized in that: The step of constructing a customer information center according to the metadata corresponding to all the business systems includes: According to the relationship identifiers of the metadata corresponding to all the business systems, the mainstream relationship identifiers in all the business systems are determined; wherein the relationship identifier is information used to represent the relationship between data determined based on specific data rules, and the mainstream relationship identifier is the relationship identifier with the highest frequency of use in all the business systems; Generate an initial AST abstract syntax tree according to the mainstream relationship identifier; the initial AST abstract syntax tree is used to record the relationship between the customer information determined according to the mainstream relationship identifier; According to other relationship identifiers except the mainstream relationship identifier in the metadata corresponding to all the business systems, the relationship between the customer information not determined in the initial AST abstract syntax tree is supplemented to obtain a target AST abstract syntax tree, wherein the target AST abstract syntax tree is used to record the relationship between the customer information in all the business systems; A customer information center is constructed based on the target AST abstract syntax tree.

7. The method for controlling money laundering risks in a securities business system according to claim 6, characterized in that: The using of the pre-trained transaction path analysis model to analyze the customer information center to obtain all transaction paths output by the transaction path analysis model includes: Determine the depth of all syntax branches in the target AST abstract syntax tree, traverse and parse each syntax branch of the AST abstract syntax tree in descending order of depth, and obtain a virtual table corresponding to each node on each syntax branch, wherein the virtual table includes one or more of table name information, column information, sub-virtual table, and table condition information; All the virtual tables are input into a pre-trained transaction path analysis model to obtain all transaction paths output by the transaction path analysis model.

8. A money laundering risk control device for a securities business system, characterized in that: The device comprises: An information acquisition module, used to acquire, for each business system, a set of customer information corresponding to the business system, and determine metadata corresponding to the business system based on the set of customer information; A relationship building module, used to build a customer information center according to the metadata corresponding to all the business systems, and the customer information center is used to record the relationship between the customer information in all the business systems; A transaction analysis module, used to analyze the customer information center using a pre-trained transaction path analysis model to obtain all transaction paths output by the transaction path analysis model; A weight analysis module, for determining, for each of the transaction paths, a risk weight of each of the customer information on the transaction path, wherein the risk weight is used to measure the possibility of money laundering transactions occurring in the corresponding customer information; The risk analysis module is used to determine the risk coefficient corresponding to each piece of customer information in the customer information center according to the risk weight of the customer information on all corresponding transaction paths, and the risk coefficient is used to measure the possibility that the corresponding customer information is a risk account.

9. A money laundering risk management and control system for a securities business system, characterized in that: The system includes: a memory storing executable program code; a processor coupled to the memory; the processor calls the executable program code stored in the memory to execute the money laundering risk control method for the securities business system as described in any one of claims 1-7.

10. A computer storage medium, characterized in that: The computer storage medium stores computer instructions, which, when called, are used to execute the money laundering risk management method for a securities business system as described in any one of claims 1 to 7.

Images