CN117974302A

CN117974302A - Transaction processing method, device, computer equipment and storage medium

Info

Publication number: CN117974302A
Application number: CN202211314985.6A
Authority: CN
Inventors: 张博海; 周美旭; 王波
Original assignee: Tencent Technology Shenzhen Co Ltd
Current assignee: Tencent Technology Shenzhen Co Ltd
Priority date: 2022-10-24
Filing date: 2022-10-24
Publication date: 2024-05-03

Abstract

The embodiment of the invention discloses a transaction processing method, a transaction processing device, computer equipment and a storage medium, wherein the method comprises the following steps: acquiring transaction description information of a target transaction to be executed, wherein the transaction description information comprises description information of at least two transaction objects related to the target transaction; acquiring target graph data of each transaction object in the target transaction according to the transaction description information; the target graph data comprises one or more nodes and connecting edges between the nodes; any node is used for storing object attributes of corresponding transaction objects; according to the object attributes in the target graph data and the corresponding association relations, analyzing and processing the target graph data to obtain shared object attributes among all transaction objects in the target graph data; based on the shared object attribute, the execution risk of the target transaction is determined, and the accuracy of transaction discrimination can be effectively improved under the condition that the computing resource of the equipment is ensured.

Description

Transaction processing method, device, computer equipment and storage medium

Technical Field

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

Background

With the continuous and deep development of computer technology, the daily production and life of people assisted by computer technology has become the development trend of current computer technology, and currently, in order to assist in judging transaction risks, a mode is generally adopted for judging and processing corresponding transactions based on transaction related data acquired offline, such as object data of transaction objects, and the like. The practice shows that the method for assisting in discriminating the transaction based on the offline collected data has the problems of higher processing pressure and lower accuracy of the computer equipment. Therefore, under the condition of ensuring equipment computing resources in a transaction scene, the accuracy of judging the transaction is improved, and the method becomes a current research hotspot.

Disclosure of Invention

The embodiment of the invention provides a transaction processing method, a transaction processing device, computer equipment and a storage medium, which can effectively improve the accuracy of transaction discrimination under the condition of ensuring the computing resources of the equipment.

In one aspect, an embodiment of the present invention provides a transaction processing method, including:

acquiring transaction description information of a target transaction to be executed, wherein the transaction description information comprises description information of at least two transaction objects related to the target transaction;

Acquiring target graph data of each transaction object in the target transaction according to the transaction description information; the target graph data comprises one or more nodes and connecting edges between the nodes; any node is used for storing object attributes of corresponding transaction objects; any one of the connection edges is used for describing the association relation between the object attributes in the two nodes which are correspondingly connected;

according to the object attributes in the target graph data and the corresponding association relations, analyzing and processing the target graph data to obtain shared object attributes among all transaction objects in the target graph data;

Based on the shared object attributes, a risk of execution of the target transaction is determined.

In still another aspect, an embodiment of the present invention provides a transaction processing apparatus, including:

An obtaining unit, configured to obtain transaction description information of a target transaction to be executed, where the transaction description information includes description information of at least two transaction objects related to the target transaction;

The acquisition unit is further used for acquiring target graph data of all transaction objects in the target transaction according to the transaction description information; the target graph data comprises one or more nodes and connecting edges between the nodes; any node is used for storing object attributes of corresponding transaction objects; any one of the connection edges is used for describing the association relation between the object attributes in the two nodes which are correspondingly connected;

The processing unit is used for analyzing and processing the target graph data according to the object attributes and the corresponding association relations in the target graph data to obtain the shared object attributes among all transaction objects in the target graph data;

the processing unit is further configured to determine an execution risk of the target transaction based on the shared object attribute.

In yet another aspect, an embodiment of the present invention provides a computer device, including a processor, an input device, an output device, and a memory, where the processor, the input device, the output device, and the memory are connected to each other, and the memory is configured to store a computer program that supports the computer device to perform the above method, where the computer program includes program instructions, and where the processor is configured to invoke the program instructions to perform the following steps:

In yet another aspect, an embodiment of the present invention provides a computer readable storage medium having stored therein program instructions, which when executed by a processor, are for executing the transaction processing method according to the first aspect.

In the embodiment of the application, the computer equipment can acquire the transaction description information of the target transaction in the execution process of the target transaction, so that the target graph data among all transaction objects related to the target transaction can be acquired from the graph database based on the acquisition of the transaction description information, and after the target graph data is acquired, the target graph data can be analyzed and processed based on the object attribute and the association relation stored in the target graph data to acquire the shared object attribute among all the transaction objects of the target transaction, so that the computer equipment can realize the real-time analysis of all the transaction objects in the target transaction based on graph calculation, and the accuracy and the reliability of the analysis of all the transaction objects in the target transaction to be executed currently by the computer equipment can be effectively improved based on the real-time analysis of all the transaction objects in the target transaction. After the computer equipment analyzes the shared object attribute of the target transaction related to the transaction object, the execution risk of the target transaction can be determined based on the shared object attribute, so that the flexibility of the computer equipment in judging the transaction risk is improved, and the accuracy of judging the transaction risk based on the shared object attribute is ensured because the shared object attribute is determined in real time in the target transaction process.

Drawings

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

FIG. 1a is a schematic diagram of graph data provided by an embodiment of the present invention;

FIG. 1b is a schematic diagram of a transaction processing system according to an embodiment of the present invention;

FIG. 2 is a schematic flow chart of a transaction processing method provided by an embodiment of the present invention;

FIG. 3a is a schematic diagram of determining a shared object attribute according to an embodiment of the present invention;

FIG. 3b is a schematic diagram of a computing platform according to an embodiment of the present invention;

FIG. 3c is a code schematic diagram of a graph data model according to an embodiment of the present invention;

FIG. 4 is a schematic flow chart of a transaction processing method provided by an embodiment of the invention;

FIG. 5a is a schematic diagram of graph data provided by an embodiment of the present invention;

FIG. 5b is a schematic diagram of performing graph data calculation according to an embodiment of the present invention;

FIG. 5c is a schematic diagram of a policy tree corresponding to a policy according to an embodiment of the present invention;

FIG. 5d is a flow chart of a transaction process provided by an embodiment of the present invention;

FIG. 6 is a schematic block diagram of a transaction processing device provided in an embodiment of the present invention;

fig. 7 is a schematic block diagram of a computer device provided by an embodiment of the present invention.

Detailed Description

The embodiment of the application provides a transaction processing method, which enables computer equipment to acquire the description information of a transaction object related to a target transaction in the execution process of the target transaction, so that the computer equipment can acquire target graph data related to each transaction object in the target transaction from a graph database based on the acquisition of the description information of the related transaction object, further enables the computer equipment to analyze the shared object attribute among each transaction object related to the target transaction from the target graph data based on the object attribute stored in each node in the target graph data and the association relation among the object attributes in the corresponding nodes indicated by the connecting edges in the target graph data, and enables the computer equipment to determine the execution risk of the target transaction by adopting the determined shared object attribute based on the determination of the shared object attribute, thereby enabling the computer equipment to realize real-time discrimination of the transaction execution risk in the transaction execution process, being beneficial to improving the real-time performance and effectiveness of the discrimination of the transaction risk and being beneficial to improving the reliability of the execution of the transaction, and further guaranteeing the reliability of the corresponding transaction. The target transaction refers to any transaction currently being executed and executed depending on the computer device, in one embodiment, the computer device may support execution of one or more transactions at the same time, and the computer device may be a terminal device or a server, which is not limited in the embodiment of the present application, and when the computer device is a server, the server may be an independent server formed by one physical structure or a server cluster formed by at least two servers, and in addition, the server may also be a cloud server.

