CN116975160A - Transaction data processing method, device, equipment, medium and product - Google Patents
Transaction data processing method, device, equipment, medium and product Download PDFInfo
- Publication number
- CN116975160A CN116975160A CN202310523679.1A CN202310523679A CN116975160A CN 116975160 A CN116975160 A CN 116975160A CN 202310523679 A CN202310523679 A CN 202310523679A CN 116975160 A CN116975160 A CN 116975160A
- Authority
- CN
- China
- Prior art keywords
- transaction
- type
- target
- transactions
- range
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000003672 processing method Methods 0.000 title claims abstract description 19
- 238000011156 evaluation Methods 0.000 claims abstract description 207
- 230000002159 abnormal effect Effects 0.000 claims abstract description 133
- 238000000034 method Methods 0.000 claims abstract description 69
- 230000000977 initiatory effect Effects 0.000 claims description 118
- 238000012545 processing Methods 0.000 claims description 29
- 238000004590 computer program Methods 0.000 claims description 13
- 238000003860 storage Methods 0.000 claims description 10
- 230000008569 process Effects 0.000 claims description 9
- 230000005856 abnormality Effects 0.000 claims description 6
- 238000005516 engineering process Methods 0.000 abstract description 16
- 238000013473 artificial intelligence Methods 0.000 abstract description 13
- 238000012546 transfer Methods 0.000 description 87
- 230000001364 causal effect Effects 0.000 description 55
- 238000009826 distribution Methods 0.000 description 26
- 238000004458 analytical method Methods 0.000 description 12
- 238000012549 training Methods 0.000 description 11
- 238000010801 machine learning Methods 0.000 description 9
- 230000006870 function Effects 0.000 description 8
- 238000010586 diagram Methods 0.000 description 6
- 238000000605 extraction Methods 0.000 description 6
- 238000007476 Maximum Likelihood Methods 0.000 description 5
- 238000002360 preparation method Methods 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 4
- 230000008859 change Effects 0.000 description 4
- 238000010206 sensitivity analysis Methods 0.000 description 4
- 238000007619 statistical method Methods 0.000 description 4
- 238000007405 data analysis Methods 0.000 description 3
- 230000008030 elimination Effects 0.000 description 3
- 238000003379 elimination reaction Methods 0.000 description 3
- 230000003993 interaction Effects 0.000 description 3
- 238000007781 pre-processing Methods 0.000 description 3
- 238000012502 risk assessment Methods 0.000 description 3
- 230000008685 targeting Effects 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000002790 cross-validation Methods 0.000 description 2
- 238000013480 data collection Methods 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 235000013305 food Nutrition 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 241000219991 Lythraceae Species 0.000 description 1
- 235000014360 Punica granatum Nutrition 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 230000003190 augmentative effect Effects 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000033228 biological regulation Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 230000006698 induction Effects 0.000 description 1
- 239000003999 initiator Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000002787 reinforcement Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012954 risk control Methods 0.000 description 1
- 238000013526 transfer learning Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/28—Databases characterised by their database models, e.g. relational or object models
- G06F16/284—Relational databases
- G06F16/285—Clustering or classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2462—Approximate or statistical queries
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/04—Trading; Exchange, e.g. stocks, commodities, derivatives or currency exchange
Abstract
The embodiment of the application discloses a transaction data processing method, a device, equipment, a medium and a product, which relate to the artificial intelligence technology, and the method comprises the following steps: acquiring a transaction reference object; the transaction reference object has at least one object range for evaluating a transaction type of the transaction; each object range has an association relationship with the transaction type of the transaction; acquiring transaction attribute information of a plurality of reference transactions; the transaction attribute information of the reference transaction comprises a transaction type and transaction reference information, wherein the type of the transaction reference information is matched with the type of the transaction reference object; calculating a transaction evaluation parameter associated with each object range based on the transaction type of each reference transaction and the object range hit by the transaction reference information of each reference transaction in at least one object range; the transaction evaluation parameters are used to evaluate the probability that the transaction type of the transaction is an abnormal transaction type. By adopting the embodiment of the application, the accuracy of evaluating the transaction type of the transaction can be improved.
Description
Technical Field
The present application relates to the field of artificial intelligence technologies, and in particular, to a transaction data processing method, apparatus, device, medium, and product.
Background
With the development of information networks, more and more transactions are performed in an online transaction mode, but online transactions cannot acquire the real situation of merchants, so that larger transaction risks exist and transaction losses are easy to cause. Currently, whether a transaction is at risk is generally evaluated according to the identity (such as a name) of a merchant associated with the transaction, for example, a list of identities of merchants at risk for the transaction may be maintained in advance, if the identity of a subsequent merchant hits the list of identities, the transaction of the merchant may be considered to be at risk (i.e., abnormal), but such a manner of evaluating the transaction risk by the identity of the merchant is not accurate in many cases.
Disclosure of Invention
The embodiment of the application provides a transaction data processing method, a device, equipment, a medium and a product, which can improve the accuracy of evaluating the transaction type of a transaction.
In a first aspect, the present application provides a transaction data processing method, including:
acquiring a transaction reference object; the transaction reference object has at least one object range for evaluating a transaction type of a transaction; the at least one object range has an association relationship with the transaction type of the transaction, wherein the transaction type of the transaction is a normal transaction type or an abnormal transaction type;
Acquiring transaction attribute information of a plurality of reference transactions; the transaction attribute information of any reference transaction comprises transaction type and transaction reference information of the any reference transaction, and the type of the transaction reference information is matched with the type of the transaction reference object;
calculating a first transaction evaluation parameter associated with each object range based on the transaction type of each reference transaction and the object range hit by the transaction reference information of each reference transaction in the at least one object range; a first transaction evaluation parameter associated with any of the object ranges for evaluating a probability that a transaction type of any of the transactions is an abnormal transaction type.
In a second aspect, the present application provides a transaction data processing apparatus comprising:
a range determining unit for acquiring a transaction reference object; the transaction reference object has at least one object range for evaluating a transaction type of a transaction; the at least one object range has an association relationship with the transaction type of the transaction, wherein the transaction type of the transaction is a normal transaction type or an abnormal transaction type;
a data acquisition unit for acquiring transaction attribute information of a plurality of reference transactions; the transaction attribute information of any reference transaction comprises transaction type and transaction reference information of the any reference transaction, and the type of the transaction reference information is matched with the type of the transaction reference object;
A data calculating unit, configured to calculate a first transaction evaluation parameter associated with each object range based on a transaction type of each reference transaction and an object range hit by transaction reference information of each reference transaction in the at least one object range, respectively; a first transaction evaluation parameter associated with any of the object ranges for evaluating a probability that a transaction type of any of the transactions is an abnormal transaction type.
Optionally, any one of the at least one object range is a target object range; the data calculation unit is specifically configured to:
determining the transaction reference information of the plurality of reference transactions as a target reference transaction, wherein the transaction reference information hits the reference transaction of the target object range;
counting the target quantity of target reference transactions with the transaction type being the normal transaction type in at least one target reference transaction;
a ratio between the target number and the total number of the at least one target reference transaction is determined as a first transaction evaluation parameter associated with the target object range.
Optionally, the plurality of reference transactions each have transaction reference information; the range determining unit is specifically configured to:
acquiring a plurality of candidate object ranges of the transaction reference object;
Determining a first reference transaction which hits each candidate object range of the transaction reference object by targeting the transaction reference information which belongs to the plurality of reference transactions;
determining a reference transaction with the abnormal transaction type as a second reference transaction hitting each candidate object range in the first reference transaction hitting each candidate object range;
determining the ratio between the number of the second reference transactions hitting each candidate object range and the number of the first reference transactions as the association degree between each candidate object range and the transaction type of the transactions respectively;
and taking the candidate object range with the association degree with the transaction type of the transaction being greater than or equal to the association degree threshold value as the object range of the transaction reference object.
Optionally, the transaction reference object includes at least one of:
the method comprises the steps of initiating time of a transaction, transferring amount of the transaction resources, geographic position of a transaction initiating object of the transaction, object type of a resource issuing object of a business resource to be acquired by the transaction, or geographic position of the resource issuing object.
Optionally, the transaction reference objects are of a plurality, each transaction reference object having at least one object range; the transaction data processing device further includes a transaction prediction unit for:
Acquiring a plurality of transaction reference information of a target transaction; the type of a transaction reference information of the target transaction is matched with the type of a transaction reference object;
acquiring an object range of each transaction reference information of the target transaction, which is hit in at least one object range of the matched transaction reference object, as a hit object range of each transaction reference information;
calculating a target transaction evaluation parameter of the target transaction based on the first transaction evaluation parameter associated with the hit object range of each transaction reference information;
wherein the target transaction evaluation parameter is a probability that the transaction type of the target transaction is an abnormal transaction type.
Optionally, the transaction prediction unit is specifically configured to:
calculating a first product between first transaction evaluation parameters associated with the hit range of each transaction reference information;
acquiring first priori information of the plurality of reference transactions, and calculating a second product between the first product and the numerical value of the first priori information; the first priori information is used for indicating the proportion of the reference transaction with the abnormal transaction type in the plurality of reference transactions;
the target transaction evaluation parameter is determined based on the second product.
Optionally, the transaction prediction unit is specifically configured to:
counting the number of reference transactions with abnormal transaction types in the plurality of reference transactions;
the ratio between the number of reference transactions of the transaction type abnormal transaction type and the total number of the plurality of reference transactions is determined as the first priori information.
Optionally, the transaction prediction unit is specifically configured to:
calculating a second transaction evaluation parameter associated with the hit object range of each transaction reference information based on the first transaction evaluation parameter associated with the hit object range of each transaction reference information; a second transaction evaluation parameter associated with any hit range for evaluating a probability that the transaction type of the target transaction is an abnormal transaction type;
calculating a third product between the second transaction evaluation parameters associated with the hit range of each transaction reference information;
acquiring second priori information of the plurality of reference transactions, and calculating a fourth product between the third product and the value of the second priori information; the second priori information is used for indicating the proportion of the reference transaction with the normal transaction type in the plurality of reference transactions;
and acquiring the sum value of the second product and the fourth product, and determining the ratio between the second product and the sum value as the target transaction evaluation parameter.
Optionally, the transaction prediction unit is specifically configured to:
counting the number of reference transactions with the transaction type being a normal transaction type in the plurality of reference transactions;
the ratio between the number of reference transactions of the transaction type normal to the total number of the plurality of reference transactions is determined as the second prior information.
Optionally, the target transaction is a transaction initiated by a transaction initiation object to a resource release object, where the transaction initiation object is configured to obtain a service resource held by the resource release object based on the initiated target transaction; the transaction data processing device further comprises a transaction processing unit for:
if the value of the target transaction evaluation parameter is greater than or equal to the parameter threshold, stopping the execution of the target transaction, and prompting the transaction initiating object that the resource issuing object has abnormality.
In a third aspect, the present application provides a computer device comprising a processor, a memory, wherein the memory is for storing a computer program comprising program instructions, the processor being configured to invoke the program instructions to perform the transaction data processing method described above.
In a fourth aspect, the present application provides a computer readable storage medium having stored therein a computer program adapted to be loaded and executed by a processor to cause a computer device having the processor to perform the transaction data processing method described above.
In a fifth aspect, the present application provides a computer program product or computer program comprising computer instructions which, when executed by a processor, implement the transaction data processing method described above.
In the embodiment of the application, the transaction reference object is obtained, and is provided with at least one object range for evaluating the transaction type of the transaction, wherein each object range has an association relationship with the transaction type of the transaction; by acquiring the transaction attribute information of the plurality of reference transactions, the first transaction evaluation parameter associated with each object range may be calculated based on the transaction type of each reference transaction and the object range in which the transaction reference information of each reference transaction hits in at least one object range, respectively. The first transaction evaluation parameter may be used to evaluate a probability that a transaction type of the transaction is an abnormal transaction type. According to the method, through calculating the first transaction evaluation parameters associated with the object ranges of the transaction reference object, the probability that the transaction type of the transaction is the abnormal transaction type under the condition of each object range can be determined, namely the probability that the transaction is the abnormal transaction type under the condition of each object range, and then various factors affecting the transaction type can be determined. For example, determining which object ranges of which transaction reference objects affect the transaction type, i.e., the probability of being an abnormal transaction type is greater under the conditions of which object ranges of which transaction reference objects and the probability of being an abnormal transaction type is less under the conditions of which object ranges of which transaction reference objects, may promote the accuracy of evaluating the transaction type of the transaction.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic diagram of a network architecture of a transaction data processing system according to an embodiment of the present application;
FIG. 2 is a flow chart of a transaction data processing method according to an embodiment of the present application;
FIG. 3 is a schematic view of an object range of a transaction reference object according to an embodiment of the present application;
FIG. 4 is a flow chart of another transaction data processing method according to an embodiment of the present application;
FIG. 5 is a schematic view of a scenario for determining target transaction evaluation parameters of a target transaction according to an embodiment of the present application;
FIG. 6 is a schematic diagram of a transaction data processing device according to an embodiment of the present application;
fig. 7 is a schematic diagram of a composition structure of a computer device according to an embodiment of the present application.
Detailed Description
The following description of the embodiments of the present application will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present application, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
Artificial intelligence (Artificial Intelligence, AI) is the theory, method, technique and application system that uses a digital computer or a machine controlled by a digital computer to simulate, extend and extend human intelligence, sense the environment, acquire knowledge and use the knowledge to obtain optimal results. In other words, artificial intelligence is an integrated technology of computer science that attempts to understand the essence of intelligence and to produce a new intelligent machine that can react in a similar way to human intelligence. Artificial intelligence, i.e. research on design principles and implementation methods of various intelligent machines, enables the machines to have functions of sensing, reasoning and decision.
The artificial intelligence technology is a comprehensive subject, and relates to the technology with wide fields, namely the technology with a hardware level and the technology with a software level. Artificial intelligence infrastructure technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technologies, operation/interaction systems, mechatronics, and the like. The artificial intelligence software technology mainly comprises a computer vision technology, a voice processing technology, a natural language processing technology, machine learning/deep learning and other directions. The scheme provided by the embodiment of the application belongs to the machine learning technology in the field of artificial intelligence.
Machine Learning (ML) is a multi-domain interdisciplinary, involving multiple disciplines such as probability theory, statistics, approximation theory, convex analysis, algorithm complexity theory, etc. It is specially studied how a computer simulates or implements learning behavior of a human to acquire new knowledge or skills, and reorganizes existing knowledge structures to continuously improve own performance. Machine learning is the core of artificial intelligence, a fundamental approach to letting computers have intelligence, which is applied throughout various areas of artificial intelligence. Machine learning and deep learning typically include techniques such as artificial neural networks, confidence networks, reinforcement learning, transfer learning, induction learning, teaching learning, and the like. For example, an artificial neural network in a machine learning technique may be employed in the present application to calculate a first transaction evaluation parameter associated with each object range based on the transaction type of each reference transaction and the object range that each reference transaction's transaction reference information hits in at least one object range, respectively, and so on.
The technical scheme of the application can be applied to the scene of evaluating transaction risks to determine the transaction type of the transaction. For example, the method can be applied to various online shopping websites, paymate and financial applications, and whether the transaction is of an abnormal transaction type can be determined by calculating the evaluation parameters of the transaction generated under various scenes, so that prompt can be carried out for the abnormal transaction, and the transaction loss is reduced. The technical scheme of the application can be applied to various scenes, including but not limited to cloud technology, intelligent traffic, auxiliary driving and the like.
It should be specifically noted that, in the embodiments of the present application, data (such as transaction attribute information, transaction reference information, etc.) related to information about objects (such as transaction initiation objects and resource release objects), when the embodiments of the present application are applied to specific products or technologies, permission or consent of the objects needs to be obtained, and collection, use and processing of related data need to comply with related laws and regulations and standards of related countries and regions. For example, the transaction initiation object and the resource issuance object may refer to users of the terminal device or the computer device.
Referring to fig. 1, fig. 1 is a schematic diagram of a network architecture of a transaction data processing system according to an embodiment of the present application, as shown in fig. 1, a computer device may perform data interaction with terminal devices, and the number of the terminal devices may be one or at least two. For example, when the number of terminal apparatuses is plural, the terminal apparatuses may include the terminal apparatus 101a, the terminal apparatus 101b, the terminal apparatus 101c, and the like in fig. 1. Taking the terminal device 101a as an example, the computer device 102 may obtain the transaction reference object and transaction attribute information of a plurality of reference transactions. Further, the computer device 102 may calculate a first transaction evaluation parameter associated with each object range based on the transaction type of each reference transaction and the object range that each reference transaction's transaction reference information hits in at least one object range, respectively. Alternatively, the computer device 102 may determine whether the transaction type of the transaction is an abnormal transaction type based on the first transaction evaluation parameter, or the computer device 102 may also send the first transaction evaluation parameter to the terminal device 101a to cause the terminal device 101a to determine whether the transaction type of the transaction is an abnormal transaction type based on the first transaction evaluation parameter, and so on.
It is understood that the computer devices mentioned in the embodiments of the present application include, but are not limited to, terminal devices or servers. In other words, the computer device may be a server or a terminal device, or may be a system formed by the server and the terminal device. The above-mentioned terminal device may be an electronic device, including, but not limited to, a mobile phone, a tablet computer, a desktop computer, a notebook computer, a palm computer, a vehicle-mounted device, an intelligent voice interaction device, an augmented Reality (AR/VR) device, a head mounted display, a wearable device, a smart speaker, a smart home appliance, an aircraft, a digital camera, a camera, and other mobile internet devices (mobile internet device, MID) with network access capability, etc. The servers mentioned above may be independent physical servers, or may be server clusters or distributed systems formed by a plurality of physical servers, or may be cloud servers that provide cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, vehicle-road collaboration, content distribution networks (Content Delivery Network, CDN), and basic cloud computing services such as big data and artificial intelligence platforms.
Further, referring to fig. 2, fig. 2 is a flow chart of a transaction data processing method according to an embodiment of the present application; as shown in fig. 2, the transaction data processing method can be applied to a computer device, and the transaction data processing method includes, but is not limited to, the following steps:
s101, acquiring a transaction reference object.
The technical scheme of the application can be applied to any application needing to conduct transactions, for example, but not limited to online shopping application, payment application and financial application, and the probability of each transaction being an abnormal transaction type can be determined by calculating the transaction evaluation parameter of each transaction data, so as to determine whether each transaction is an abnormal transaction or not. When the transaction is determined to be abnormal, the transaction can be processed, such as intercepting the transaction execution or limiting the transaction execution, so that the transaction safety is ensured, and the transaction risk is reduced.
In embodiments of the present application, the transaction reference object may be an object for evaluating a transaction, e.g., the transaction reference object may refer to certain factors that can affect the transaction, e.g., the transaction reference object may include the time of initiation of the transaction, the amount of resource transfer of the transaction, the type of transaction, etc. By determining these factors that can affect the transaction, the probability that the transaction is at risk can be determined, and thus whether the transaction is of an abnormal transaction type.
Optionally, the transaction reference object may include at least one of: the method comprises the steps of initiating time of a transaction, resource transfer amount of the transaction, geographic position of a transaction initiating object of the transaction, object type of a resource issuing object of a business resource to be acquired by the transaction, geographic position of the resource issuing object, type of the transaction or registration time of the resource issuing object. The resource transfer amount may refer to, for example, an amount transferred by a transaction exchange, the transaction initiating object may refer to a transaction initiator, for example, the transaction is initiated by the object a to the object b, and the transaction initiating object may refer to the object a, and the object b may be referred to as a resource issuing object, that is, the object a initiates the transaction to the object b, and the object b issues the service resource to the object a. For example, in the scene of an online shopping website, the object A initiates a transaction to the object B, the object A pays the resource corresponding to the transaction to the object B, such as transferring accounts, and the object B issues business resources such as commodities to the object A. The geographic location where the transaction initiating object is located may refer to, for example, the city where the transaction initiating object is located. The object type of the resource issuing object may refer to a business scope, and may include, for example, types of clothing, food, department stores, and the like. The geographic location where the resource provisioning object is located may refer to, for example, a city where the resource provisioning object is located. The types of transactions may include, for example, electronic card transactions or online banking transactions, among others. The registration time of the resource issuing object may refer to, for example, a time when the merchant registers with the registration authority, and so on.
Wherein the transaction reference object has at least one object range for evaluating a transaction type of the transaction; at least one object range has an association relationship with the transaction type of the transaction, and the transaction type of the transaction is a normal transaction type or an abnormal transaction type. A normal transaction type may refer to the transaction having no risk or a probability of having a risk less than a risk threshold, and an abnormal transaction type may refer to the transaction having a risk or a probability of having a risk greater than or equal to a risk threshold. For example, the abnormal transaction types may include, but are not limited to, fraud risk types, no-shipment risk types for collection, no-qualifying risk types for quality of goods, and so forth. The object range may be used, for example, to indicate an interval to which the transaction reference object belongs, and an object range may refer to an interval, such as a time interval, a resource transfer amount interval, a geographic location interval, an object type interval, and so on. For example, the time intervals may include, but are not limited to, intervals of [0,8 ], [8, 16), and [16, 24), the resource transfer amount intervals may include, but are not limited to, intervals of [0,500 ], [500,1000 ], [1000,1500 ], + -infinity), and the like, the geographic location interval may refer to a jurisdiction to which a city belongs, for example, the geographic location interval may include, but is not limited to, intervals of Chongqing, shenzhen, chengdu, beijing, shanghai, and the like. The object type interval may include intervals of clothing, food, department, and the like.
In one possible implementation, the association that the object scope has with the transaction type of the transaction may include, but is not limited to, a causal relationship, e.g., the object scope has a causal relationship with the transaction type of the transaction, i.e., the object scope affects the transaction type of the transaction. For example, the transaction reference object may have a different range of objects and the transaction type of the transaction may be different. It will be appreciated that transaction reference objects are of different types and have different numbers of object ranges. For example, when the transaction reference object is the initiation time of the transaction, the number of object ranges that can be present may be 3, such as [0,8 ], [8, 16), and [16, 24). Where the transaction reference object is a resource transfer amount of a transaction, the number of object ranges that may be present may be 4, for example, [0,500 ], [500,1000 ], [1000,1500 ], [1500, ++ infinity). The transaction reference object is the geographic position of the transaction initiating object of the transaction, and the number of the object ranges can be 5, such as Chongqing, shenzhen, chengdu, beijing, shanghai, and the like. Each transaction reference object may have an object range that includes more or less objects, which is not limited by embodiments of the present application.
Alternatively, the type of transaction reference object, the number of each type of transaction reference object, and the object range of each type of transaction reference object, i.e., which aspects of the transaction need to be evaluated, may be preset, so that these aspects of the transaction attribute information of the reference transaction may be evaluated when the reference transaction is subsequently acquired.
S102, acquiring transaction attribute information of a plurality of reference transactions.
In the embodiment of the application, the transaction attribute information of a plurality of reference transactions can be obtained from a database, or the transaction attribute information of a plurality of reference transactions can be obtained from various transaction platforms, which is not limited in the embodiment of the application. Wherein the transaction attribute information may be used to reflect transaction content of the reference transaction, for example, the transaction attribute information of the reference transaction may include transaction data and related data of the reference transaction. The transaction data may include, for example, but is not limited to, the time of initiation of the transaction, the amount of resource transfer of the transaction, the geographic location where the transaction initiation object is located, and so forth. The relevant data may include, for example, but not limited to, a geographic location where the resource provisioning object is located, an object type of the resource provisioning object to which the business resource to be acquired by the transaction belongs, a registration time of the resource provisioning object, and so on.
Wherein the transaction attribute information of any reference transaction includes a transaction type of any reference transaction and transaction reference information, the type of the transaction reference information matching the type of the transaction reference object. The type of transaction reference information matching the type of transaction reference object may refer to: a transaction reference message may refer to a value of a transaction reference object. For example, the transaction reference object is the initiation time of the transaction, the transaction reference information may refer to a specific initiation time of the transaction, for example, the specific initiation time of the transaction is 10. Alternatively, the transaction reference object includes a specific initiation time of the transaction and a specific resource transfer amount of the transaction, and the transaction reference information may refer to the specific initiation time of the transaction (e.g., initiation time of 10) and the specific resource transfer amount of the transaction (e.g., resource transfer amount of 700).
That is, if the type of the transaction reference object includes the initiation time of the transaction, the transaction reference information of the reference transaction includes the specific initiation time of the reference transaction. If the type of transaction reference object includes a resource transfer amount for a transaction, the transaction reference information for the reference transaction includes a specific resource transfer amount for the reference transaction. The transaction attribute information includes a transaction type of the reference transaction and transaction reference information of the reference transaction. The transaction type may be used to indicate that the reference transaction is either an abnormal transaction type or a normal transaction type, and the transaction reference information may include a transaction reference object for the reference transaction and a value corresponding to each type of transaction reference object. For example, the transaction attribute information of the reference transaction 1 may include [ reference transaction 1, abnormal transaction type, initiation time of transaction 10, resource transfer amount of transaction 700], and the transaction attribute information of the reference transaction 2 may include [ reference transaction 2, normal transaction type, initiation time of transaction 5, resource transfer amount of transaction 200].