In one embodiment, the transaction object related to the target transaction refers to an object related to the target transaction, and may specifically include one or more of the following: an initiating object of a target transaction, a receiving object of a target transaction, one or more initiating associated objects related to the initiating object, one or more receiving associated objects associated with the receiving object, wherein the initiating associated object (or receiving object) is, for example, an object that has a historical electronic resource transfer with the initiating object, or an object that has a relationship, a friendship, and other statutory relationships, etc. The description information (i.e., object description information) of the corresponding transaction object refers to information describing various features of the corresponding object, where the features of the corresponding object may include one or more of behavior features, portrait features, trip features, transaction features, and the like of the corresponding object, and in one embodiment, the behavior features are used to describe behavior habits of the corresponding object, such as network access habits, resource transfer habits, and the like of the corresponding object, the portrait features are used to describe basic properties of the corresponding object, such as resource conditions, basic identity information, and the like of the corresponding object, the trip features are features used to describe trip conditions of the corresponding object, and the transaction features are features used to describe resource transfer conditions of the corresponding object. In one embodiment, if the target transaction is that the object a transfers 10 electronic resources to the object B, the transaction object related to the target transaction may include the object a and the object B, and may further include an associated object of the object a and/or an associated object of the object B, etc., further, the description information about the transaction object acquired by the computer device may be one or more of a behavior feature, a transaction feature, and an portrait feature of the corresponding object (such as the object a described above).

The graph database accessed by the computer device when acquiring the description information of the corresponding objects is a database which stores data in the form of a graph, wherein the graph (or graph data) is a data structure for representing the association relationship between different objects, the graph is described by using a vertex (or called a node, and can be represented by V) and an edge (or called a connected edge, and can be represented by E), the vertex represents an object (i.e. an object attribute for storing the corresponding object), the edge represents the association relationship between the objects (i.e. the association relationship between the corresponding objects of the corresponding vertex for indicating connection), in addition, the connected edge contained in the graph data can be the connected edge without the pointing relationship, or can also be the connected edge with the pointing relationship, wherein when the edge in the graph data is the edge without the pointing relationship, the association relationship described by the edge is mutually established between the two connected nodes, and when the edge in the graph data is the edge with the pointing relationship, the association relationship representing the edge connection is established based on the pointing relationship, wherein when the edge in the graph data is the edge with the pointing relationship, the graph data is represented by the vertex data of the graph= (V) and the graph data is represented by the graph data of the graph 1, if the graph data is represented by the graph data of the graph 1. In one embodiment, in order to avoid information security of object information related to an object (such as description information about the object), when the graph database stores information in the form of a graph (such as object attributes stored in vertices or association relationships stored in a borderline), the computer device may encrypt corresponding information first to obtain corresponding encrypted information and store the corresponding encrypted information.

In one embodiment, the target graph data acquired by the computer device is graph data related to the transaction object, that is, the target graph data acquired by the computer device is graph data including nodes and a continuous edge structure, based on graph data stored in the graph database, further, the computer device can analyze and process the target graph data based on object attributes included in each node in the target graph data to obtain a shared object attribute of each transaction object indicated by the target transaction, and since the shared object attribute is obtained based on graph data analysis and is related to attribute characteristics of each transaction party, after obtaining the shared object attribute, the computer device can determine the execution risk of the target transaction based on the shared object attribute. The shared object attribute refers to an attribute related to each transaction object related to the target transaction, and the execution risk determined by the computer device is used for indicating whether any transaction object related to the corresponding transaction generates electronic resource loss after the corresponding transaction is executed. Based on the determination of the transaction execution risk, the computer device may output corresponding risk prompt information to each transaction object involved in the target transaction based on the determined execution risk, so that the corresponding transaction object may determine whether to continue to execute the target transaction based on the output risk prompt information, or the computer device may also directly trigger to stop subsequent execution of the target transaction when the determined target transaction has a transaction risk, and send corresponding risk prompt information and transaction stop information to each transaction object.

The transaction processing method provided by the embodiment of the application can be applied to a transaction processing system as shown in fig. 1b, wherein the transaction processing system comprises an object device 10 of at least two transaction objects and a computer device 11 for supporting the execution of the transaction, and the computer device 11 can analyze and extract related transaction object graph data based on the target transaction in the execution process of the at least two transaction objects through corresponding object devices to obtain the shared object attribute of the related transaction objects, so that the computer device 11 can determine the association degree of the related transaction objects under a transaction scene based on the shared object attribute, and can determine the execution risk of the target transaction based on the association degree. The computer equipment can acquire the association characteristics related to the transaction objects of all parties based on the analysis process of the graph data, and the accuracy and the effectiveness of the transaction risk discrimination of the computer equipment can be improved based on the interpretability and the better distinguishing property of the association characteristics related to the transaction objects of all parties.

Referring to fig. 2, a schematic flow chart of a transaction processing method according to an embodiment of the application is shown, wherein the transaction processing method may be executed by the above-mentioned computer device, and as shown in fig. 2, the method may include:

S201, acquiring transaction description information of a target transaction to be executed, wherein the transaction description information comprises description information of at least two transaction objects related to the target transaction.

S202, acquiring target graph data of each transaction object in target transaction according to transaction description information; the target graph data comprises one or more nodes and connecting edges between the nodes; any node is used for storing object attributes of corresponding transaction objects; any one of the links is used to describe the association between the object properties in the two nodes of the respective connection.

In step S201 and step S202, the computer device may acquire the transaction description information of the target transaction during the execution of the target transaction, where, since the computer device is a computer device including an air control engine and a graph computing platform, and the air control engine is used for performing transaction detection on a transaction to be executed in the computer device, that is, for acquiring, in real time, the transaction to be executed implemented based on the support of the computer device, then the air control engine may take any transaction to be executed as the target transaction when the computer device has the transaction to be executed, and perform the acquisition of the transaction description information based on the target transaction. In one embodiment, the obtained transaction description information is description information for indicating a transaction object related to the target transaction, and after the computer device obtains the transaction description information, target graph data of the transaction object related to the target transaction can be obtained based on the transaction description information. The obtained transaction description information may be an object identifier or an object name of the related transaction object.

The computer equipment is used for acquiring target graph data of a transaction object, the graph computing platform mainly comprises a graph data supporting platform, a graph data importing platform, a graph data storing platform and a graph data application platform from bottom to top, wherein the graph data supporting platform comprises main functions of graph computing platform bottom layer authority management, cluster management, graph data schema management (summary management), data deletion and verification and the like, the graph data importing platform is connected with the graph data storing platform, it can be understood that the graph data importing platform is used for importing graph data to be stored into the graph data storing platform, the graph data storing platform is used for storing the graph data, and in addition, the graph data application platform can be used for executing operations of deletion and verification and the like on the stored graph data based on graph query language (or graph query service). In one embodiment, the computer device performs the detection of the transaction during the execution of the transaction, and the specific architecture of the graph computing platform included in the computer device may be shown in fig. 3a, and based on the specific architecture of the graph computing platform, the computer device may access the graph computing platform based on the graph query service to achieve the acquisition of the target graph data related to the transaction object from the graph computing platform.