In one embodiment, a plurality of candidate object ranges may be obtained, and an object range having an association relationship with a transaction type of a reference transaction is obtained from the plurality of candidate object ranges, so that subsequent processing is performed. Specifically, the procedure of acquiring at least one object range of the transaction reference object may be as follows:
Acquiring a plurality of candidate object ranges of a transaction reference object; determining the reference transaction of each candidate object range of the transaction reference object as a first reference transaction hit in each candidate object range by hit transaction reference information belonging to the plurality of reference transactions; determining a reference transaction with the abnormal transaction type as a second reference transaction hitting each candidate object range in the first reference transaction hitting each candidate object range; determining the ratio between the number of the second reference transactions hitting each candidate object range and the number of the first reference transactions as the association degree between each candidate object range and the transaction type of the transactions respectively; and taking the candidate object range with the association degree with the transaction type of the transaction being greater than or equal to the association degree threshold value as the object range of the transaction reference object.
Wherein each of the plurality of reference transactions has transaction reference information. For example, the transaction reference object is the initiation time of the transaction, the candidate object range may refer to a plurality of intervals corresponding to the transaction reference object's transaction initiation time, e.g., the candidate object range may include [0,8 ], [8, 16), and [16, 24). For example, a reference transaction reference object, such as a reference transaction number of 10, includes the time of initiation of the transaction and the amount of resource transfer of the transaction, the transaction type may include 0 or 1, and a transaction type of 0 may indicate that the transaction is not risky, i.e., the transaction is a normal transaction type. A transaction type of 1 may indicate that there is a risk for the transaction, i.e., the transaction is an abnormal transaction type, and 10 reference transactions may be as shown in table 1:
TABLE 1
| Resource transfer amount | Initiation time | Transaction type |
| 200 | 5 | 0 |
| 700 | 10 | 1 |
| 1200 | 8 | 1 |
| 1300 | 20 | 1 |
| 800 | 16 | 0 |
| 300 | 4 | 0 |
| 1400 | 12 | 1 |
| 600 | 18 | 0 |
| 900 | 6 | 0 |
| 1000 | 14 | 1 |
By statistically analyzing 10 reference transactions in table 1, a correspondence relationship between the resource transfer amount of the reference transaction and the transaction type of the reference transaction can be obtained, as shown in table 2:
TABLE 2
As shown in table 2, taking the transaction reference object as an example of the resource transfer amount of the transaction, the number of the plurality of candidate object ranges of the transaction reference object may include 4, for example, [0,500 ], [500,1000 ], [1000,1500 ], [1500, + ], respectively. There are 2 reference transaction hit candidate ranges [0,500 ] among 10 reference transactions, there are 3 reference transaction hit candidate ranges [500, 1000), there are 4 reference transaction hit candidate ranges 1000,1500, there are 1 reference transaction hit candidate ranges 1500, +++). Taking the candidate range [0,500) as an example, the reference transaction with the abnormal transaction type as the transaction type among the 3 first reference transactions hitting the candidate range [0,500) is determined as the second reference transaction hitting the candidate range [0,500), and the number of the second reference transactions is 0. The ratio between the number of second reference transactions and the number of first reference transactions hitting the candidate range, respectively, i.e. 0/2=0, is determined as the degree of association between the candidate range [0,500) and the transaction type of the reference transaction. From this, it can be calculated that the degree of association between the candidate object range [500, 1000) and the transaction type of the reference transaction is 2/3=0.67, the degree of association between the candidate object range [1000, 1500) and the transaction type of the reference transaction is 3/4=0.75, and the degree of association between the candidate object range [1500, + ] and the transaction type of the reference transaction is 0/1=0, respectively.
Further, a candidate object range with a degree of association greater than the degree of association threshold may be taken as the object range of the transaction reference object. For example, if the association threshold is 0.5, the candidate object range [500,1000 ] and the candidate object range [1000, 1500) are set as the object ranges of the transaction reference object. I.e. the number of object ranges of the resource transfer amount of the transaction may be 2, object ranges [500,1000 ] and object ranges [1000,1500 ], respectively.
It is understood that if the association degree between the plurality of candidate object ranges and the transaction type of the transaction is greater than the association degree threshold, the plurality of candidate object ranges may be used as object ranges of the transaction reference object.
In the embodiment of the application, the association degrees between the plurality of candidate object ranges of the transaction reference object and the transaction types of the transaction are respectively calculated, so that the object range with higher association degree with the transaction types can be determined, thereby establishing the association relation between the plurality of object ranges of the transaction reference object and the transaction types. When a new transaction exists subsequently, transaction evaluation parameters of the new transaction can be calculated by combining the association relationship, so that the probability that the new transaction is of an abnormal transaction type is determined, and whether the new transaction is an abnormal transaction or not is further determined. If the transaction reference object is of another kind, such as the initiation time of the transaction, the geographic position of the transaction initiation object of the transaction, the object type of the resource issuing object to which the service resource to be acquired by the transaction belongs, or the geographic position of the resource issuing object, the association relationship between the multiple object ranges of each transaction reference object and the transaction type can be calculated respectively by referring to the method, so as to determine the association relationship between the transaction type and each transaction reference object.
In one embodiment, a causal relationship between at least one object range of the transaction reference object and the transaction type of the transaction may be constructed, for example, the resource transfer amount of the transaction may be divided into 4 intervals of [0,500 ], [500,1000 ], [1000,1500 ], [1500, + ], respectively, and the initiation time of the transaction into 3 intervals of [0,8 ], [8, 16), and [16, 24), respectively, in advance. Further, the interval corresponding to each resource transfer amount and the interval corresponding to the initiation time may be taken as one node, and the transaction type may be taken as one node, for example, 8 nodes may be determined, where 7 factor nodes are respectively: resource transfer amount [0,500 ], resource transfer amount [500, 1000), resource transfer amount [1000, 1500), resource transfer amount [1500, + -infinity), initiation times [0, 8), [8, 16), and [16, 24), and 1 fruit node are transaction types.
Further, by calculating the probability that the transaction type is an abnormal transaction type under the condition of each factor node respectively, if the probability that the transaction type is an abnormal transaction type under the condition of a certain factor node is calculated to be greater than the association threshold, the factor node and the transaction type node are considered to have a causal relation, one side between the factor node and the transaction type node can be constructed, the side between each factor node and the transaction type node is constructed respectively, and the node with the side is taken as the node in the causal graph, so that the causal graph can be constructed. It can be understood that more cause nodes may be further included, for example, a plurality of intervals corresponding to a geographic location where a transaction initiating object of a transaction is located, a plurality of intervals corresponding to an object type of a resource issuing object where a service resource to be acquired by the transaction belongs, or a plurality of intervals corresponding to a geographic location where the resource issuing object is located, so as to determine whether edges exist between each cause node and a transaction type node, and further construct a causal graph including more cause nodes, where a causal relationship in the causal graph may represent a correlation relationship between an object range of a transaction reference object and a transaction type of the transaction.
In the embodiment of the application, the causal relationship between each node in the causal graph can be determined by constructing the causal graph, so that the association relationship between the object range of the transaction reference object and the transaction type of the transaction is determined, namely, a plurality of factors possibly influencing the transaction type of the transaction are determined, and the transaction is evaluated by combining the plurality of factors possibly influencing the transaction type of the transaction, so that the probability that the transaction is an abnormal transaction type is determined. By establishing causal relationships, causal relationships between different variables (a variable may refer to an object range or transaction type of a transaction reference object), such as causal relationships between object ranges and transaction risks and influencing factors, may be taken into account, so that accuracy of transaction prediction may be improved, and accuracy and reliability of risk prediction may be further improved.
In one embodiment, data preparation may be performed in advance, the data preparation being the first step of transaction prediction, the purpose of the data preparation being to collect transaction data and related data of a reference transaction to construct a predictive model, whereby the probability that a transaction is an abnormal transaction type may be predicted using the predictive model. Data preparation may include the steps of data acquisition, data cleansing, data feature extraction. In particular, data collection may refer to preparing a plurality of reference transaction data in advance so that transaction data and other relevant data of the reference transaction may be collected, the transaction data may include, for example, but not limited to, a time of initiation of the transaction, an amount of resource transfer of the transaction, a geographic location where the transaction initiation object is located, and so on. The relevant data may include, for example, but not limited to, a geographic location where the resource provisioning object is located, an object type of the resource provisioning object to which the business resource to be acquired by the transaction belongs, a registration time of the resource provisioning object, and so on.
Further, data cleansing may refer to data cleansing of acquired transaction data and other related data of a reference transaction, i.e., cleansing and preprocessing of such data, such as data deduplication, data population, outlier processing, and so forth. Where data deduplication may refer to deduplication of data, data population may refer to population of missing data with a median or mode, outlier processing may include outlier removal or outlier modification, for example. Further, the data feature extraction may refer to feature extraction of the collected data, and the feature extraction may be followed by causal graph construction and data prediction. The feature extraction method may be, for example, statistical analysis or machine learning. For example, after the transaction data and the related data are obtained, the feature is obtained by counting the transaction data and the related data, and the feature refers to the counted transaction data and the related data. Further, the transaction data and related data may be data set partitioned to obtain training and testing sets for subsequent data prediction and evaluation.
After data preparation, the data may be subjected to a causal relationship analysis, the purpose of which is to analyze causal relationships between different variables in the trading data and construct a causal graph. Wherein, the causal relationship analysis can comprise the steps of variable selection, data analysis, causal reasoning and the like. The variable selection may be to select a variable that has an influence on the transaction as an analysis object, for example, the variable may include a resource transfer amount of the transaction, an initiation time of the transaction, a type of the transaction, a geographic location where the resource issuing object is located, and the like. Data analysis may refer to analysis of collected variables, and data analysis methods may include statistical analysis, machine learning, and the like. For example, correlations, distributions, etc. between different variables may be analyzed. Causal reasoning can refer to analyzing causal relationships between different variables using causal reasoning techniques and building a causal graph. The purpose of causal reasoning is to determine causal relationships between variables, i.e. how a change in a certain variable affects other variables, e.g. how the time of initiation of a trade affects the type of trade, or how the amount of resource transfer of a trade affects the type of trade, or how the type of trade affects the type of trade, etc. by means of a causal graph. Causal reasoning can be implemented using reasoning tools or statistical methods, such as cause alnex, pomegranate, a graph model and probabilistic model toolkit, etc.
Optionally, after the causal graph is constructed, it may be further determined whether some variables in the causal graph have no direct relationship with the transaction type or have no obvious influence on the transaction type, and if the variables exist, the variables may be removed from the causal graph, so as to simplify the causal graph, and also simplify the subsequent prediction model. Through the flow, the causal graph can be constructed, so that the object range of the transaction reference object with the association relation with the transaction type of the transaction can be determined, and the subsequent evaluation can be performed by combining the object range of the transaction reference object with the association relation with the transaction type of the transaction when evaluating the risk type of the transaction, so that the transaction evaluation efficiency can be improved.
S103, calculating a first transaction evaluation parameter associated with each object range based on the transaction type of each reference transaction and the object range hit by the transaction reference information of each reference transaction in at least one object range.
In an embodiment of the present application, for any one of at least one object scope, a first transaction evaluation parameter associated with the any one object scope is used to evaluate a probability that a transaction type of any transaction is an abnormal transaction type. That is, by calculating the first transaction evaluation parameter associated with each object range, the probability that the transaction type of the transaction is an abnormal transaction type can be determined, thereby determining whether the transaction in that object range is an abnormal transaction. It will be appreciated that after determining the probability that the transaction type of the transaction is an abnormal transaction type, the probability that the transaction type of the transaction is a normal transaction type may also be determined, thereby determining the transaction type of the transaction based on the probability of the abnormal transaction type and the probability of the normal transaction type. If the probability of the abnormal transaction type is greater than the probability of the normal transaction type, determining that the transaction type of the transaction is the abnormal transaction type. If the probability of the abnormal transaction type is smaller than the probability of the normal transaction type, determining that the transaction type of the transaction is the normal transaction type.
In one embodiment, any one of the at least one object scope is represented as a target object scope, and a process of calculating the first transaction evaluation parameter associated with the target object scope is described below:
determining a reference transaction of which transaction reference information in the plurality of reference transactions hits a target object range as a target reference transaction; counting the target quantity of target reference transactions with the transaction type being the normal transaction type in at least one target reference transaction; a ratio between the target number and the total number of at least one target reference transaction is determined as a first transaction evaluation parameter associated with the target object range.
Describing the reference transactions listed in table 2, the target object range is, for example, [500, 1000), 3 reference transactions hit [500,1000 ] among 10 reference transactions are determined as target reference transactions, the target number of target reference transactions of which the transaction type is the normal transaction type among the 3 target reference transactions is counted as 1, and the ratio of 1/3 between the target number and the total number of target reference transactions is determined as the first transaction evaluation parameter associated with the target object range [500, 1000). The larger the first transaction evaluation parameter associated with a target object range, the greater the probability that a transaction belonging to the target object range is an abnormal transaction type. The first transaction evaluation parameter associated with the target object range may be used to evaluate a probability that a transaction type of a transaction belonging to the target object range is an abnormal transaction type, e.g., the smaller the first transaction evaluation parameter associated with the target object range, the smaller the probability that a transaction belonging to the target object range is an abnormal transaction type.
The first transaction evaluation parameter associated with the target object range may refer to a conditional probability that the transaction type of the transaction is an abnormal transaction type under the condition that the transaction reference object is the target object range, and by using the method, the probability that the transaction type of the transaction is the abnormal transaction type under each condition can be calculated. It will be appreciated that, for at least one object range other than the target object range, the first transaction evaluation parameter associated with each object range may also be calculated with reference to the process, which is not described in any greater detail in the embodiments of the present application.
Alternatively, a bayesian formula may be used to calculate a first transaction evaluation parameter associated with the target object range, that is, a parameter of a conditional probability table of the transaction type under the condition of a plurality of object ranges corresponding to the resource transfer amounts of the transactions. For example, if the transaction reference object is the resource transfer amount of the transaction, the object range of the transaction reference object may be a numerical range of the resource transfer amount, and parameters of the conditional probability table for calculating the transaction type under the condition of a plurality of object ranges corresponding to the resource transfer amount of the transaction are respectively as follows:
/>
wherein X is 3 =0 represents normal transaction type, X 3 =1 denotes the abnormal transaction type, X 1 An object range (interval) representing that the transaction reference object is the resource transfer amount of the transaction, such as X 1 E [0,500), the object range at this time is the resource transfer amount range [0,500 ] (representing a range of 0 to 500).The conditional probability of each transaction type under the condition that the amount of the resource transfer of the transaction belongs to a plurality of intervals is represented, namely, the probability that the transaction type of the transaction is a normal transaction type or an abnormal transaction type under the condition that the amount of the resource transfer of the transaction belongs to a certain interval is represented, namely, the probability that the amount of the resource transfer of the transaction has a first transaction evaluation parameter associated with each object range in a plurality of object ranges. Here, a->The subscript of the parameter has two parameters, and the parameter on the left may represent the sequence number of the interval of the resource transfer amount. For example, the total interval of the resource transfer amount is [0, + -infinity), and the resource transfer amount can be divided into a plurality of intervals of [0,500 ], [500,1000 ], [1000,1500 ], [1500, + -infinity), and each interval can be provided with a corresponding sequence number, for example, the sequence number of [0,500) can be provided0, [500, 1000) may be 1, [1000, 1500) may be 2, [1500, ++ infinity) may be 3. The right parameter may represent the transaction type (e.g., the right parameter is 1, the right parameter is 0, the right parameter is a normal transaction type), for example, two parameters of the subscript are (0, 0) to indicate the resource transfer amount interval is [0,500), and the transaction type is a normal transaction type. Two parameters of the subscript are (1, 0) to indicate that the resource transfer amount interval is [500, 1000), and the transaction type is a normal transaction type. Two parameters of the subscript are (2, 0) to indicate that the resource transfer amount interval is [1000, 1500), and the transaction type is a normal transaction type. Two parameters of the subscript are (2, 1) to indicate that the resource transfer amount interval is [1000, 1500), and the transaction type is an abnormal transaction type. Two parameters of the subscript are (3, 0) which indicates that the resource transfer amount interval is [1500, ++ ] and the transaction type is a normal transaction type; two parameters of the subscript are (3, 1) to indicate that the resource transfer amount interval is [1500, ++ -infinity), and the transaction type is an abnormal transaction type.
For example, the number of the cells to be processed,may be expressed as a probability that the transaction is of a normal transaction type, provided that the amount of resource transfer of the transaction belongs to [0,500). />May be expressed as a probability that the transaction is of a normal transaction type, provided that the amount of resource transfer of the transaction belongs to [500,1000 ]. />May be expressed as a probability that the transaction is of a normal transaction type, provided that the amount of resource transfer for the transaction belongs to [1000,1500 ]. />May be expressed as a probability that the transaction is of an abnormal transaction type, provided that the amount of resource transfer for the transaction belongs to [1000,1500 ]. />It may be indicated that the amount of resource transfer at the transaction belongs to 1500, + -infinity), the probability that the transaction is of the normal transaction type.
Optionally, if the transaction reference object is the time of the transaction, the object range of the transaction reference object may be the time range, and the parameters of the conditional probability table of the transaction type of the transaction under the condition of the initiation time of the transaction may be calculated by the method as follows:
/>
in the embodiment of the application, X 2 Representing a transaction reference object as an object range (interval), e.g. X, of the time of initiation of a transaction 2 E [0, 8), the object range at this time is time range [0,8 ] (representing the range from 0 point to 8 points). Conditions representing a plurality of intervals to which the time of initiation of a transaction belongsThe conditional probability of each transaction type can be expressed as the probability that the transaction type of the transaction is a normal transaction type or an abnormal transaction type under the condition that the initiating time of the transaction is a certain interval, namely, the first transaction evaluation parameter associated with each object range in a plurality of object ranges of the initiating time of the transaction. Here, a->The subscript of the parameter has two parameters, the parameter on the left may represent the sequence number of the interval of the initiation time. For example, the total interval of the initiation time is [0,24 ], which can be divided into a plurality of intervals of [0,8 ], [8,16 ], [16, 24), and each interval can be provided with a corresponding sequence number, for example, the sequence number of [0,8 ] can be 0, the sequence number of [8, 16) can be 1, and the sequence number of [16, 24) can be 2. The right parameter may represent the transaction type (e.g., a right parameter of 1, an abnormal transaction type, a right parameter of 0, a normal transaction type). Two parameters, such as the subscript, are (0, 0) indicating an initiation time interval of [0, 8), and the transaction type is a normal transaction type. Two parameters of the subscript are (1, 0) indicating an initiation time interval of [8, 16), and the transaction type is a normal transaction type. Two parameters of the subscript are (2, 1) indicating an initiation time interval of [16, 24), and the transaction type is an abnormal transaction type.
For example, the number of the cells to be processed,may be expressed as a probability that the transaction is of a normal transaction type, provided that the time of initiation of the transaction belongs to [0,8 ]. />May be expressed as a probability that the transaction is of an abnormal transaction type, provided that the time of initiation of the transaction belongs to [0,8 ]. />May be expressed as a probability that the transaction is of an abnormal transaction type, provided that the time of initiation of the transaction belongs to [16, 24).
By the method, the first transaction evaluation parameters associated with a plurality of object ranges of any kind of transaction reference objects can be calculated, for example, the first transaction evaluation parameters associated with a plurality of object ranges and provided at the time of initiating a transaction, the first transaction evaluation parameters associated with a plurality of object ranges and provided at the time of transferring a transaction resource, the first transaction evaluation parameters associated with a plurality of object ranges and provided at the geographic position of the transaction initiating object, the first transaction evaluation parameters associated with a plurality of object ranges and provided at the geographic position of the resource issuing object, and the like can be calculated respectively. By calculating the first transaction evaluation parameters associated with the object ranges of each type of transaction reference object, the influence of the object ranges of each type of transaction reference object on the transaction type of the transaction can be determined, so that when a new transaction is acquired, the transaction evaluation parameters of the new transaction can be determined by combining the transaction reference information of the new transaction with the first transaction evaluation parameters associated with the object ranges of each type of transaction reference object, and the probability that the new transaction is of an abnormal transaction type can be determined. If the new transaction is of an abnormal transaction type, the new transaction can be further processed such as transaction interception, transaction quota and the like, so that transaction loss is reduced.
Referring to fig. 3, fig. 3 is a schematic view of an object range of a transaction reference object according to an embodiment of the present application, as shown in fig. 3, for example, the transaction reference object may include 3 types, each of which has a plurality of object ranges. For example, the transaction reference object includes the time of initiation of the transaction, the amount of resource transfer of the transaction, and the geographic location of the resource issuing object (i.e., the location of the merchant). Wherein the initiation time of the transaction may include 3 object ranges of [0,8 ], [8, 16), and [16,24 ], respectively. The resource transfer amount of a transaction may include 4 object ranges of [0,500 ], [500,1000 ], [1000, 1500) and [1500, ++ infinity A kind of electronic device. The geographical location where the resource issuing object is located may include 2 object ranges of "Chongqing" and "Shenzhen", respectively. By the method, the first transaction evaluation parameter associated with each object range can be calculated by combining a plurality of reference transactions, for example, the first transaction evaluation parameter associated with the object range [0,8 ] of the initiating time of the transaction is 0, the first transaction evaluation parameter associated with the object range [8,16 ] of the initiating time of the transaction is 0.5, and the first transaction evaluation parameter associated with the object range [16, 24) of the initiating time of the transaction is 0.33. The first transaction evaluation parameter associated with the object range [0,500 ] of the resource transfer amount of the transaction is 0, the first transaction evaluation parameter associated with the object range [500,1000 ] of the resource transfer amount of the transaction is 0.67, the first transaction evaluation parameter associated with the object range [1000, 1500) of the resource transfer amount of the transaction is 0.75, and the first transaction evaluation parameter associated with the object range [1500, + ] of the resource transfer amount of the transaction is 0. The first transaction evaluation parameter associated with the object range of Chongqing is 0.4, and the first transaction evaluation parameter associated with the object range of Shenzhen is 0.6.
In an alternative implementation, after constructing the causal graph, a predictive model may be utilized to determine a conditional probability distribution for each node in the causal graph, i.e., to determine a conditional probability that the transaction type is an abnormal transaction type under a plurality of object ranges for a plurality of transaction reference objects, i.e., to determine a first transaction evaluation parameter associated with each object range using the predictive model. Specifically, the prediction model may be trained, where the purpose of the training of the prediction model is to determine a conditional probability distribution of each node, and optimize parameters of the prediction model, and model training may be implemented, for example, by using methods such as maximum likelihood estimation or bayesian estimation. Taking the maximum likelihood estimation as an example, let the training set be d= { x 1 ,x 2 ,…x n X, where x i ＝(t i ,m i ,a i ,l i ,s i ,c i ) Representing transaction data and related data and transaction types. t is t i Representing the initiation time of a transaction, m i Representing the amount of resource transfer of a transaction, a i Indicating the geographic location of the transaction initiating object of the transaction, l i Representing object types of resource release objects s i Representing the geographic location of the resource issuing object, c i Indicating the registration time of the resource issuance object. The parameters of the predictive model may be estimated by maximizing the likelihood function of the training set, which may be expressed as:
Wherein L (θ) represents a first transaction evaluation parameter, i.e., a conditional probability of transaction type under each variable, θ represents a model parameter, x i Representing the variables, i.e. the object range, the maximization of the likelihood function can be achieved by maximum likelihood estimation. The goal of maximum likelihood estimation is to find the parameter values that maximize the likelihood function. For example, the parameter value of the maximum likelihood estimate may be obtained by solving the partial derivative of the likelihood function with respect to the parameter and making the partial derivative 0. The conditional probability distribution of the respective variables can be determined by likelihood functions.
In the embodiment of the application, the first transaction evaluation parameters associated with the object ranges of each type of transaction reference object, namely the conditional probability distribution of each variable, can be respectively determined through a prediction model, and the influence of the object ranges of each type of transaction reference object on the transaction type of the transaction can be determined, so that when a new transaction is acquired, the transaction evaluation parameters of the new transaction can be determined by combining the transaction reference information of the new transaction with the first transaction evaluation parameters associated with the object ranges of each type of transaction reference object, and the probability that the new transaction is of an abnormal transaction type can be determined.
Optionally, after the training of the prediction model is completed, the prediction model may be evaluated to evaluate the prediction accuracy and robustness of the prediction model. Model evaluation may be implemented using cross-validation or the like. The cross-validation is a common model evaluation method, and the performance of the prediction model is evaluated by dividing a training set into a plurality of subsets, using one subset as a test set each time, using the rest of the subsets as the training set, repeating for a plurality of times and taking an average value. Alternatively, model evaluation may be accomplished using a sensitivity analysis method, where the purpose of sensitivity analysis is to analyze how sensitive the model output is to the model input. The sensitivity analysis may be implemented using a variety of methods, such as a variable elimination method, a variable addition method, a parameter sensitivity analysis, and the like. Taking the variable elimination method as an example for explanation, the variable elimination method is used for analyzing the change condition of the model output by eliminating one or more variables. For example, one or more variables may be excluded, excluding a variable may mean excluding a parameter corresponding to the variable in the calculation formula, excluding data corresponding to the variable from the data set, excluding a node corresponding to the variable in the causal graph, reestablishing a prediction model, calculating a posterior probability distribution of a transaction type, comparing the posterior probability distribution with a result of an original model (a prediction model before excluding the variable), and analyzing a change condition of model output, thereby evaluating the model according to the change condition of model output and improving accuracy and robustness of model prediction.