In one embodiment, the main role of the graph computing platform of the computer device in the execution flow of the embodiment of the present application is to store the graph data of the total transaction objects, and feed back the graph data of the transaction objects (such as at least two transaction objects related to the target transaction) to be queried for the computer device based on the graph query service, where the graph computing platform performs the process of storing the graph data and feeding back the required graph data based on the query service may be as shown in fig. 3 b. Specifically, the map data stored by the map computing platform includes offline map data and real-time map data, that is, the map data maintained in the map data storage platform included by the map computing platform includes an offline map database and a real-time map database, and in one embodiment, the real-time map database updates the data in the real-time map database based on a detected real-time event, where the real-time event may be, for example, a friend adding event implemented by an object based on the computer device, and then, based on execution of the friend adding event, the corresponding map writing engine updates the real-time map database based on the event. The real-time event refers to an event that a time interval between an occurrence time of the time and a current time is smaller than an interval threshold, the interval threshold may be, for example, 2 minutes, 5 minutes, or 10 minutes, and the current time refers to a time when an execution risk of the target transaction needs to be determined in an execution process of the target transaction, that is, if the occurrence time of the friend adding event is within the interval threshold range from the current time, the computer device may determine that the friend adding event is a real-time event, or else determine that the friend adding time is not a real-time, so that effective update of a real-time graph database may be realized when the execution risk of the target transaction is determined.

When the computer device updates the real-time graph database based on the real-time event, the real-time event may be used to update the real-time graph database when the computer device determines that the amount of updates required to update the graph data in the real-time graph database based on the real-time event is small. In one embodiment, when a real-time event occurs, the computer device further updates an offline graph database based on the real-time event, and the graph data contained in the offline graph database is written into the real-time graph database based on the writing service of the offline event processing, so as to ensure that the object attribute stored in the graph data contained in the real-time graph database at the current moment is accurate and up-to-date attribute information. Because the updating of the real-time graph database by the computer equipment is performed when the updating amount triggered by the real-time event is smaller, the computer equipment can only trigger the updating of the offline graph database when the updating amount triggered by the real-time event is larger, thereby realizing the updating of the real-time graph database based on the updating of the offline graph database, reducing the frequent disturbance to the real-time graph database, and improving the data stability of the real-time graph database while ensuring the accuracy and the effectiveness of the real-time graph database. The graph data contained in the real-time graph database and the offline graph database of the graph computing platform may be described and stored based on a graph data model, which may be as shown in fig. 3 c.

In one embodiment, the graph data stored in the real-time graph database and the offline graph database of the graph computing platform are databases for storing object attributes of related transaction objects in the form of graphs, wherein one node of the graph data can be used for storing one or more object attributes of the related transaction objects, and a connecting edge of the node can be used for describing association relations between object attributes stored in two connected nodes or association relations between corresponding objects of the two connected nodes. The object attribute is an object feature, and in order to obtain the object feature of each object, the object attribute is obtained through four stages of data mining, data preprocessing, feature engine processing and feature selection. The distributed high-performance large data storage computing platform can be constructed by means of the existing open source tool in the data mining stage so as to bear the splicing of high-dimensional features and the high-performance parallel computing work, so that various features of the object, such as age-layer features, academic-layer features and the like, can be extracted from basic data and daily behaviors of the corresponding object, and suspicious behaviors of the object in the transaction process can be estimated. In the data preprocessing stage, the original object data for executing data mining money can be processed by adopting modes of data cleaning, data integration, data conversion, data reduction and the like so as to improve the data mining quality, and in the feature engine processing stage, feature data which is adapted to an offline platform can be generated and stored in a feature table according to feature engineering projects (such as engineering projects related to transactions) configured by the object, and feature formats of the mined features can be determined. In addition, the feature selection stage is to select features with larger influence on the transaction from the mined features, so that the finally selected mined features are the object features (namely object attributes) of the corresponding objects, and the corresponding object attributes are stored in the corresponding nodes.

In one embodiment, as shown in fig. 3b, the computer device accesses the real-time graph database in the graph computing platform to obtain the target graph data of the transaction object, and obtains the target graph data of the related object, and after the computer device obtains the target graph data of the transaction object related to the target transaction, the target graph data can be analyzed based on the object attribute and the corresponding association relationship contained in the target graph data to obtain the shared object attribute of the transaction object related to the target transaction, that is, step S203 is performed.

S203, according to the object attributes and the corresponding association relations in the target graph data, analyzing and processing the target graph data to obtain the shared object attributes among all the transaction objects in the target graph data.

S204, determining the execution risk of the target transaction based on the shared object attribute.

After the target graph data is obtained in step S202 and step S203, the computer device may analyze the target graph data according to the object attribute and the association relationship of the target graph data, and in one embodiment, the process of obtaining and analyzing the target graph data may also be understood as a process of performing real-time graph calculation, where the graph calculation is a calculation manner of modeling data according to a graph manner and calculating a data structure in the graph to obtain a corresponding result, and based on the data structure including nodes and edges, it may be understood that the graph calculation is a calculation method of taking the graph as a data model and performing association calculation on points/edges in the graph, and then the graph calculation platform (or the graph calculation system) is used for solving the problem related to performing the graph calculation. Then, based on the computer device acquiring the target graph data related to the transaction object in the execution process of the target transaction, the computer device can perform association calculation on the target graph data based on the node and the continuous edge contained in the target graph data, and since the node contained in the target graph data is used for storing the object attribute and the continuous edge contained in the target graph data is used for storing the corresponding association relation, the computer device performs association calculation on the target graph data, namely, performs association calculation on the object attribute in the target graph data.

In one embodiment, since one object attribute in the target graph data is associated with a corresponding transaction object, the computer device performs an association calculation on the object attribute in the target graph data, that is, determines the tightness degree between the transaction objects of the corresponding object attribute, wherein the computer device may obtain the shared object attribute between the transaction objects in the target graph data after performing an analysis processing on the target graph data, that is, may determine the shared object attribute belonging to each different transaction object from the object attribute contained in the target graph data based on the analysis processing on the target graph data by the computer device, and may determine the tightness degree (intimacy degree) between the corresponding transaction objects based on the determination on the shared object attribute, and may determine the execution risk of the target transaction based on the determined tightness degree between the objects.

In one embodiment, after the computer device obtains the shared object attribute of each transaction object based on the analysis processing of the target graph data, the computer device may further determine the total number and type of the shared object attributes, so as to combine the category of the corresponding shared object attribute and the determined total number of the shared object attributes to obtain the tightness degree between the corresponding transaction objects, and further determine the execution risk of the target transaction based on the tightness degree. In general, the computer device may determine that the more compact the respective transaction objects are when the total number of determined shared object attributes is greater, and that the less compact the respective transaction objects are when the total number of determined shared object attributes is less; in addition, when the computer device determines the object tightness based on the determined type of the shared object attribute, the computer device may determine that the tightness between the corresponding transaction objects is stronger when the correlation between the determined type of the shared object attribute and the transaction type is higher, and otherwise, determine that the tightness before the corresponding transaction objects is lower.