In the embodiment of the application, the transaction reference object is obtained, and is provided with at least one object range for evaluating the transaction type of the transaction, wherein each object range has an association relationship with the transaction type of the transaction; by acquiring the transaction attribute information of the plurality of reference transactions, the first transaction evaluation parameter associated with each object range may be calculated based on the transaction type of each reference transaction and the object range in which the transaction reference information of each reference transaction hits in at least one object range, respectively. The first transaction evaluation parameter may be used to evaluate a probability that a transaction type of the transaction is an abnormal transaction type. According to the method, through calculating the first transaction evaluation parameters associated with the object ranges of the transaction reference object, the probability that the transaction type of the transaction is the abnormal transaction type under the condition of each object range can be determined, namely the probability that the transaction is the abnormal transaction type under the condition of each object range, and then various factors affecting the transaction type can be determined. For example, the method and the device for determining the transaction type can influence the transaction type by determining which object ranges of which transaction reference objects have a larger probability of being of an abnormal transaction type under the condition of which object ranges of which transaction reference objects and a smaller probability of being of an abnormal transaction type under the condition of which object ranges of which transaction reference objects, so that the transaction type of the transaction can be accurately evaluated through the first transaction evaluation parameters associated with each object range of the transaction reference objects.
Optionally, referring to fig. 4, fig. 4 is a flow chart of another transaction data processing method according to an embodiment of the present application. The transaction data processing method can be applied to computer equipment; as shown in fig. 3, the transaction data processing method includes, but is not limited to, the following steps:
s201, acquiring a transaction reference object.
S202, transaction attribute information of a plurality of reference transactions is acquired.
S203, calculating a first transaction evaluation parameter associated with each object scope based on the transaction type of each reference transaction and the object scope hit by the transaction reference information of each reference transaction in at least one object scope.
In the embodiment of the present application, the specific implementation manner of step S201 to step S203 may refer to the implementation manner of step S101 to step S103, which will not be described herein. The first transaction evaluation parameters associated with each object range can be calculated in the mode. For example, when the types of the transaction reference objects are multiple, and each transaction reference object corresponds to multiple object ranges, the method can obtain the first transaction evaluation parameters associated with the multiple object ranges corresponding to each transaction reference object in the multiple transaction reference objects.
S204, acquiring a plurality of transaction reference information of the target transaction.
In the embodiment of the present application, the plurality of transaction reference information of the target transaction may include, but is not limited to, a time of initiation of the target transaction, a resource transfer amount of the target transaction, a geographic location where a transaction initiation object of the target transaction is located, an object type of a resource release object to which a service resource to be acquired by the target transaction belongs, or a geographic location where the resource release object is located, and so on. For example, the initiation time of the target transaction may be 10 and the resource transfer amount of the target transaction may be 800.
The transaction reference objects can be multiple, and each transaction reference object has at least one object range. The type of a transaction reference information of the target transaction matches the type of a transaction reference object. For example, the type of transaction reference object may include 2 types, a resource transfer amount of the transaction and an initiation time of the transaction, the resource transfer amount of the transaction has 4 object ranges of [0,500 ], [500,1000 ], [1000,1500 ], [1500, + ], respectively, and the initiation time of the transaction has 3 object ranges of [0,8 ], [8, 16) and [16, 24), respectively. The target transaction comprises 2 transaction reference information, wherein the initiation time of the target transaction is 10, and the resource transfer amount of the target transaction is 800.
S205, an object range in which each transaction reference information of the target transaction hits in at least one object range of the matched transaction reference objects is acquired as a hit object range of each transaction reference information.
In an embodiment of the present application, the transaction reference information of the target transaction includes, for example, a target transaction initiation time of 10 and a target transaction resource transfer amount 800. For example, the transaction reference information of the target transaction includes that the initiation time of the target transaction is 10, the transaction reference object matched with the transaction reference information may refer to the initiation time of the transaction, the initiation time of the transaction may have 3 object ranges of [0,8 ], [8, 16), and [16,24 ], respectively), and then the object range hit by the initiation time of the target transaction is [8, 16), and the object range [8, 16) is taken as the hit object range of the reference transaction information. For example, the transaction reference information of the target transaction includes that the resource transfer amount of the target transaction is 800, the transaction reference object matched with the transaction reference information may refer to the resource transfer amount of the transaction, the resource transfer amount of the transaction may have 4 object ranges of [0,500 ], [500,1000 ], [1000, 1500), [1500, + ], respectively, the object range hit by the resource transfer amount of the target transaction is [500, 1000), and the object range [500, 1000) is taken as the hit object range of the reference transaction information. The hit range of each transaction reference information of the target transaction includes the initiation time [8, 16) and the resource transfer amount [500, 1000).
In one possible implementation, if each transaction reference information of the target transaction does not match a node in the causal graph, that is, each transaction reference information of the target transaction is not acquired, that is, an object range hit by each transaction reference information of the target transaction in at least one object range of the matched transaction reference object, it may be directly determined that the target transaction evaluation parameter of the target object is less than the parameter threshold. Because the causal graph can comprise transaction type nodes and other nodes which have causal relation with the transaction type nodes, if the transaction reference information of the target transaction does not exist in the causal graph, the probability that the target transaction is of an abnormal transaction type is smaller, the transaction reference information of the target transaction can be directly determined to be of a normal transaction type without adopting a prediction model to carry out subsequent processing, and the transaction risk prediction efficiency can be improved.
If each transaction reference information of the target transaction is matched with the node in the causal graph, that is, the object range of each transaction reference information of the target transaction, which is hit in at least one object range of the matched transaction reference object, is acquired, the target transaction evaluation parameters of the target transaction can be calculated continuously based on the first transaction evaluation parameters associated with the hit object range of the transaction reference information of the target transaction.
S206, calculating target transaction evaluation parameters of the target transaction based on the first transaction evaluation parameters associated with the hit object range of each transaction reference information.
In the embodiment of the application, since the first transaction evaluation parameters associated with the object ranges corresponding to each transaction reference object in the plurality of transaction reference objects are calculated, the target transaction evaluation parameters of the target transaction can be calculated based on the first transaction evaluation parameters associated with the hit object ranges of each transaction reference information. The target transaction evaluation parameters of the target transaction, i.e. the conditional probabilities of the transaction types, may be calculated, for example, using bayesian formulas of bayesian networks. Wherein the target transaction evaluation parameter is a probability that the transaction type of the target transaction is an abnormal transaction type.
Referring to fig. 5, fig. 5 is a schematic view of a scenario for determining a target transaction evaluation parameter of a target transaction according to an embodiment of the present application, as shown in fig. 5, for example, transaction reference objects may include 3 types, each of which has a plurality of object ranges. For example, the transaction reference object includes a transaction initiation time, a transaction resource transfer amount and a geographic location of the resource issuing object, and the transaction initiation time has a first transaction evaluation parameter associated with a range of 3 objects of 0, 0.5 and 0.33, respectively. The first transaction evaluation parameters associated with the 4 object ranges of the resource transfer amount of the transaction are 0, 0.67, 0.75 and 0 respectively. The first transaction evaluation parameters associated with the 2 object ranges of the geographic position where the resource issuing object is located are respectively 0.4 and 0.6. The transaction reference information of the target transaction is the resource transfer quantity 800, the initiation time 10 and the geographical position 'Shenzhen' where the resource issuing object is located respectively, and then the transaction reference information of the target transaction is the object range [500,1000 ] hit by the resource transfer quantity 800, namely the object range indicated by the dotted line, and the first transaction evaluation parameter associated with the object range [500,1000 ] is 0.67. The transaction reference information for the target transaction is object range [8,16 ] hit at the initiation time 10, and the first transaction evaluation parameter associated with the object range [8,16 ] is 0.5. The target transaction evaluation parameters of the target transaction can be calculated based on 3 first transaction evaluation parameters (0.67, 0.5 and 0.6) when the transaction reference information of the target transaction is that the hit object range of the geographic position Shenzhen where the resource issuing object is located is Shenzhen and the first transaction evaluation parameter associated with the object range Shenzhen is 0.6.
In one embodiment, a method of calculating a target transaction evaluation parameter for a target transaction based on a first transaction evaluation parameter associated with a hit range for each transaction reference information may include: calculating a first product between first transaction evaluation parameters associated with hit ranges of each transaction reference information; acquiring first priori information of a plurality of reference transactions, and calculating a second product between the first product and the numerical value of the first priori information; the first priori information is used for indicating the proportion of the reference transaction with the abnormal transaction type in the plurality of reference transactions; a target transaction evaluation parameter is determined based on the second product.
For example, the first transaction evaluation parameter associated with the hit range of the initiation time of the target transaction is P (t|r), the first transaction evaluation parameter associated with the hit range of the resource transfer amount of the target transaction is P (m|r), then a first product between the first transaction evaluation parameters associated with the hit range of each transaction reference information is P (t|r) ×p (m|r), for example, the first priori information of the plurality of reference transactions is P (R), then a second product between the first product and the value of the first priori information is P (t|r) ×p (m|r) ×p (R), and then the target transaction evaluation parameter may be determined based on the second product.
Because the first transaction evaluation parameters associated with the hit object range of each transaction reference information of the target transaction are combined when the target transaction evaluation parameters of the target transaction are calculated, the obtained target transaction evaluation parameters can reflect the probability that the target transaction is of an abnormal transaction type under various conditions, the accuracy of risk prediction can be improved, and the accuracy of determining transaction data is further improved.
In one embodiment, the first priori information for the plurality of reference transactions may be obtained in the following manner. Specifically, the number of reference transactions with abnormal transaction types in the plurality of reference transactions can be counted; the ratio between the number of reference transactions of the transaction type abnormal transaction type and the total number of the plurality of reference transactions is determined as the first priori information.
For example, the total number of the plurality of reference transactions is 10, and the reference transaction of which the transaction type is an abnormal transaction type is 6 among the 10 reference transactions, the first priori information may be 6/10=0.6. The first priori information may refer to, for example, a priori probability, i.e., a probability obtained from past experience and analysis, and may be determined, for example, from the total number of reference transactions and the number of reference transactions of the abnormal transaction type in the reference transactions. By determining the prior probability, the posterior probability, namely the target transaction evaluation parameter of the target transaction, can be obtained by combining the prior probability and the conditional probability, and the accuracy of determining the posterior probability is improved.
In one embodiment, a method of determining a target transaction evaluation parameter based on a second product may include: calculating a second transaction evaluation parameter associated with the hit object range of each transaction reference information based on the first transaction evaluation parameter associated with the hit object range of each transaction reference information; calculating a third product between the second transaction evaluation parameters associated with the hit range of each transaction reference information; acquiring second priori information of the plurality of reference transactions, and calculating a fourth product between the third product and the numerical value of the second priori information; and acquiring the sum value of the second product and the fourth product, and determining the ratio between the second product and the sum value as a target transaction evaluation parameter. Wherein, any hit target range associated second transaction evaluation parameter is used for evaluating the probability that the transaction type of the target transaction is an abnormal transaction type; the second prior information is used to indicate a proportion of reference transactions of the plurality of reference transactions for which the transaction type is a normal transaction type.
In the embodiment of the present application, the second transaction evaluation parameter associated with the hit object range of each transaction reference information is obtained by calculation, and since the target transaction may include a plurality of transaction reference information, for example, including the initiation time of the target transaction and the resource transfer amount of the target transaction, the second transaction evaluation parameter associated with the hit object range of the initiation time of the target transaction is P (t|r), the second transaction evaluation parameter associated with the hit object range of the resource transfer amount of the target transaction is P (m|r), and then the third product between the second transaction evaluation parameters associated with the hit object range of the plurality of transaction reference information included in the target transaction is P (t|r), for example, the second priori information is P (1-R). The fourth product is P (t|r) P (m|r) P (1-R). The sum of the second product and the fourth product is P (t|r) P (m|r) P (R) +p (t|r) P (m|r) P (1-R), and the ratio between the second product and the sum is determined as the target transaction evaluation parameter, i.e., the target transaction evaluation parameter is [ P (t|r) P (m|r) P (R) ]/[ P (t|r) P (m|r) P (R) +p (t|r) P (m|r) P (1-R) ].
In one embodiment, the second prior information for the plurality of reference transactions may be obtained in the following manner. Specifically, the number of reference transactions with the normal transaction type in the transaction types in the plurality of reference transactions can be counted; the ratio between the number of reference transactions of the transaction type normal to the total number of the plurality of reference transactions is determined as the second prior information.
For example, the total number of the plurality of reference transactions is 10, and the reference transaction of which the transaction type is the normal transaction type is 4 among the 10 reference transactions, the second priori information may be 4/10=0.4. The second prior information may for example refer to a prior probability, i.e. a probability obtained from past experience and analysis, e.g. may be determined from the total number of reference transactions and the number of reference transactions of the normal transaction type in the reference transactions. By determining the prior probability, the posterior probability, namely the target transaction evaluation parameter of the target transaction, can be obtained by combining the prior probability and the conditional probability, and the accuracy of determining the posterior probability is improved.
Illustratively, taking the transaction reference object including the initiation time of the transaction and the resource transfer amount of the transaction as an example, the transaction reference information of the target transaction includes the initiation time of the target transaction and the resource transfer amount of the target transaction, the conditional probability of the transaction type may be calculated using a bayesian formula of a bayesian network, where the bayesian formula may be as follows:
Wherein, P (R|T, M) can represent posterior probability, i.e. probability that the transaction is an abnormal transaction type and a normal transaction type respectively, R ε 0 represents the transaction is a normal transaction type, R ε 1 represents the transaction is an abnormal transaction type. P (R) may represent prior information such as first prior information or second prior information, P (t|r) may refer to a conditional probability distribution, i.e., a first transaction evaluation parameter corresponding to an initiation time of a transaction, and P (m|r) may refer to a conditional probability distribution, i.e., a first transaction evaluation parameter corresponding to a resource transfer amount of a transaction.
Optionally, if the transaction reference information of the target transaction further includes the geographic location where the transaction initiating object is located, P (a|r) may also be included in the bayesian formula. If the transaction reference information of the target transaction further includes an object type of the resource release object, P (l|r) may also be included in the bayesian formula. If the transaction reference information of the target transaction further includes the geographic location where the resource issuing object is located, the bayesian formula may further include P (s|r). If the transaction reference information of the target transaction further includes the registration time of the resource issuing object, P (c|r) may be further included in the bayesian formula, that is, P (t|r), P (m|r), P (a|r), P (l|r), P (s|r), P (c|r) represent the conditional probability distribution of each variable under the condition that the transaction risk is 0 or 1.
For example, the Bayes' formula containing P (T|R), P (M|R), P (A|R), P (L|R), P (S|R), P (C|R) can be converted into:
for example, the plurality of transaction reference information for the target transaction includes an initiation time of the target transaction and a resource transfer amount of the target transaction, for example, the initiation time of the target transaction is 10 and the resource transfer amount of the target transaction is 800, so that a probability that the target transaction is an abnormal transaction type under the two transaction reference information can be predicted.
Specifically, since the resource transfer amount 800 belongs to the [500,1000 ] interval, the initiation time 10 of the transaction belongs to the [8, 16) interval, the target transaction evaluation parameter of the target transaction can be calculated through the bayesian formula, and the data is substituted into the bayesian formula to obtain:
since the initiation time of the transaction and the resource transfer amount of the transaction are direct causes affecting the transaction type, the bayesian formula can be brought into the calculated first transaction evaluation parameter associated with the object range of the transaction reference object, so as to obtain:
that is, in the case where the transaction amount of the target transaction is 800 and the initiation time of the transaction is [8,16 ], the target transaction evaluation parameter is 0.667, i.e., the probability that the transaction type of the target transaction is an abnormal transaction type is 0.667. The target transaction evaluation parameter may be further compared with a probability threshold, e.g., if the target transaction evaluation parameter is greater than the probability threshold, then the target transaction is considered to be of an abnormal transaction type; and if the target transaction evaluation parameter is smaller than or equal to the probability threshold, the target transaction is considered to be of a normal transaction type. It will be appreciated that the probability that the transaction type of the target transaction is a normal transaction type may also be calculated by the bayesian formula described above, and so on.
In the embodiment of the application, after causal relation analysis is completed, a causal graph is utilized to construct a Bayesian network model, namely a prediction model, which can be used for predicting the risk of a merchant, namely predicting the probability that a transaction is an abnormal transaction type. The bayesian network model is a graph model for representing the dependency relationship between variables, and is inferred by using bayesian formulas in probability theory, and can be used for describing the dependency relationship between random variables. In the embodiment of the application, a Bayesian network can be adopted to build a prediction model, a causal graph is utilized to build a Bayesian network model, and a training set is utilized to train and optimize the model. The purpose of model training is to determine the conditional probability distribution for each node and optimize the parameters of the model. The steps of bayesian network modeling may be as follows:
node selection: before the bayesian network model is built, nodes in the bayesian network model can be determined, and the nodes can be transaction reference objects (i.e. transaction data and related data) such as the resource transfer amount of a transaction, the initiation time of the transaction, the geographic position of the transaction initiation object of the transaction, the object type of the resource release object, the geographic position of the resource release object, the registration time of the resource release object and the like.
Determination of a conditional probability distribution: after determining the nodes, it is necessary to determine a conditional probability distribution of each node, so that a posterior probability distribution of the risk of the merchant, that is, a probability that the transaction is of an abnormal transaction type, is calculated based on the conditional probability distribution of each node. The conditional probability distribution represents the conditional probability distribution of the values of a certain variable (node) under certain conditions. For example, the conditional probability distribution of each node may be calculated by the bayesian formula, or the conditional probability distribution of each node may be determined by using the likelihood function in the prediction model, or other tools or statistical methods may be used to determine the conditional probability distribution of each node, that is, the conditional probability distribution between the transaction type node and other cause nodes, which is not limited in the embodiment of the present application. After the conditional probability distribution of each node is calculated, the prior probability and the conditional probability of each node can be combined to calculate the posterior probability, namely, after the conditional probability distribution of each node is calculated, the posterior probability can be calculated through the Bayesian formula, and the posterior probability can indicate the probability that the target transaction is of an abnormal transaction type, so that the target transaction evaluation parameter of the target transaction, namely, the probability that the target transaction is of the abnormal transaction type, can be output through the model.
In the embodiment of the application, after the Bayesian network model is constructed based on the causal graph, that is, after the prediction model is constructed, the probability that the target transaction is of an abnormal transaction type can be predicted by adopting the prediction model, so as to determine whether the target transaction has risk. Because the prediction model is constructed based on the causal relationship in the causal graph, and the causal graph can reflect the dependency relationship between different nodes, the accuracy of determining the transaction risk result can be improved.
In the embodiment of the application, after the training and optimization of the prediction model are completed, the prediction model can be applied to real-time transaction data processing and analysis so as to realize real-time prediction of the transaction type. The real-time data processing and analysis includes the following steps: data collection, i.e., collecting transaction data and related data for a target transaction in real-time. And data preprocessing, namely cleaning and preprocessing the collected transaction data and related data of the target transaction, so as to ensure the accuracy and the integrity of the data. Data feature extraction, i.e., extracting features from the collected data for use in a predictive model. The transaction type prediction, i.e. predicting the target transaction in real time by using a trained prediction model, for example, taking a plurality of transaction reference information, such as transaction data, and related data, of the target transaction collected in real time as the input of the prediction model, calculating the conditional probability distribution of each node through the prediction model, and further calculating the posterior probability distribution of the target transaction, for example, calculating the target transaction evaluation parameter of the target transaction, thereby determining whether the transaction type of the target transaction is an abnormal transaction type.
In the embodiment of the application, transaction attribute information of a transaction reference object and a plurality of reference transactions can be acquired, and a prediction model constructed through a Bayesian network is utilized to calculate a first transaction evaluation parameter associated with each object range based on the transaction type of each reference transaction and the object range of each transaction reference information hit in at least one object range. Further, a plurality of transaction reference information of the target transaction can be acquired, and an object range, in which each transaction reference information of the target transaction hits in at least one object range of the matched transaction reference object, is acquired by utilizing a prediction model constructed based on a Bayesian network and is used as a hit object range of each transaction reference information; and further, calculating target transaction evaluation parameters of the target transaction based on first transaction evaluation parameters associated with hit object ranges of each transaction reference information, so that the target transaction evaluation parameters of the target transaction are output through a prediction model constructed based on a Bayesian network, and further, the probability that the target transaction is of an abnormal transaction type can be determined, and therefore whether the target transaction has risks is determined.
In one embodiment, the corresponding operations may be performed based on the predicted outcome of the target transaction. Specifically, if the value of the target transaction evaluation parameter is greater than or equal to the parameter threshold, stopping execution of the target transaction, and prompting the transaction initiation object that the resource issuing object has an abnormality. The target transaction is a transaction initiated by a transaction initiating object to a resource issuing object, and the transaction initiating object is used for acquiring a business resource held by the resource issuing object based on the initiated target transaction. For example, the object a initiates a transaction to the object b, where the transaction is online purchasing of a commodity, the object a may refer to a transaction initiating object, the object b may refer to a resource issuing object, and after the object a initiates the transaction, the object b pays the resource corresponding to the commodity, such as an amount corresponding to the commodity, and the object b issues the service resource, such as the commodity, to the object a.
It may be appreciated that if the value of the target transaction evaluation parameter is greater than or equal to the parameter threshold, it may indicate that the target transaction is of an abnormal transaction type, and indicate that the target transaction is at risk, then the execution of the target transaction may be stopped, so as to avoid causing a loss to the transaction initiating object. Stopping execution of the target transaction may refer to stopping the transaction from continuing, such as stopping the transfer of the transaction initiating object to the resource issuing object. Alternatively, if the value of the target transaction evaluation parameter is greater than or equal to the parameter threshold, a quota process may be performed on the transaction, i.e., the amount of the resource transfer amount of the target transaction is limited, for example, the amount of the resource transfer amount is adjusted from 1000 to 500 or less, so as to reduce the loss of the transaction initiating object. By prompting the transaction initiating object that the resource issuing object is abnormal, the transaction initiating object can acquire the reasons of the transaction abnormality, and loss is reduced.
In the embodiment of the application, after the transaction risk prediction is completed, risk assessment and decision implementation can be carried out on the transaction, corresponding decision suggestion is provided, and transaction loss is reduced. The implementation decision is a process of quickly deciding the current situation according to a preset rule and model on the basis of real-time data processing and analysis. Specifically, for example, risk assessment may be performed on a transaction according to a predicted result of the transaction, where the risk assessment may refer to determining whether the transaction is at risk, for example, a target transaction assessment parameter of the transaction is greater than a parameter threshold, and may indicate that a transaction type of the transaction is an abnormal transaction type. The transaction can be further judged, for example, the target transaction evaluation parameter of the transaction is sent to an evaluation mechanism, the evaluation mechanism is used for further evaluating the transaction to determine whether the transaction is of an abnormal transaction type, and the accuracy of the transaction evaluation can be improved. Further, decision advice can be further performed on the transaction, the decision advice can be that the pointer provides corresponding decision advice for risk situations of the transaction, for example, the transaction is of an abnormal transaction type, and risk control measures such as limiting transaction amount, suspending the transaction and the like can be adopted, so that transaction loss is reduced, and transaction safety is improved.
In the embodiment of the application, in the transaction type prediction, the risk of future transactions is predicted and evaluated by analyzing and modeling historical transaction data and other related data. By establishing a causal graph, the accuracy and the credibility of risk prediction can be improved by taking causal relations and influence factors among different variables into consideration. By analyzing and modeling the causal relationship among things, the reason and the result of occurrence of a certain event are deduced, and the prediction accuracy can be improved. In the model prediction process, the causal relationship is modeled and optimized by using a Bayesian network, so that the fitting property and the prediction capability of the model can be improved. In the transaction prediction process, based on the established causal relationship model and the Bayesian network model, transaction data can be analyzed and processed in real time, and real-time decision is supported. Because the causal relationship and the influencing factors are comprehensively considered, the accuracy and the credibility of the transaction type prediction can be improved. The embodiment of the application can realize real-time data processing and analysis and provide powerful support for real-time decision making of transactions. In practical application, the embodiment of the application can be applied to various scenes needing transaction prediction, such as but not limited to scenes of online shopping websites, paytables, financial platforms and the like, and scenes of transaction type prediction can be enriched.