Since the target transaction involves a transaction object comprising at least two, the computer device, when determining the shared object properties between the respective transaction objects, may determine the shared object properties between any two transaction objects to obtain a degree of tightness between the two transaction objects. In one embodiment, because the resource circulation related to the target transaction is mainly generated between the resource transfer-out object and the resource transfer-in object, when the computer equipment determines the attribute of the shared object based on the target graph data, the computer equipment only determines the attribute of the shared object between the object related to the resource transfer-out object and the object related to the resource transfer-in object based on the target graph data, so that the computing pressure of the computer equipment can be reduced, and the efficiency and the effectiveness of judging the transaction execution risk of the computer equipment can be improved. Specifically, if the transaction object related to the target transaction determined by the computer device includes an object a, an object B and an object C, and the object related to the target transaction is the object a, the object related to the transfer of the resource is the object B, the computer device may determine only the shared object attribute between the object a and the object B, without referring to the shared object attribute between other objects, when analyzing the corresponding target graph data and determining the shared object attribute, so as to effectively improve the processing efficiency of the computer device.

Based on the determined shared object attribute, the computer device, when determining the execution risk of the target transaction, may determine that the target transaction is not at risk of execution when the determined shared object attribute indicates that the degree of tightness between the respective objects is greater than a degree threshold, and may determine that the target transaction is at risk of execution when the determined degree of tightness is less than the degree threshold. When the computer equipment determines that the corresponding target transaction has execution risk, the risk prompt can be output, and the transaction termination is triggered, so that the security and reliability of the transaction object for carrying out the transaction based on the computer equipment are ensured.

Referring to fig. 4, a schematic flow chart of a transaction processing method according to an embodiment of the present application may also be applied to the computer device, where the transaction processing method mainly includes four processes of a data preparation phase, a feature calculation phase, a feature verification phase and a feature application phase, and referring to fig. 4, the method may specifically include:

S401, acquiring transaction description information of a target transaction to be executed, wherein the transaction description information comprises description information of at least two transaction objects related to the target transaction.

S402, acquiring target graph data of each transaction object in the target transaction according to the transaction description information; the target graph data comprises one or more nodes and connecting edges between the nodes; any node is used for storing object attributes of corresponding transaction objects; any one of the links is used to describe the association between the object properties in the two nodes of the respective connection.

In step S401 and step S402, the computer device acquires the target graph data from the stored graph data, which is the graph data contained in the real-time graph database stored in the graph computing platform, when acquiring the target graph data according to the transaction description information, and the graph data contained in the real-time graph database is determined in the data preparation stage. In one embodiment, in the data preparation stage, a graph structure design is performed based on an original scene (such as a transaction scene in the embodiment of the present application), where the graph structure design is used to determine a storage structure and a storage attribute of a node (and/or an edge) included in the real-time graph database when performing graph data storage, so that after determining the designed graph structure, the determined graph structure is deployed and online into the graph computing platform, so that the real-time graph database included in the graph computing platform performs graph data storage according to the determined storage structure and the determined storage attribute. In addition, the computer device accesses and acquires the real-time graph database based on the graph query service when acquiring the target graph data from the real-time graph database contained in the graph computing platform, so that in the data preparation stage, the data result corresponding to the graph data returned based on the graph query service can be predefined based on the query language of the graph computing platform.

In one embodiment, since the map data included in the real-time map database includes map data transmitted from the offline map database, the computer device specifically includes an offline map data preparation process and an offline map data database delivery process when preparing the real-time map database in the data preparation stage. The offline diagram data preparation process mainly comprises the following steps: since the graph can be represented by g= (V, E), typical offline graph data preparation is to prepare point data and the number of edges, and the prepared point data can be represented as: v= (V ₁,V₂…V_n), the prepared edge data may be expressed as: e= { V ₁,V₂),(V₁,V₃),…,(V_i,V_n) }, where n, i are positive integers greater than 1, and i < n, and V _i represents the i-th point data (i.e., the data stored by the i-th node), (V _i,V_n) represents the continuous edge connecting the i-th point data and the n-th point data, wherein the prepared point data and the map data can be shown in fig. 5 a. Based on the data structure of the offline map data determined in the offline map data preparation process, the computer device can execute the offline map data database-out process after determining that the map data of the corresponding data structure is stored and backed up, wherein the offline map data database-out is based on the real-time calculation requirement of the map data, the prepared offline map data is imported into a map data import layer based on a map calculation platform to carry out data database-out, namely the offline map data is pushed to a map database storage platform, so that the offline map data prepared in the offline map data process is sent to a real-time map database for storage. In one embodiment, when the offline map database (and/or the real-time map database) stores data (such as storing object attributes), a Key-value (KV) storage manner is generally adopted, that is, a Key value (Key) for storing corresponding data is generally used in a node of the corresponding map database, and real data is stored in other storage positions associated with the Key value, so that the database can also be called a KV database based on the storage manner of the KV storage of the database, and the KV storage manner of the corresponding database based on the KV storage manner of the computer device, so that the computer device can effectively reduce the data amount of the database and improve the storage efficiency of the corresponding database.

Based on the preparation of the real-time graph database in the data preparation stage, when the computer equipment acquires the target graph data of each transaction object in the target transaction according to the transaction description information, the computer equipment can firstly determine the reference node corresponding to any transaction object in at least two transaction objects indicated by the transaction description information from the stored graph data (namely the graph data contained in the real-time graph database), further can determine the association node with the connection relation with the reference node from the stored graph data according to the association relation described by the continuous edge of the stored graph data and combines the reference node, and takes the graph data formed by the reference node corresponding to each transaction object, the association node and the corresponding connection relation as the target graph data. In one embodiment, based on a graph database determined by a computer device in a data preparation stage, if the computer device determines that a transaction object related to a target transaction includes a transaction object a and a transaction object B, the computer device further obtains other reference nodes having connection relation with the node of the transaction object a or the node of the transaction object B after determining a data storage node of the transaction object a and a data storage node of the transaction object B from the graph database, and based on the obtaining of the reference nodes, the computer device can obtain more attribute data related to the corresponding transaction object, so that the comprehensiveness and accuracy of description of the target transaction by the target graph data obtained by the computer device can be improved, and further, the reliability of risk analysis and discrimination based on the target graph data can be effectively improved.

That is, the target graph data is acquired from a graph database belonging to a graph data storage platform in the graph computing platform and used for storing graph data of the total transaction objects for performing transaction execution based on the graph computing platform; the object attributes contained in the graph data of each transaction object stored in the graph database are encryption attributes stored after encrypting the object attributes of the corresponding transaction object. In addition, the graph computing platform further comprises a graph data importing platform, the graph database stored by the graph data storage platform is a real-time graph database, and the graph database further comprises a graph departure database; the map data import platform is used for sending the offline map data stored in the offline map database to the real-time map database, so that the map database for obtaining the target map data is the real-time map database containing the offline map data.

S403, analyzing and processing the target graph data according to the object attributes and the corresponding association relations in the target graph data to obtain the shared object attributes among all the transaction objects in the target graph data.

After the computer device obtains the target graph data of each transaction object in the target transaction, graph calculation analysis can be performed on the obtained target graph data in the feature calculation stage to obtain the shared object attribute among each transaction object in the target transaction, that is, the feature calculation stage triggers the feature calculation process of the predefined graph calculation platform based on the transaction original request, and generates the predefined graph calculation feature (that is, the shared object attribute). In one embodiment, the computer device, when executing the feature calculation phase, will also first perform the design process of the real-time graph features. In one embodiment, since at least two transaction objects of a transaction are generally involved in a transaction scenario, the graph calculation paradigm is based on calculation of graph data of the transaction scenario, and may be a double-bond departure graph calculation paradigm, that is, query and calculation of object attributes are simultaneously performed on payment objects (i.e., resource roll-out objects) and collection objects (i.e., resource receiving objects) in a transaction, so as to obtain the number and types of shared object attributes, and the obtained graph features (i.e., the obtained shared object attributes) may be deployed after offline testing or online gray-scale testing. The types familiar to the shared object include: a zone type, a work type, a device type, etc.

That is, the shared object attribute determined by the computer device is obtained based on the query and calculation of the object attribute, and when the computer device analyzes and processes the object graph data according to the object attribute and the corresponding association relationship in the object graph data to obtain the shared object attribute between all the transaction objects in the object graph data, one or more attribute types of the object attribute contained in the object graph data and the number of the object attributes contained in each attribute type can be determined according to the object attribute stored in the object graph data and the corresponding association relationship; then, based on the number of the object attributes contained in each attribute type, determining the attribute type of which the number of the object attributes correspondingly contained in one or more attribute types meets a number threshold as a shared attribute type; further, the computer device may use the object attribute of each transaction object in the target graph data as the shared object attribute of each transaction object in the object attributes included in the shared attribute type.

The target graph data acquired by the computer device may be graph data as shown in fig. 1a based on the mode of designing the graph calculation paradigm into the double-bond departure graph calculation paradigm, and when the target graph data is subjected to graph calculation processing based on the double-bond departure graph calculation paradigm to obtain the corresponding shared object attribute, the aggregation calculation is performed on the node V _i,V_j existing in the target graph data based on the double-bond departure graph calculation paradigm. Specifically, the computer device may analyze the target graph data according to the object attribute and the corresponding association relationship in the target graph data, when obtaining the shared object attribute between each transaction object in the target graph data, first obtain a storage mode of each node of the target graph data for the object attribute of the corresponding transaction object, obtain a graph calculation rule related to the storage mode based on the storage mode, and may analyze the target graph data by combining the graph calculation rule with the object attribute stored in the target graph data and the corresponding association relationship after obtaining the graph calculation rule related to the storage mode, to obtain the shared object attribute between each transaction object in the target graph data. The storage mode of each node of the target graph data is that any node of the target graph data is used for storing one object attribute of a corresponding transaction object, and in this case, the graph calculation rule corresponding to the storage mode obtained by the computer device is an attribute calculation rule, and the attribute calculation rule can be shown in formula 1.

Wherein V _ifi represents the fi object attribute stored at the Vi node, V _jfi represents the fi object attribute stored at the Vj node, hash () represents hash calculation of the object attribute stored at the node, n represents the total number of object attributes of each transaction object under a certain attribute type, and f (V _i,V_j) represents the shared object attribute. Then, by the equation 1, when one node of the target graph data stores one object attribute, it is determined whether the object attributes stored by the corresponding node are equal according to whether the hash values of the object attributes stored by the respective nodes are equal, and when the object attributes stored by the corresponding node are equal, it is determined that the object attributes are shared object attributes, and based on the equation 1, the number of shared object attributes of each transaction object under a certain type can be calculated. Specifically, the computer equipment analyzes and processes the target graph data by adopting an attribute calculation rule and combining object attributes stored in the target graph data with corresponding association relations, and the mode of obtaining the shared object attributes among all transaction objects in the target graph data is that first object attributes (namely object attributes stored in any node) are selected from the target graph data according to the object attributes and the association relations stored in the target graph data; wherein the first object attribute belongs to a first shared attribute type and is associated with a first transaction object in a target transaction, the target transaction also comprises other transaction objects, and the other transaction objects comprise one or more second object attributes under the first shared attribute type; further, the computer device may compare the first object attribute with any one of the second object attributes (i.e. the object attributes stored in the other nodes), respectively, and when the comparison result indicates that the first object attribute is the same as the corresponding second object attribute, the same object attribute is used as one shared object attribute in the first shared attribute type.

Specifically, if the computer device determines that the object attribute of the transaction object a corresponding to the attribute type of the target transaction is a location type and the object attribute of the transaction object B corresponding to the attribute type of the target transaction is a location type contains 5, and the object attribute of the transaction object B corresponding to the attribute type of the target transaction is 7, the computer device may compare the object attribute of the transaction object a corresponding to the 5 location types with the object attribute of the transaction object B corresponding to the 7 location types, and if it is determined that 3 locations are the same, the obtained shared object attribute contains a location, and the corresponding value is 3, and the expression represented by equation 1 may be f (V _i,V_j)＝(V_i =3).

In one embodiment, the second method for storing the corresponding object by each node of the target graph data is that any node of the target graph data is used for storing all object attributes of the corresponding transaction object, and then the graph calculation rule corresponding to the storage method obtained by the computer device is a neighbor heterogeneous node calculation rule, where the neighbor heterogeneous node calculation rule may be as shown in formula 2.

Where V _i and V _j represent two nodes in the target graph data, N () represents a neighbor heterogeneous algorithm, and N represents the total number of object attributes of each transaction object in a type where the shared attribute type is shared by neighboring nodes. That is, when any node of the target graph data stores all object attributes of the corresponding transaction object, the determined type of the shared object attribute may include a type shared by neighboring nodes, and further, when the computer determines that the shared attribute type is the number of object attributes included in the type shared by the neighboring nodes by adopting the rule of calculating the neighboring heterogeneous nodes, the computer may determine, from the target graph data, a reference node having a connection relationship with a node corresponding to any transaction object according to the transaction object indicated by the object attribute stored in the target graph data and the association relationship stored in the target graph data; further, the number of nodes of the reference node may be obtained, and the number of nodes of the reference node may be used as the number of object attributes included in any transaction object under the type shared by the neighboring nodes.

In one embodiment, after the computer device is designed based on the real-time graph features, the real-time graph features can be triggered to calculate through the real-time graph features in the policy engine, and the result returned by calculation is used as a discrimination condition of the policy node. Wherein the computer device performs the query and calculation process of the transaction object attribute based on the real-time graph database as shown in fig. 5 b. In another implementation manner, after the computer device obtains the shared object attribute between each transaction object in the target graph data, the execution risk of the target transaction may be further determined based on the shared object attribute, that is, the subsequent feature verification stage and the feature application stage are executed, and step S404 is executed instead. In addition, it should be noted that, the object attribute stored in the target graph data obtained by the computer device may be an encrypted object attribute, and then the shared object attribute obtained by analyzing and processing the target graph data is the shared encrypted object attribute (or shared encrypted physical attribute), and based on the obtaining of the related information of the corresponding transaction and the obtaining of the object attribute of the corresponding transaction object by the computer device in the transaction process, the computer device outputs the obtaining authorization information of the related object attribute to the related transaction object in advance, and obtains the attribute information of the corresponding transaction object after obtaining the authorization confirmation of the corresponding transaction object, so as to ensure the information security of the corresponding object.

S404, acquiring a wind control strategy related to the target transaction, wherein the wind control strategy comprises one or more strategy nodes, and any strategy node comprises a distinguishing feature and a corresponding distinguishing condition.

S405, determining a target discrimination feature matched with the shared object attribute and a discrimination condition corresponding to the target discrimination feature from discrimination features included in the strategy node of the wind control strategy.

S406, determining the execution risk of the target transaction based on the shared object attribute and the corresponding discrimination conditions.

In step S404 to step S406, after obtaining the shared object attribute of each transaction object, when performing risk discrimination based on the shared object attribute, the computer device may first obtain a policy for wind control related to the target transaction, where the policy broadly refers to a policy or a strategy, in the wind control field, the policy (rule) is a set of preset risk control conditions and a set of follow-up actions, and represents logic for judging and executing by the wind control system, and the object controls the service risk based on setting control conditions and a follow-up trigger action type. In one embodiment, in the field of wind control, the strategy (rule) is used as a risk identification and control method, and has the characteristics of intuitiveness, good interpretability and the like. In general, a complex wind control system needs to build a series of policy combinations, and policies can be visualized through a tree structure, and typical policies include root nodes (start nodes), internal nodes (feature judgment conditions), leaf nodes (end nodes); the quality of the strategy directly influences the quality of the wind control system, the quality index of the strategy can be embodied through the accuracy and the coverage rate, and the key for determining the quality of the strategy is the judging condition of each characteristic of the internal nodes of the strategy. In a transaction wind control system, the characteristic judgment conditions with good discrimination can greatly reduce the interference of a strategy on normal transactions and mainly intercept abnormal transactions, and the embodiment of the application can obtain the shared object attribute with strong interpretation and good discrimination through the graph calculation of target graph data and takes the shared object attribute as a strategy judgment node to assist in strategy processing. In one embodiment, the tree structure of each node included in a typical policy for performing policy discrimination may be as shown in fig. 5c, and then the tree structure based on the policy may implement performing risk discrimination.

In one embodiment, the obtaining of the wind control policy by the computer device is obtaining a tree structure of the policy, and the policy node included in the tree structure corresponding to the policy includes a distinguishing feature, such as feature a or feature B in fig. 5c, and a distinguishing condition including a corresponding distinguishing feature, such as shown in fig. 5c, where the distinguishing condition corresponding to feature a is greater than a threshold 1. Based on the acquisition of the distinguishing features and distinguishing conditions in the wind control strategy by the computer equipment, the determined shared object attribute can be used as a strategy feature, and the target distinguishing features and distinguishing conditions matched with the strategy feature are determined from the tree structure of the wind control strategy, so that the shared object attribute serving as the strategy feature can be distinguished based on the selected distinguishing conditions. Wherein the policy feature is a component of the policy, belongs to the minimum part of the policy preset wind control, and the shared object attribute (namely the shared encryption physical attribute feature) is a special graph calculation feature and is generated based on the wind control graph calculation system, so that the computer equipment can measure the trusted information of the transaction pair based on the number by counting the number of the shared encryption physical attributes of the transaction pair.

In one embodiment, since the number of the shared object attributes determined by the computer device is one or more, when the computer device determines the execution risk of the target transaction based on the shared object attributes and the corresponding discrimination conditions, the computer device may determine the object relationship between the target transaction and the transaction objects according to each shared object attribute and the corresponding discrimination conditions; when the number of the shared object attributes meeting the corresponding discrimination conditions is greater than or equal to the preset number, determining that the object relationship between the target transaction related transaction objects is a close relationship (namely, the relationship that the tightness degree between the objects reaches a degree threshold), otherwise, determining that the object relationship is a non-close relationship; further, the computer device may determine that the target transaction is not at risk of execution when the determined object relationship is affinity, and otherwise determine that the target transaction is at risk of execution.

In one embodiment, based on the shared object attribute determined by the computer device, the computer device may further perform verification processing on feasibility of using the corresponding shared object attribute as a policy feature in a feature verification stage, where the computer device may use the shared object attribute as a policy feature (or a graph computation feature), and deploy the obtained graph computation feature into a test environment for test verification, and may further calculate based on positive and negative samples to obtain a distinguishing effect of the graph computation feature on the positive and negative samples. Based on the acquisition of the positive and negative samples, the computer equipment can acquire a certain number of positive and negative samples from the online transaction floor log, and determine the accuracy and coverage rate of the graph calculation characteristics in risk judgment in a risk strategy based on the positive and negative samples. In one embodiment, the positive samples are transaction data for which a floor transaction is determined to be risk-free, and the negative samples are transaction data for which a floor transaction is determined to be risk-free. That is, the computer device may first obtain transaction log data when performing feasibility discrimination on the determined shared object attribute as a policy feature; the transaction log data comprises one or more executed transaction data and corresponding risk labels, the risk labels are used for indicating whether the corresponding executed transaction data has execution risks, further, the determined shared object attributes can be adopted to determine the execution risk of any executed transaction data contained in the transaction log data, the corresponding accuracy of the corresponding shared object attributes in risk judgment is obtained based on the execution risk of any executed transaction data and the corresponding risk labels, and risk judgment application is carried out on the shared object attributes based on the accuracy.

In one embodiment, the number of the determined shared object attributes is one or more, the execution risk of any executed transaction data contained in the transaction log data is determined by part or all of the one or more shared object attributes, and the determined accuracy of performing the execution risk determination includes the accuracy of performing the risk determination on any shared object attribute, so when the computer device performs the risk determination on the shared object attribute based on the accuracy, the computer device may select the shared object attribute with the accuracy reaching the accuracy threshold from the one or more shared object attributes, and perform the risk determination application on the selected shared object attribute. That is, based on the feasibility discrimination of the shared object attribute in the feature verification stage, the shared object attribute meeting the requirements (i.e. the policy accuracy and coverage rate described above) can be deployed as a graph feature to the online policy for application, so as to assist the policy in determining the risk behavior. The accuracy and coverage rate of risk discrimination by taking a certain shared object attribute determined by the computer community device as a policy feature under different risk categories can be shown in table 1.

TABLE 1

Based on the above explanation, in the process of designing a policy, a computer device needs to combine important features and models, the policy design needs to consider the distinguishing effect of the features and models on positive and negative samples, and the quality of the policy features can be measured by combining accuracy and recall, meanwhile, since the policy is mainly set manually, the features also need to have a certain interpretability, so the computer device can design real-time transaction pair sharing object attributes as policy features based on a real-time graph computing platform and graph data, wherein the computing and application process of the sharing object features based on real-time graph computation can be shown in fig. 5 d. The method comprises an offline diagram data preparation stage, an offline diagram data database output stage, a real-time diagram computing platform construction, a real-time diagram computing design, a real-time diagram computing feature calculation, feature effect verification and strategy application, wherein the specific real-time modes of the stages can be seen from the specific description in the above-mentioned figures 2 and 4.

In the embodiment of the application, the computer equipment can acquire target graph data of a target transaction related to a transaction object in the execution process of the target transaction so as to obtain the object attribute of the corresponding transaction object and the association relation between different object attributes, wherein the object attribute can be understood as an object feature, and then, the time delay of feature acquisition can be divided into two major types of offline features and real-time features, wherein the offline features are features obtained through offline calculation, the real-time features are features obtained through real-time calculation of the transaction, and because the time ductility requirement of the transaction is higher, most of the real-time features are single features mainly due to higher requirements on the calculation stability and time delay of the real-time features, namely, the related features of a payer or a payee in the calculation transaction; the real-time graph calculation feature is also one of the real-time features, so that the computer equipment can trigger the output transaction pair association relationship feature through both the transaction pair resource transfer-out object and the resource transfer-in object simultaneously by introducing the association relationship of the graph, and the real-time features (namely object attributes shared by both sides) with strong interpretability and good distinction between normal transaction and abnormal transaction are obtained, so that the interpretability of the features determined by the computer equipment is improved. Based on the determination of the attribute of the shared object by the computer equipment, the computer equipment can also determine the transaction execution risk based on the attribute of the shared object, judge the feasibility of the determined attribute of the shared object as a risk policy feature, obtain the accuracy and feasibility of the attribute of the corresponding shared object when judging the policy feature, and further improve the flexibility and reliability of the computer equipment for generating the wind control policy.

Based on the above description of the embodiments of the transaction processing method, the embodiments of the present invention also provide a transaction processing apparatus, which may be a computer program (including program code) running in the above-mentioned computer device. The transaction processing device may be used to perform the transaction processing method as described in fig. 2 and 4, referring to fig. 6, the transaction processing device includes: an acquisition unit 601 and a processing unit 602.

An obtaining unit 601, configured to obtain transaction description information of a target transaction to be executed, where the transaction description information includes description information of at least two transaction objects related to the target transaction;

The acquiring unit 601 is further configured to acquire target graph data of each transaction object in the target transaction according to the transaction description information; the target graph data comprises one or more nodes and connecting edges between the nodes; any node is used for storing object attributes of corresponding transaction objects; any one of the connection edges is used for describing the association relation between the object attributes in the two nodes which are correspondingly connected;

The processing unit 602 is configured to analyze and process the target graph data according to the object attribute and the corresponding association relationship in the target graph data, so as to obtain a shared object attribute between each transaction object in the target graph data;

The processing unit 602 is further configured to determine an execution risk of the target transaction based on the shared object attribute.

In one embodiment, the obtaining unit 601 is specifically configured to:

determining a reference node corresponding to any one of at least two transaction objects indicated by the transaction description information from stored graph data;

According to the association relation of the continuous edge description of the stored graph data and the reference node, determining the association node with the connection relation with the reference node from the stored graph data;

And taking the graph data formed by the reference nodes, the association nodes and the corresponding connection relations corresponding to the transaction objects as target graph data.

In one embodiment, the target graph data is obtained from a graph database belonging to a graph data storage platform in a graph computing platform and used for storing graph data of a full-scale transaction object for performing transaction execution based on the graph computing platform;

the object attributes contained in the graph data of each transaction object stored in the graph database are encryption attributes stored after the object attributes of the corresponding transaction objects are encrypted;

the map computing platform further comprises a map data importing platform, a map database stored by the map data storage platform is a real-time map database, and the map database further comprises a departure map database;

the map data import platform is used for sending the offline map data stored in the offline map database to the real-time map database; the map database for obtaining the target map data is a real-time map database containing offline map data.

In one embodiment, the processing unit 602 is specifically configured to:

Acquiring a storage mode of object attributes of corresponding transaction objects by each node of the target graph data, and acquiring graph calculation rules related to the storage mode based on the storage mode;

And analyzing and processing the target graph data by adopting the graph calculation rule and combining the object attributes stored in the target graph data and the corresponding association relation to obtain the shared object attributes among all transaction objects in the target graph data.

In one embodiment, the processing unit 602 is specifically configured to:

If the storage mode indicates that any node of the target graph data is used for storing one object attribute of the corresponding transaction object, acquiring an attribute calculation rule, and taking the acquired attribute calculation rule as a graph calculation rule related to the storage mode;

If the storage mode indicates that any node of the target graph data is used for storing all object attributes of the corresponding transaction object, acquiring a neighbor heterogeneous node calculation rule, and taking the acquired neighbor heterogeneous node calculation rule as a graph calculation rule related to the storage mode.

In one embodiment, the processing unit 602 is specifically configured to:

According to the object attribute stored in the target graph data and the corresponding association relation, determining one or more attribute types of the object attribute contained in the target graph data and the number of the object attribute contained in each attribute type;

based on the number of the object attributes contained in each attribute type, determining the attribute type of which the number of the object attributes correspondingly contained in one or more attribute types meets a number threshold as a shared attribute type;

and taking the object attributes respectively belonging to all the transaction objects in the target graph data as the shared object attributes of all the transaction objects in the object attributes contained in the shared attribute type.

In one embodiment, if the graph calculation rule is an attribute calculation rule, each node in the target graph data is configured to store an object attribute of a corresponding transaction object; the processing unit 602 is specifically configured to:

Selecting a first object attribute from the target graph data according to the object attribute and the association relation stored in the target graph data; wherein the first object attribute belongs to a first shared attribute type and is associated with a first transaction object in the target transaction, the target transaction further comprises other transaction objects, and the other transaction objects comprise one or more second object attributes in the first shared attribute type;

And comparing the first object attribute with any one of the second object attributes, and taking the same object attribute as one shared object attribute under the first shared attribute type when the comparison result indicates that the first object attribute and the corresponding second object attribute are the same attribute.

In one embodiment, if the graph calculation rule is a neighboring heterogeneous node calculation rule, each node in the target graph data is configured to store all object attributes of a corresponding transaction object, and the determined shared attribute type is a type shared by neighboring nodes; the processing unit 602 is specifically configured to:

Determining a reference node with a connection relation with a node corresponding to any transaction object from the target graph data according to the transaction object indicated by the object attribute stored in the target graph data and the association relation stored in the target graph data;

and acquiring the node number of the reference node, and taking the node number of the reference node as the number of object attributes contained by any transaction object under the type shared by the neighbor nodes.

In one embodiment, the processing unit 602 is specifically configured to:

acquiring a wind control strategy related to the target transaction, wherein the wind control strategy comprises one or more strategy nodes, and any strategy node comprises a distinguishing feature and a corresponding distinguishing condition;

determining a target discrimination feature matched with the shared object attribute and a discrimination condition corresponding to the target discrimination feature from discrimination features included in the strategy node of the wind control strategy;

and determining the execution risk of the target transaction based on the shared object attribute and the corresponding discrimination conditions.

In one embodiment, the determined number of shared object attributes is one or more; the processing unit 602 is specifically configured to:

determining the object relation among the target transaction related transaction objects according to each shared object attribute and the corresponding discrimination conditions; when the number of the shared object attributes meeting the corresponding judging conditions is larger than or equal to the preset number, determining that the object relationship among the target transaction related transaction objects is intimate relationship, otherwise, determining that the object relationship is non-intimate relationship;

and when the determined object relationship is the intimate relationship, determining that the target transaction does not have the execution risk, otherwise, determining that the target transaction has the execution risk.

In one embodiment, the obtaining unit 601 is further configured to obtain transaction log data; the transaction log data comprises one or more executed transaction data and corresponding risk tags, wherein the risk tags are used for indicating whether the executed transaction data has execution risks or not;

The processing unit 602 is further configured to determine an execution risk of any executed transaction data included in the transaction log data by using the determined shared object attribute;

The processing unit 602 is further configured to obtain, based on the execution risk of any executed transaction data and the corresponding risk tag, an accuracy corresponding to the corresponding shared object attribute when performing risk discrimination, and perform risk discrimination application on the shared object attribute based on the accuracy.

In one embodiment, the number of the determined shared object attributes is one or more, the execution risk of any executed transaction data contained in the transaction log data is determined by part or all of the one or more shared object attributes, and the determined accuracy of performing the execution risk determination includes the accuracy of performing the risk determination of any shared object attribute;

The processing unit 602 is specifically configured to: and selecting the shared object attribute with the corresponding accuracy reaching an accuracy threshold from the one or more shared object attributes, and carrying out risk discrimination application on the selected shared object attribute.

In the embodiment of the present application, the acquiring unit 601 may acquire the transaction description information of the target transaction during the execution of the target transaction, so that the acquiring unit 602 may acquire, from the graph database, the target graph data between the transaction objects related to the target transaction based on the acquisition of the transaction description information, and after acquiring the target graph data, the processing unit 602 may analyze and process the target graph data based on the object attribute and the association relationship stored in the target graph data, so as to obtain the shared object attribute between the transaction objects related to the target transaction, and based on the real-time analysis of the transaction objects in the target transaction, the accuracy and the reliability of the analysis of the transaction objects related to the target transaction to be executed at present may be effectively improved. After the processing unit 602 analyzes the shared object attribute related to the transaction object, the execution risk of the target transaction can be determined based on the shared object attribute, so that the flexibility of determining the transaction risk is improved, and the accuracy of determining the transaction risk based on the shared object attribute is ensured because the shared object attribute is determined in real time in the target transaction process.

Fig. 7 is a schematic block diagram of a computer device according to an embodiment of the present invention. The computer device in the present embodiment as shown in fig. 7 may include: one or more processors 701; one or more input devices 702, one or more output devices 703 and a memory 704. The processor 701, the input device 702, the output device 703, and the memory 704 are connected by a bus 705. The memory 704 is used for storing a computer program comprising program instructions, and the processor 701 is used for executing the program instructions stored in the memory 704.

The memory 704 may include volatile memory (RAM), such as random-access memory (RAM); the memory 704 may also include a non-volatile memory (non-volatile memory), such as a flash memory (flash memory), a solid state disk (solid-state drive-STATE DRIVE, SSD), etc.; memory 704 may also include combinations of the above types of memory.

The processor 701 may be a central processing unit (central processing unit, CPU). The processor 701 may further comprise a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (programmable logic device, PLD), or the like. The PLD may be a field-programmable gate array (FPGA) of field-programmable GATE ARRAY, generic array logic (GENERIC ARRAY logic, GAL), or the like. The processor 701 may also be a combination of the above structures.

In an embodiment of the present invention, the memory 704 is configured to store a computer program, where the computer program includes program instructions, and the processor 701 is configured to execute the program instructions stored in the memory 704, to implement the steps of the corresponding method shown in fig. 2 and fig. 4.

In one embodiment, the processor 701 is configured to call the program instructions for executing:

In one embodiment, if the graph calculation rule is an attribute calculation rule, each node in the target graph data is configured to store an object attribute of a corresponding transaction object; the processor 701 is configured to call the program instructions for executing:

In one embodiment, if the graph calculation rule is a neighboring heterogeneous node calculation rule, each node in the target graph data is configured to store all object attributes of a corresponding transaction object, and the determined shared attribute type is a type shared by neighboring nodes; the processor 701 is configured to call the program instructions for executing:

In one embodiment, the determined number of shared object attributes is one or more; the processor 701 is configured to call the program instructions for executing:

Acquiring transaction log data; the transaction log data comprises one or more executed transaction data and corresponding risk tags, wherein the risk tags are used for indicating whether the executed transaction data has execution risks or not;

determining the execution risk of any executed transaction data contained in the transaction log data by adopting the determined shared object attribute;

And obtaining the accuracy corresponding to the corresponding shared object attribute when performing risk discrimination based on the execution risk of any executed transaction data and the corresponding risk label, and performing risk discrimination application on the shared object attribute based on the accuracy.

the processor 701 is configured to call the program instructions for executing: and selecting the shared object attribute with the corresponding accuracy reaching an accuracy threshold from the one or more shared object attributes, and carrying out risk discrimination application on the selected shared object attribute.

Embodiments of the present invention provide a computer program product or computer program comprising computer instructions stored in a computer readable storage medium. The processor of the computer device reads the computer instructions from the computer readable storage medium and executes the computer instructions to cause the computer device to perform the method embodiments described above as shown in fig. 2 or fig. 4. The computer readable 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 foregoing disclosure is merely illustrative of some embodiments of the present invention and it is not to be construed as limiting the scope of the invention, as a person of ordinary skill in the art will appreciate that all or part of the above-described embodiments may be practiced with equivalent variations which fall within the scope of the invention as defined in the appended claims.

Claims

1. A transaction processing method, comprising:

2. The method of claim 1, wherein the obtaining target graph data for each transaction object in the target transaction based on the transaction description information comprises:

3. The method of claim 1 or 2, wherein the target graph data is obtained from a graph database belonging to a graph data storage platform in a graph computing platform and for storing graph data of a full volume of transaction objects for performing a transaction based on the graph computing platform;

4. The method of claim 1, wherein the analyzing the target graph data according to the object attribute and the corresponding association relationship in the target graph data to obtain the shared object attribute between each transaction object in the target graph data includes:

5. The method of claim 4, wherein the obtaining, based on the storage manner, a graph computation rule associated with the storage manner comprises:

6. The method of claim 1, wherein the analyzing the target graph data according to the object attribute and the corresponding association relationship in the target graph data to obtain the shared object attribute between each transaction object in the target graph data includes:

7. The method of claim 4, wherein if the graph computation rule is an attribute computation rule, each node in the target graph data is configured to store an object attribute of a corresponding transaction object; the method for analyzing and processing the target graph data by adopting the attribute calculation rule and combining the object attributes stored in the target graph data and the corresponding association relation to obtain the shared object attributes among all transaction objects in the target graph data comprises the following steps:

8. The method of claim 4, wherein if the graph computation rule is a neighboring heterogeneous node computation rule, each node in the target graph data is configured to store all object attributes of a corresponding transaction object, and the determined type of the shared attribute is a type shared by neighboring nodes; the method for determining the number of the object attributes contained in the shared attribute type which is the type shared by the neighbor nodes by adopting the calculation rule of the neighbor heterogeneous nodes comprises the following steps:

9. The method of claim 1, wherein the determining the execution risk of the target transaction based on the shared object attribute comprises:

10. The method of claim 9, wherein the determined number of shared object attributes is one or more; the determining the execution risk of the target transaction based on the shared object attribute and the corresponding discrimination conditions comprises the following steps:

11. The method of claim 1, wherein the method further comprises:

12. The method of claim 11, wherein the determined number of shared object attributes is one or more, the transaction log data includes any risk of execution of executed transaction data determined from part or all of the one or more shared object attributes, and the determined accuracy of performing risk determination includes the accuracy of performing risk determination for any shared object attribute;

the applying risk discrimination on the shared object attribute based on the accuracy includes: and selecting the shared object attribute with the corresponding accuracy reaching an accuracy threshold from the one or more shared object attributes, and carrying out risk discrimination application on the selected shared object attribute.

13. A transaction processing device, comprising:

14. A computer device comprising a processor, an input device, an output device and a memory, the processor, the input device, the output device and the memory being interconnected, wherein the memory is adapted to store a computer program comprising program instructions, the processor being configured to invoke the program instructions to perform the method of any of claims 1-12.

15. A computer readable storage medium, characterized in that the computer readable storage medium stores a computer program comprising program instructions which, when executed by a processor, cause the processor to perform the method of any of claims 1-12.