In the embodiment of the application, by respectively calculating the first transaction evaluation parameters associated with the object ranges of the transaction reference objects, when a new transaction such as a target transaction is acquired, the first transaction evaluation parameters associated with the hit object ranges of each transaction reference information of the target transaction can be determined by combining the object ranges of the transaction reference objects hit by each transaction reference information of the target transaction, so that the target transaction evaluation parameters of the target transaction are calculated based on the first transaction evaluation parameters associated with the hit object ranges of the transaction reference information of the target transaction. That is, by searching which object ranges the transaction reference information of the target transaction belongs to, the target transaction evaluation parameter of the target object can be determined according to the first transaction evaluation parameter corresponding to the object range to which the transaction reference information of the target object belongs, and further the probability that the target transaction has risk under the conditions of the transaction reference information is determined, so that the accuracy of determining the transaction risk can be improved.
The method of the embodiment of the application is described above, and the device of the embodiment of the application is described below.
Referring to fig. 6, fig. 6 is a schematic diagram of a composition structure of a transaction data processing apparatus according to an embodiment of the present application, where the transaction data processing apparatus may be deployed on a computer device; the transaction data processing device may be used to perform the corresponding steps in the transaction data processing method provided by the embodiment of the application. The transaction data processing device 60 includes:
A range determining unit 601, configured to obtain a transaction reference object; the transaction reference object has at least one object range for evaluating a transaction type of a transaction; the at least one object range has an association relationship with the transaction type of the transaction, wherein the transaction type of the transaction is a normal transaction type or an abnormal transaction type;
a data acquisition unit 602, configured to acquire transaction attribute information of a plurality of reference transactions; the transaction attribute information of any reference transaction comprises transaction type and transaction reference information of the any reference transaction, and the type of the transaction reference information is matched with the type of the transaction reference object;
a data calculating unit 603, configured to calculate a first transaction evaluation parameter associated with each object range based on a transaction type of each reference transaction and an object range hit by transaction reference information of each reference transaction in the at least one object range, respectively; a first transaction evaluation parameter associated with any of the object ranges for evaluating a probability that a transaction type of any of the transactions is an abnormal transaction type.
Optionally, any one of the at least one object range is a target object range; the data calculation unit 603 is specifically configured to:
Determining the transaction reference information of the plurality of reference transactions as a target reference transaction, wherein the transaction reference information hits the reference transaction of the target object range;
counting the target quantity of target reference transactions with the transaction type being the normal transaction type in at least one target reference transaction;
a ratio between the target number and the total number of the at least one target reference transaction is determined as a first transaction evaluation parameter associated with the target object range.
Optionally, the plurality of reference transactions each have transaction reference information; the range determining unit 601 is specifically configured to:
acquiring a plurality of candidate object ranges of the transaction reference object;
determining a first reference transaction which hits each candidate object range of the transaction reference object by targeting the transaction reference information which belongs to the plurality of reference transactions;
determining a reference transaction with the abnormal transaction type as a second reference transaction hitting each candidate object range in the first reference transaction hitting each candidate object range;
determining the ratio between the number of the second reference transactions hitting each candidate object range and the number of the first reference transactions as the association degree between each candidate object range and the transaction type of the transactions respectively;
And taking the candidate object range with the association degree with the transaction type of the transaction being greater than or equal to the association degree threshold value as the object range of the transaction reference object.
Optionally, the transaction reference object includes at least one of:
the method comprises the steps of initiating time of a transaction, transferring amount of the transaction resources, geographic position of a transaction initiating object of the transaction, object type of a resource issuing object of a business resource to be acquired by the transaction, or geographic position of the resource issuing object.
Optionally, the transaction reference objects are of a plurality, each transaction reference object having at least one object range; the transaction data processing device 60 further comprises a transaction prediction unit 604, the transaction prediction unit 604 being configured to:
acquiring a plurality of transaction reference information of a target transaction; the type of a transaction reference information of the target transaction is matched with the type of a transaction reference object;
acquiring an object range of each transaction reference information of the target transaction, which is hit in at least one object range of the matched transaction reference object, as a hit object range of each transaction reference information;
calculating a target transaction evaluation parameter of the target transaction based on the first transaction evaluation parameter associated with the hit object range of each transaction reference information;
Wherein the target transaction evaluation parameter is a probability that the transaction type of the target transaction is an abnormal transaction type.
Optionally, the transaction prediction unit 604 is specifically configured to:
calculating a first product between first transaction evaluation parameters associated with the hit range of each transaction reference information;
acquiring first priori information of the plurality of reference transactions, and calculating a second product between the first product and the numerical value of the first priori information; the first priori information is used for indicating the proportion of the reference transaction with the abnormal transaction type in the plurality of reference transactions;
the target transaction evaluation parameter is determined based on the second product.
Optionally, the transaction prediction unit 604 is specifically configured to:
counting the number of reference transactions with abnormal transaction types in the plurality of reference transactions;
the ratio between the number of reference transactions of the transaction type abnormal transaction type and the total number of the plurality of reference transactions is determined as the first priori information.
Optionally, the transaction prediction unit 604 is specifically configured to:
calculating a second transaction evaluation parameter associated with the hit object range of each transaction reference information based on the first transaction evaluation parameter associated with the hit object range of each transaction reference information; a second transaction evaluation parameter associated with any hit range for evaluating a probability that the transaction type of the target transaction is an abnormal transaction type;
Calculating a third product between the second transaction evaluation parameters associated with the hit range of each transaction reference information;
acquiring second priori information of the plurality of reference transactions, and calculating a fourth product between the third product and the value of the second priori information; the second priori information is used for indicating the proportion of the reference transaction with the normal transaction type in the plurality of reference transactions;
and acquiring the sum value of the second product and the fourth product, and determining the ratio between the second product and the sum value as the target transaction evaluation parameter.
Optionally, the transaction prediction unit 604 is specifically configured to:
counting the number of reference transactions with the transaction type being a normal transaction type in the plurality of reference transactions;
the ratio between the number of reference transactions of the transaction type normal to the total number of the plurality of reference transactions is determined as the second prior information.
Optionally, the target transaction is a transaction initiated by a transaction initiation object to a resource release object, where the transaction initiation object is configured to obtain a service resource held by the resource release object based on the initiated target transaction; the transaction data processing device 60 further comprises a transaction processing unit 605, the transaction processing unit 605 being adapted to:
If the value of the target transaction evaluation parameter is greater than or equal to the parameter threshold, stopping the execution of the target transaction, and prompting the transaction initiating object that the resource issuing object has abnormality.
It should be noted that, in the embodiment corresponding to fig. 6, the content not mentioned may be referred to the description of the method embodiment, and will not be repeated here.
In the embodiment of the application, the transaction reference object is obtained, and is provided with at least one object range for evaluating the transaction type of the transaction, wherein each object range has an association relationship with the transaction type of the transaction; by acquiring the transaction attribute information of the plurality of reference transactions, the first transaction evaluation parameter associated with each object range may be calculated based on the transaction type of each reference transaction and the object range in which the transaction reference information of each reference transaction hits in at least one object range, respectively. The first transaction evaluation parameter may be used to evaluate a probability that a transaction type of the transaction is an abnormal transaction type. According to the method, through calculating the first transaction evaluation parameters associated with the object ranges of the transaction reference object, the probability that the transaction type of the transaction is the abnormal transaction type under the condition of each object range can be determined, namely the probability that the transaction is the abnormal transaction type under the condition of each object range, and then various factors affecting the transaction type can be determined. For example, the method and the device for determining the transaction type can influence the transaction type by determining which object ranges of which transaction reference objects have a larger probability of being of an abnormal transaction type under the condition of which object ranges of which transaction reference objects and a smaller probability of being of an abnormal transaction type under the condition of which object ranges of which transaction reference objects, so that the transaction type of the transaction can be accurately evaluated through the first transaction evaluation parameters associated with each object range of the transaction reference objects.
Referring to fig. 7, fig. 7 is a schematic diagram of a composition structure of a computer device according to an embodiment of the present application. As shown in fig. 7, the above-mentioned computer device may include: a processor 701 and a memory 702. Optionally, the computer device may further include a network interface or a power module. Data may be exchanged between the processor 701 and the memory 702.
The processor 701 may be a central processing unit (Central Processing Unit, CPU) which may also be other general purpose processors, digital signal processors (Digital Signal Processor, DSP), application specific integrated circuits (Application Specific Integrated Circuit, ASIC), off-the-shelf programmable gate arrays (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or the like. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
The network interface may include input devices, such as a control panel, microphone, receiver, etc., and/or output devices, such as a display screen, transmitter, etc., which are not shown.
The memory 702 may include read only memory and random access memory and provides program instructions and data to the processor 701. A portion of the memory 702 may also include non-volatile random access memory. Wherein the processor 701, when calling the program instructions, is configured to execute:
acquiring a transaction reference object; the transaction reference object has at least one object range for evaluating a transaction type of a transaction; the at least one object range has an association relationship with the transaction type of the transaction, wherein the transaction type of the transaction is a normal transaction type or an abnormal transaction type;
acquiring transaction attribute information of a plurality of reference transactions; the transaction attribute information of any reference transaction comprises transaction type and transaction reference information of the any reference transaction, and the type of the transaction reference information is matched with the type of the transaction reference object;
calculating a first transaction evaluation parameter associated with each object range based on the transaction type of each reference transaction and the object range hit by the transaction reference information of each reference transaction in the at least one object range; a first transaction evaluation parameter associated with any of the object ranges for evaluating a probability that a transaction type of any of the transactions is an abnormal transaction type.
Optionally, any one of the at least one object range is a target object range; the processor 701 is specifically configured to:
determining the transaction reference information of the plurality of reference transactions as a target reference transaction, wherein the transaction reference information hits the reference transaction of the target object range;
counting the target quantity of target reference transactions with the transaction type being the normal transaction type in at least one target reference transaction;
a ratio between the target number and the total number of the at least one target reference transaction is determined as a first transaction evaluation parameter associated with the target object range.
Optionally, the plurality of reference transactions each have transaction reference information; the processor 701 is specifically configured to:
acquiring a plurality of candidate object ranges of the transaction reference object;
determining a first reference transaction which hits each candidate object range of the transaction reference object by targeting the transaction reference information which belongs to the plurality of reference transactions;
determining a reference transaction with the abnormal transaction type as a second reference transaction hitting each candidate object range in the first reference transaction hitting each candidate object range;
Determining the ratio between the number of the second reference transactions hitting each candidate object range and the number of the first reference transactions as the association degree between each candidate object range and the transaction type of the transactions respectively;
and taking the candidate object range with the association degree with the transaction type of the transaction being greater than or equal to the association degree threshold value as the object range of the transaction reference object.
Optionally, the transaction reference object includes at least one of:
the method comprises the steps of initiating time of a transaction, transferring amount of the transaction resources, geographic position of a transaction initiating object of the transaction, object type of a resource issuing object of a business resource to be acquired by the transaction, or geographic position of the resource issuing object.
Optionally, the transaction reference objects are of a plurality, each transaction reference object having at least one object range; the processor 701 is further configured to:
acquiring a plurality of transaction reference information of a target transaction; the type of a transaction reference information of the target transaction is matched with the type of a transaction reference object;
acquiring an object range of each transaction reference information of the target transaction, which is hit in at least one object range of the matched transaction reference object, as a hit object range of each transaction reference information;
Calculating a target transaction evaluation parameter of the target transaction based on the first transaction evaluation parameter associated with the hit object range of each transaction reference information;
wherein the target transaction evaluation parameter is a probability that the transaction type of the target transaction is an abnormal transaction type.
Optionally, the processor 701 is specifically configured to:
calculating a first product between first transaction evaluation parameters associated with the hit range of each transaction reference information;
acquiring first priori information of the plurality of reference transactions, and calculating a second product between the first product and the numerical value of the first priori information; the first priori information is used for indicating the proportion of the reference transaction with the abnormal transaction type in the plurality of reference transactions;
the target transaction evaluation parameter is determined based on the second product.
Optionally, the processor 701 is specifically configured to:
counting the number of reference transactions with abnormal transaction types in the plurality of reference transactions;
the ratio between the number of reference transactions of the transaction type abnormal transaction type and the total number of the plurality of reference transactions is determined as the first priori information.
Optionally, the processor 701 is specifically configured to:
Calculating a second transaction evaluation parameter associated with the hit object range of each transaction reference information based on the first transaction evaluation parameter associated with the hit object range of each transaction reference information; a second transaction evaluation parameter associated with any hit range for evaluating a probability that the transaction type of the target transaction is an abnormal transaction type;
calculating a third product between the second transaction evaluation parameters associated with the hit range of each transaction reference information;
acquiring second priori information of the plurality of reference transactions, and calculating a fourth product between the third product and the value of the second priori information; the second priori information is used for indicating the proportion of the reference transaction with the normal transaction type in the plurality of reference transactions;
and acquiring the sum value of the second product and the fourth product, and determining the ratio between the second product and the sum value as the target transaction evaluation parameter.
Optionally, the processor 701 is specifically configured to:
counting the number of reference transactions with the transaction type being a normal transaction type in the plurality of reference transactions;
the ratio between the number of reference transactions of the transaction type normal to the total number of the plurality of reference transactions is determined as the second prior information.
Optionally, the target transaction is a transaction initiated by a transaction initiation object to a resource release object, where the transaction initiation object is configured to obtain a service resource held by the resource release object based on the initiated target transaction; the processor 701 is further configured to:
if the value of the target transaction evaluation parameter is greater than or equal to the parameter threshold, stopping the execution of the target transaction, and prompting the transaction initiating object that the resource issuing object has abnormality.
In the embodiment of the application, the transaction reference object is obtained, and is provided with at least one object range for evaluating the transaction type of the transaction, wherein each object range has an association relationship with the transaction type of the transaction; by acquiring the transaction attribute information of the plurality of reference transactions, the first transaction evaluation parameter associated with each object range may be calculated based on the transaction type of each reference transaction and the object range in which the transaction reference information of each reference transaction hits in at least one object range, respectively. The first transaction evaluation parameter may be used to evaluate a probability that a transaction type of the transaction is an abnormal transaction type. According to the method, through calculating the first transaction evaluation parameters associated with the object ranges of the transaction reference object, the probability that the transaction type of the transaction is the abnormal transaction type under the condition of each object range can be determined, namely the probability that the transaction is the abnormal transaction type under the condition of each object range, and then various factors affecting the transaction type can be determined. For example, the method and the device for determining the transaction type can influence the transaction type by determining which object ranges of which transaction reference objects have a larger probability of being of an abnormal transaction type under the condition of which object ranges of which transaction reference objects and a smaller probability of being of an abnormal transaction type under the condition of which object ranges of which transaction reference objects, so that the transaction type of the transaction can be accurately evaluated through the first transaction evaluation parameters associated with each object range of the transaction reference objects.
Optionally, the program instructions may further implement other steps of the method in the above embodiment when executed by the processor, which is not described herein.
The embodiments of the present application also provide a computer readable storage medium storing a computer program comprising program instructions which, when executed by a computer, cause the computer to perform a method as in the previous embodiments, the computer being part of a computer device as mentioned above. As an example, the program instructions may be executed on one computer device or on multiple computer devices located at one site, or alternatively, on multiple computer devices distributed across multiple sites and interconnected by a communication network, which may constitute a blockchain network.
Embodiments of the present application also provide a computer program product or computer program comprising computer instructions which, when executed by a processor, implement some or all of the steps of the above-described method. For example, the computer instructions are stored in a computer readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device performs the steps performed in the embodiments of the methods described above.
Those skilled in the art will appreciate that all or part of the processes in the methods of the above embodiments may be implemented by a computer program for instructing relevant hardware, where the program may be stored in a computer readable storage medium, and where the program, when executed, may include processes of the embodiments of the methods as described above. The 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 illustrative of the present application and is not to be construed as limiting the scope of the application, which is defined by the appended claims.
Claims (14)
1. A transaction data processing method, the method comprising:
acquiring a transaction reference object; the transaction reference object has at least one object range for evaluating a transaction type of a transaction; the at least one object range has an association relationship with the transaction type of the transaction, wherein the transaction type of the transaction is a normal transaction type or an abnormal transaction type;
acquiring transaction attribute information of a plurality of reference transactions; the transaction attribute information of any reference transaction comprises transaction type and transaction reference information of the any reference transaction, wherein the type of the transaction reference information is matched with the type of the transaction reference object;
Calculating a first transaction evaluation parameter associated with each object range based on the transaction type of each reference transaction and the object range hit by the transaction reference information of each reference transaction in the at least one object range; a first transaction evaluation parameter associated with any of the object ranges for evaluating a probability that a transaction type of any of the transactions is an abnormal transaction type.
2. The method of claim 1, wherein any of the at least one object range is a target object range; the calculating a first transaction evaluation parameter associated with each object scope based on the transaction type of each reference transaction and the object scope hit by the transaction reference information of each reference transaction in the at least one object scope, including:
determining the transaction reference information of the plurality of reference transactions as a target reference transaction, wherein the transaction reference information hits the reference transaction of the target object range;
counting the target quantity of target reference transactions with the transaction type being the normal transaction type in at least one target reference transaction;
a ratio between the target number and the total number of the at least one target reference transaction is determined as a first transaction evaluation parameter associated with the target object range.
3. The method of claim 1, wherein each of the plurality of reference transactions has transaction reference information; a process for obtaining the at least one object range of the transaction reference object, comprising:
acquiring a plurality of candidate object ranges of the transaction reference object;
determining the reference transaction of each candidate object range of the transaction reference object as a first reference transaction of the hit of transaction reference information belonging to the plurality of reference transactions;
determining a reference transaction with the abnormal transaction type as a reference transaction which hits the range of each candidate object in the first reference transaction which hits the range of each candidate object;
determining the ratio between the number of the second reference transactions hitting each candidate object range and the number of the first reference transactions as the association degree between each candidate object range and the transaction type of the transactions respectively;
and taking the candidate object range with the association degree with the transaction type of the transaction being greater than or equal to the association degree threshold value as the object range of the transaction reference object.
4. The method of claim 1, wherein the transaction reference object comprises at least one of:
The method comprises the steps of initiating time of a transaction, transferring amount of the transaction resources, geographic position of a transaction initiating object of the transaction, object type of a resource issuing object of a business resource to be acquired by the transaction, or geographic position of the resource issuing object.
5. The method of claim 1, wherein the transaction reference objects are of a plurality, each transaction reference object having at least one object range; the method further comprises the steps of:
acquiring a plurality of transaction reference information of a target transaction; the type of a transaction reference information of the target transaction is matched with the type of a transaction reference object;
acquiring an object range of each transaction reference information of the target transaction, which is hit in at least one object range of the matched transaction reference object, as a hit object range of each transaction reference information;
calculating target transaction evaluation parameters of the target transaction based on the first transaction evaluation parameters associated with the hit object range of each transaction reference information;
wherein the target transaction evaluation parameter is a probability that the transaction type of the target transaction is an abnormal transaction type.
6. The method of claim 5, wherein calculating the target transaction evaluation parameters for the target transaction based on the first transaction evaluation parameters associated with the hit range for each transaction reference information comprises:
Calculating a first product between first transaction evaluation parameters associated with the hit object range of each transaction reference information;
acquiring first priori information of the plurality of reference transactions, and calculating a second product between the first product and the numerical value of the first priori information; the first priori information is used for indicating the proportion of the reference transaction with the abnormal transaction type in the plurality of reference transactions;
the target transaction evaluation parameter is determined based on the second product.
7. The method of claim 6, wherein the obtaining the first prior information for the plurality of reference transactions comprises:
counting the number of reference transactions with abnormal transaction types in the plurality of reference transactions;
determining, as the first priori information, a ratio between a number of reference transactions of a transaction type that is an abnormal transaction type and a total number of the plurality of reference transactions.
8. The method of claim 6, wherein the determining the target transaction evaluation parameter based on the second product comprises:
calculating a second transaction evaluation parameter associated with the hit object range of each transaction reference information based on the first transaction evaluation parameter associated with the hit object range of each transaction reference information; a second transaction evaluation parameter associated with any hit range for evaluating a probability that the transaction type of the target transaction is an abnormal transaction type;
Calculating a third product between second transaction evaluation parameters associated with the hit object range of each transaction reference information;
acquiring second priori information of the plurality of reference transactions, and calculating a fourth product between the third product and the value of the second priori information; the second priori information is used for indicating the proportion of the reference transaction with the normal transaction type in the plurality of reference transactions;
and acquiring the sum value of the second product and the fourth product, and determining the ratio between the second product and the sum value as the target transaction evaluation parameter.
9. The method of claim 8, wherein the obtaining second prior information for the plurality of reference transactions comprises:
counting the number of reference transactions with normal transaction types in the plurality of reference transactions;
determining a ratio between a number of reference transactions of a transaction type that is a normal transaction type and a total number of the plurality of reference transactions as the second priori information.
10. The method of claim 5, wherein the target transaction is a transaction initiated by a transaction initiation object to a resource issuance object, the transaction initiation object to obtain a business resource held by the resource issuance object based on the initiated target transaction; the method further comprises the steps of:
And if the value of the target transaction evaluation parameter is greater than or equal to a parameter threshold, stopping the execution of the target transaction, and prompting the transaction initiating object that the resource issuing object has an abnormality.
11. A transaction data processing device, the device comprising:
a range determining unit for acquiring a transaction reference object; the transaction reference object has at least one object range for evaluating a transaction type of a transaction; the at least one object range has an association relationship with the transaction type of the transaction, wherein the transaction type of the transaction is a normal transaction type or an abnormal transaction type;
a data acquisition unit for acquiring transaction attribute information of a plurality of reference transactions; the transaction attribute information of any reference transaction comprises transaction type and transaction reference information of the any reference transaction, wherein the type of the transaction reference information is matched with the type of the transaction reference object;
a data calculating unit, configured to calculate a first transaction evaluation parameter associated with each object range based on a transaction type of each reference transaction and an object range hit by transaction reference information of each reference transaction in the at least one object range, respectively; a first transaction evaluation parameter associated with any of the object ranges for evaluating a probability that a transaction type of any of the transactions is an abnormal transaction type.
12. A computer device comprising a processor, a memory, wherein the memory is for storing a computer program, the computer program comprising program instructions, the processor being configured to invoke the program instructions to perform the method of any of claims 1-10.
13. A computer readable storage medium, characterized in that the computer readable storage medium stores a computer program adapted to be loaded and executed by a processor to cause a computer device having a processor to perform the method of any of claims 1-10.
14. A computer program product, characterized in that the computer program product comprises computer instructions which, when executed by a processor, implement the method according to any one of claims 1-10.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310523679.1A CN116975160A (en) | 2023-05-10 | 2023-05-10 | Transaction data processing method, device, equipment, medium and product |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310523679.1A CN116975160A (en) | 2023-05-10 | 2023-05-10 | Transaction data processing method, device, equipment, medium and product |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116975160A true CN116975160A (en) | 2023-10-31 |
Family
ID=88477443
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310523679.1A Pending CN116975160A (en) | 2023-05-10 | 2023-05-10 | Transaction data processing method, device, equipment, medium and product |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116975160A (en) |
-
2023
- 2023-05-10 CN CN202310523679.1A patent/CN116975160A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110009174B (en) | Risk recognition model training method and device and server | |
| JP2020522832A (en) | System and method for issuing a loan to a consumer determined to be creditworthy | |
| CN111371767B (en) | Malicious account identification method, malicious account identification device, medium and electronic device | |
| US20140143329A1 (en) | Discovering signature of electronic social networks | |
| EP3482353A1 (en) | Machine learning and prediction using graph communities | |
| CN110147925B (en) | Risk decision method, device, equipment and system | |
| CN112700252B (en) | Information security detection method and device, electronic equipment and storage medium | |
| WO2021254027A1 (en) | Method and apparatus for identifying suspicious community, and storage medium and computer device | |
| KR20120040589A (en) | Optimum tender price prediction method and system | |
| CN112801670B (en) | Risk assessment method and device for payment operation | |
| WO2022142903A1 (en) | Identity recognition method and apparatus, electronic device, and related product | |
| CN112785157A (en) | Risk identification system updating method and device and risk identification method and device | |
| WO2023045691A1 (en) | Object recognition method and apparatus, and electronic device and storage medium | |
| CN107256231B (en) | Team member identification device, method and system | |
| CN112733995A (en) | Method for training neural network, behavior detection method and behavior detection device | |
| CN114202417A (en) | Abnormal transaction detection method, apparatus, device, medium, and program product | |
| CN109829593B (en) | Credit determining method and device for target object, storage medium and electronic device | |
| CN111245815B (en) | Data processing method and device, storage medium and electronic equipment | |
| CN113569162A (en) | Data processing method, device, equipment and storage medium | |
| CN101449284A (en) | Scoring quality of traffic to network sites using interrelated traffic parameters | |
| CN116975160A (en) | Transaction data processing method, device, equipment, medium and product | |
| CN113298641A (en) | Integrity degree cognition method and device | |
| CN112613796A (en) | Financial risk prediction method, and financial risk score card generation method and device | |
| CN113850669A (en) | User grouping method and device, computer equipment and computer readable storage medium | |
| CN113052509A (en) | Model evaluation method, model evaluation apparatus, electronic device, and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication |