Resource classification method and device, computer equipment and readable storage medium

Abstract

The invention relates to the field of artificial intelligence, and particularly discloses a resource classification method and device, computer equipment, a computer readable storage medium and a computer program product. The method comprises the steps of receiving a to-be-classified resource which carries a resource transfer-in message, searching whether a target resource classification strategy matched with the to-be-classified resource exists in preset resource classification strategies or not according to the resource transfer-in message, and when the target resource classification strategy is searched, classifying the to-be-classified resource according to the resource transfer-in message and the target resource classification strategy. And determining a resource registration book corresponding to the to-be-classified resource, identifying a service type of the to-be-classified resource according to the resource transfer-in message under the condition that the target resource classification strategy is not found, and determining the resource registration book corresponding to the to-be-classified resource based on the service type. By adopting the method, the resource classification efficiency can be improved.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to a resource classification method, apparatus, computer equipment, computer-readable storage medium, and computer program product. Background Art

[0002] With the development of information technology, resource transfer accounts for an increasing proportion in resource management business, and the ways of resource transfer are becoming more and more diversified. In large organizations and enterprises, in the process of resource transfer, it is usually necessary to classify resources to distinguish the type, source, purpose, etc. of the resources transferred in or out. The way of resource classification is crucial to the efficiency of resource management.

[0003] In traditional solutions, when transferring resources through internal resource transfer channels, Internet resource transfer channels, etc., the transferred resources are usually first temporarily stored in a main account and registered in the default sub-register, and then classified through internal adjustments and registered in different sub-registers. However, the above solution is suitable for simple resource classification. When faced with the current diversified resource transfer channels and a large number of resources to be transferred, the above resource classification solution still has the problem of low classification efficiency. Summary of the invention

[0004] Based on this, it is necessary to provide a resource classification method, apparatus, computer equipment, computer-readable storage medium and computer program product that can improve the efficiency of resource classification in response to the above technical problems.

[0005] In a first aspect, the present application provides a resource classification method, comprising:

[0006] receiving a resource to be classified, wherein the resource to be classified carries a resource transfer message;

[0007] According to the resource transfer message, searching for a target resource classification strategy matching the resource to be classified in the preset resource classification strategies;

[0008] When the target resource classification policy is found, determining a resource register corresponding to the resource to be classified according to the resource transfer message and the target resource classification policy;

[0009] In the case that the target resource classification strategy is not found, the service type of the resource to be classified is identified according to the resource transfer message, and the resource register corresponding to the resource to be classified is determined based on the service type.

[0010] In one embodiment, the resource register includes a register of resources to be cleared; and determining the resource register corresponding to the resource to be classified based on the business type includes:

[0011] In the case where the business type is the first business type, the resources to be classified are transferred into a register of resources to be cleared, and a resource transfer operation prompt message is pushed to the device logged in by the resource management account, wherein the resource transfer operation prompt message is used to instruct the resource management personnel to determine the resource register corresponding to the resources to be classified;

[0012] In the case where the service type is the second service type, taking the resource transfer message as input, calling the trained resource classification model, and obtaining a resource register corresponding to the resource to be classified;

[0013] The trained resource classification model is obtained by training based on historical resources to be classified, and the historical resources to be classified carry resource register tags.

[0014] In one embodiment, after taking the resource transfer message as input, calling the trained resource classification model, and obtaining the resource register corresponding to the resource to be classified, the method further includes:

[0015] Pushing a resource transfer operation prompt message to the device logged in by the resource management account, wherein the resource transfer operation prompt message is used to instruct the resource management personnel to determine the resource register corresponding to the resource to be classified;

[0016] receiving a resource register corresponding to the resource to be classified determined by the resource management personnel fed back by the device;

[0017] Comparing the resource register determined by the resource management personnel and the resource register determined by the resource classification model to determine whether they are consistent, and obtaining the classification accuracy of the trained resource classification model;

[0018] When the classification accuracy is less than a preset accuracy threshold, an alarm message is generated and pushed to the device.

[0019] In one embodiment, after comparing the resource register determined by the resource management personnel and the resource register determined by the resource classification model to see whether they are consistent, the method further includes:

[0020] In the case where it is determined that the resource register determined by the resource management personnel is inconsistent with the resource register determined by the resource classification model, determining the resource to be classified as a negative sample;

[0021] Based on the negative example samples, the resource classification model is updated.

[0022] In one of the embodiments, the resource transfer message includes at least one of a resource register tag, a resource transferee identifier, and resource transfer remark information;

[0023] Determining the resource register corresponding to the resource to be classified according to the resource transfer message and the target resource classification strategy includes:

[0024] In the case where the resource transfer-in message includes the resource register book tag, determining the resource register book corresponding to the resource register book tag as the resource register book corresponding to the resource to be classified;

[0025] In the case where the resource transfer-in message does not include the resource register tag, determining the resource register corresponding to the resource to be classified based on the target resource classification strategy, the resource transfer-in party identifier and / or resource transfer remark information;

[0026] In one of the embodiments, the identifying the service type of the resource to be classified according to the resource transfer message includes:

[0027] Extracting an account type field from the resource transfer-in message;

[0028] Based on the account type field, the business type of the resource to be classified is identified.

[0029] In one embodiment, the preset resource classification strategy is obtained based on the following method:

[0030] Receiving a resource mapping relationship input by a resource transfer request object, wherein the resource mapping relationship is used to represent a mapping relationship between the resource to be classified and a resource register;

[0031] Performing a validity check on the resource mapping relationship, and screening out the resource mapping relationship that passes the validity check;

[0032] Generate a resource classification strategy based on the filtered resource mapping relationship.

[0033] In a second aspect, the present application also provides a resource classification device, including:

[0034] A data acquisition module, used for receiving resources to be classified, wherein the resources to be classified carry a resource transfer message;

[0035] A policy matching module, used to search, according to the resource transfer message, whether there is a target resource classification policy matching the resource to be classified in the preset resource classification policies;

[0036] A first resource classification module is used to determine the resource register corresponding to the resource to be classified according to the resource transfer message and the target resource classification strategy when the target resource classification strategy is found;

[0037] The second resource classification module is used to identify the business type of the resource to be classified according to the resource transfer message when the target resource classification strategy is not found, and determine the resource registration book corresponding to the resource to be classified based on the business type.

[0038] In one embodiment, the resource register includes a register of resources to be cleared, and the second resource classification module includes a first classification unit and a second classification unit;

[0039] The first classification unit is used to transfer the resources to be classified into a register of resources to be cleared when the business type is the first business type, and push a resource transfer operation prompt message to the device logged in by the resource management account, wherein the resource transfer operation prompt message is used to instruct the resource management personnel to determine the resource register corresponding to the resources to be classified;

[0040] The second classification unit is used to, when the service type is the second service type, take the resource transfer message as input, call the trained resource classification model, and obtain a resource register corresponding to the resource to be classified;

[0041] The trained resource classification model is obtained by training based on historical resources to be classified, and the historical resources to be classified carry resource register tags.

[0042] In one embodiment, the second resource classification module further includes a message pushing unit, a data receiving unit, a comparison unit, and an alarm unit;

[0043] The message push unit is used to push a resource transfer operation prompt message to the device logged in by the resource management account, and the resource transfer operation prompt message is used to instruct the resource management personnel to determine the resource registration book corresponding to the resource to be classified;

[0044] The data receiving unit is used to receive the resource register corresponding to the resource to be classified determined by the resource management personnel fed back by the device;

[0045] The comparison unit is used to compare the resource register determined by the resource management personnel and the resource register determined by the resource classification model to see whether they are consistent, and obtain the classification accuracy of the trained resource classification model;

[0046] The alarm unit is used to generate and push alarm information to the device when the classification accuracy is less than a preset accuracy threshold.

[0047] In one embodiment, the second resource classification module further includes a sample determination unit and a model updating unit;

[0048] The sample determination unit is used to determine the resource to be classified as a negative sample when it is determined that the resource register determined by the resource management personnel is inconsistent with the resource register determined by the resource classification model;

[0049] The model updating unit is used to update the resource classification model based on the negative example samples.

[0050] In one embodiment, the resource transfer message includes at least one of a resource register tag, a resource transferee identifier, and resource transfer remark information;

[0051] The first resource classification module includes a label classification unit and an information classification unit;

[0052] The label classification unit is used to determine the resource register corresponding to the resource register tag as the resource register corresponding to the resource to be classified when the resource transfer message includes the resource register tag;

[0053] The information classification unit is used to determine the resource register corresponding to the resource to be classified based on the target resource classification strategy, the resource transferee identifier and / or resource transfer note information when the resource transfer-in message does not include the resource register tag.

[0054] In one embodiment, the second resource classification module further includes a field extraction unit and a business type identification unit;

[0055] The field extraction unit is used to extract the account type field from the resource transfer message;

[0056] The business type identification unit is used to identify the business type of the resource to be classified based on the account type field.

[0057] In one embodiment, the device further includes a mapping relationship acquisition module, a legality verification module, and a resource classification strategy determination module;

[0058] The mapping relationship acquisition module is used to receive the resource mapping relationship input by the resource transfer request object, and the resource mapping relationship is used to represent the mapping relationship between the resource to be classified and the resource register;

[0059] The legality verification module is used to perform legality verification on the resource mapping relationship and select the resource mapping relationship that passes the legality verification;

[0060] The resource classification strategy determination module is used to generate a resource classification strategy based on the screened resource mapping relationship.

[0061] In a third aspect, the present application further provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps in the above-mentioned resource classification method embodiments when executing the computer program.

[0062] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps in the above-mentioned resource classification method embodiments are implemented.

[0063] In a fifth aspect, the present application also provides a computer program product, including a computer program, which implements the steps in the above-mentioned resource classification method embodiments when executed by a processor.

[0064] The resource classification method, device, computer equipment, computer-readable storage medium and computer program product described above receive resources to be classified carrying resource transfer messages, and search for a target resource classification strategy that matches the resources to be classified in the preset resource classification strategy according to the resource transfer message. When the target resource classification strategy is found, the resource registration book corresponding to the resources to be classified is determined according to the resource transfer message and the target resource classification strategy. When the target resource classification strategy is not found, the business type of the resources to be classified is identified according to the resource transfer message, and the resource registration book corresponding to the resources to be classified is determined based on the business type. Different from the traditional scheme, the scheme described above sets different resource classification strategies for different types of resources to be classified. After receiving the resources to be classified, it first searches for a target resource classification strategy that matches the resources to be classified. If so, the resource registration book corresponding to the resources to be classified can be determined directly based on the target resource classification strategy and the resource transfer message, without first transferring to the default registration book for further classification, thereby simplifying the classification procedure and improving the classification efficiency. Furthermore, when faced with a large number of resources to be classified, this solution takes into account that there may be some resources to be classified that do not have corresponding target resource classification strategies. In this case, this solution can still adopt different classification methods for the resources to be classified based on their business types, thereby quickly classifying the resources to be classified and determining the resource registration book corresponding to the resources to be classified, thereby improving the classification efficiency of the resources to be classified and further improving the management efficiency of the transferred resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the drawings required for use in the embodiments of the present application or related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0066] Figure 1 An application environment diagram of a resource classification method in an embodiment;

[0067] Figure 2 A schematic diagram of a process flow of a resource classification method in one embodiment;

[0068] Figure 3 A schematic diagram of a multi-level account in one embodiment;

[0069] Figure 4 A schematic diagram of a flow chart of a resource classification method in another embodiment;

[0070] Figure 5 A schematic diagram of a process of determining a resource register corresponding to a to-be-classified resource of a second service type in an embodiment;

[0071] Figure 6 A schematic diagram of a flow chart of a resource classification method in another embodiment;

[0072] Figure 7 A schematic diagram of a process for generating a target resource classification strategy in an embodiment;

[0073] Figure 8 A schematic diagram of a process flow of a resource classification method in a detailed embodiment;

[0074] Fig. 9 is a structural block diagram of a resource classification device in an embodiment;

[0075] Fig.10 is a structural block diagram of a resource classification device in another embodiment;

[0076] Fig.11 is a structural block diagram of a resource classification device in yet another embodiment;

[0077] Fig.12 A structural block diagram of a resource classification device in another embodiment;

[0078] Fig.13 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0079] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application. It should be noted that in the embodiments of the present application, certain software, groups, models and other existing solutions in the industry may be mentioned, which should be considered as exemplary, and their purpose is only to illustrate the feasibility of the implementation of the technical solution of the present application, but it does not mean that the applicant has or will necessarily use the solution.

[0080] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations. The acquisition, storage, use, and processing of data in the technical solution of this application are in compliance with relevant provisions of national laws and regulations.

[0081] The resource classification method provided in the embodiment of the present application can be applied to Figure 1 In the application environment shown, the client 102 communicates with the server 104 via a network. The data storage system can store data that the server 104 needs to process. The data storage system can be integrated on the server 104, or placed on the cloud or other network servers.

[0082] Among them, the data storage system connected to the server 104 can store resource classification strategies in advance. The resources to be classified can be funds to be classified. In this case, the data storage system also stores a fund classification template in advance, and the fund classification template includes a general register, a management register, a register to be cleared, and a resource classification management template of several sub-registers. Specifically, when a user transfers a resource to be classified carrying a resource transfer message from an entity account through the user terminal 102, the server 104 searches the data storage system for a target resource allocation strategy that matches the resource to be classified in the preset resource classification strategy according to the resource transfer message. In the case of finding out the target resource classification strategy, the server 104 determines the resource register corresponding to the resource to be classified according to the resource transfer message and the target resource classification strategy. In the case of not finding out the resource target resource classification strategy, the server 104 identifies the business type of the resource to be classified according to the resource transfer message, and determines the resource register corresponding to the resource to be classified based on the resource business type.

[0083] The client 102 may be, but is not limited to, various personal computers, laptops, smart phones, and tablet computers. The server 104 may be an independent physical server, a server cluster or a distributed system composed of multiple physical servers, or a cloud server that provides cloud computing services.

[0084] In an exemplary embodiment, Figure 2 As shown, a resource classification method is provided, which is applied to Figure 1 Taking the server 104 in the example as an example, the following steps are included:

[0085] S100, receiving resources to be classified, the resources to be classified carrying a resource transfer message.

[0086] Among them, the resources to be classified refer to resources that need to be classified and managed according to specific rules or logic, such as financial resources, computing resources, network resources and other types of resources. For example, in a distributed computing scenario, different network resources may need to be allocated to different servers or different nodes. In this process, the classification and matching process of network resources is also involved. Specifically, when the resource to be classified is a network resource carrying a resource transfer message, the resource transfer message may record information such as the server identifier corresponding to the network resource, the source of the network resource, and the purpose of the network resource.

[0087] In the fund classification scenario, the resources to be classified can be financial resources, such as funds transferred to a certain unit's settlement account. In order to efficiently manage the transferred funds, the transferred funds need to be classified. Resource transfer messages refer to messages that carry detailed information about resources during the resource transfer process, including but not limited to the type, quantity, source, etc. of the resources, which can be used as the basis for resource classification.

[0088] For example, taking the resource to be classified as financial resources as an example, in the fund classification scenario, a multi-level account structure can be set, such as Figure 3As shown, it includes entity accounts (such as current settlement accounts), general registers, management registers, pending clearing registers, and several bottom registers. Each register is assigned an independent number, among which the pending clearing register is a register for registering funds that have not yet been classified. In addition, in a multi-level account structure, the general register and management register do not directly enter or withdraw funds, but are used to count the funds registered in the registers below them. When the funds in the registers of the lower level change, the funds in the registers of the upper level will be linked to change. Specifically, the account balance of the management register is equal to the sum of the account balances of all the sub-registers under it, and the balance of the current settlement account is equal to the registered balance of the general register, which is also equal to the sum of the registered balances of the pending clearing register and all the bottom registers. For example, when a yuan of funds are credited to the current settlement account, and the credited funds of a yuan are registered in a certain bottom register, the management register above the bottom register also records a yuan of funds. It should be noted that the bottom register can be added or deleted.

[0089] The process of classifying funds based on a multi-level account structure is equivalent to registering different funds transferred into a physical account in different underlying registers to achieve functions such as distinguishing the source and purpose of the transferred funds. When funds in a current settlement account change, the changed funds can be registered in the pending clearing register by default, or directly registered in the underlying register through a preset resource classification strategy.

[0090] S200 , searching, according to the resource transfer message, whether there is a target resource classification policy matching the resource to be classified in the preset resource classification policies.

[0091] The preset resource classification strategy may be a classification rule preset by a user, which includes a field that matches the resource transfer message.

[0092] Following the above steps, taking the fund classification scenario as an example, the resource to be classified is the transferred funds transferred into the entity account, wherein the resource transfer message carried by the transferred funds may include but is not limited to fields such as the source of funds and the purpose of transfer. The resource transfer message is parsed and the above fields are extracted, and then the extracted fields are matched with the preset classification resource strategy to determine the target resource classification strategy. Exemplarily, when the resource transfer message includes "Account xxx of Company A transfers xx yuan to the current settlement account", the resource transfer message includes "Account xxx of Company A", and its corresponding resource classification strategy can be "When the transferee is Account xxx of Company A, register the incoming funds in the resource register 0001".

[0093] S300, when the target resource classification policy is found, the resource register corresponding to the resource to be classified is determined according to the resource transfer message and the target resource classification policy.

[0094] Furthermore, when the target resource classification strategy is found, the resource register corresponding to the resource to be classified can be determined directly according to the resource transfer message and the target resource classification strategy. Continuing from the above content, when the resource transfer message includes "Account xxx of Company A transfers xx yuan to the current settlement account", its corresponding resource classification strategy is "When the transferee is Account xxx of Company A, register the incoming funds in Resource Register 0001", at this time, directly register the transferred funds in Resource Register 0001.

[0095] In other embodiments, if there is a transfer of funds into an entity account, the resource transfer message it carries may be "Zhang San's bank card transfers xx yuan into the entity account", and the target resource classification strategy that matches it is "When the transferee is Zhang San's bank card, the incoming funds will be registered in the resource registration book 0002". At this time, the transferred funds can be directly registered in the resource registration book 0002.

[0096] S400, when the target resource classification strategy is not found, identifying the business type of the resource to be classified according to the resource transfer message, and determining the resource register corresponding to the resource to be classified based on the business type.

[0097] Among them, the business type may include but is not limited to personal business, corporate business, etc. In the fund classification scenario, personal business may be an individual transferring personal funds into a physical account, and the personal funds are classified through a multi-level account structure and registered in a corresponding register. Corporate business may be a large enterprise transferring funds to be settled into a public current settlement account (physical account), and the funds to be settled are classified through a multi-level account structure and registered in a corresponding register.

[0098] It should be noted that since the preset resource classification strategy may be difficult to meet the current diverse resource transfer channels, if the target resource classification strategy is not found, the business type of the resource to be classified can be first identified based on the resource transfer message, and further based on the business type, the resource registration book corresponding to the resource to be classified can be determined.

[0099] For example, when the business type is personal business and the funds to be classified are personal funds, since the classification requirements for personal funds are not high, the priority of classification efficiency is higher than the classification accuracy, so a high-efficiency method can be selected to classify personal funds and determine the resource register, such as a deep learning model. When the business type is corporate business and the funds to be classified are the company's pending settlement funds, since the classification requirements for the company's pending settlement funds are high, the priority of classification accuracy is higher than classification efficiency, so a high-accuracy method can be selected to classify the pending settlement funds, such as through manual classification, or adding a resource classification strategy for the pending settlement funds.

[0100] The above resource classification method receives a resource to be classified carrying a resource transfer message, searches for a target resource classification strategy that matches the resource to be classified in the preset resource classification strategy based on the resource transfer message, and determines the resource registration book corresponding to the resource to be classified based on the resource transfer message and the target resource classification strategy when the target resource classification strategy is found. If the target resource classification strategy is not found, identifies the business type of the resource to be classified based on the resource transfer message, and determines the resource registration book corresponding to the resource to be classified based on the business type. Different from the traditional scheme, the above scheme can quickly and efficiently analyze whether there is a corresponding target resource classification strategy for the resource to be classified based on the resource transfer message by matching. If so, the resource registration book corresponding to the resource to be classified can be determined directly based on the target resource classification strategy and the resource transfer message, without first transferring to the default registration book for further classification, which saves the classification procedure and improves the classification efficiency. Furthermore, when faced with a large number of resources to be classified, this solution takes into account that there may be some resources to be classified that do not have corresponding target resource classification strategies. In this case, this solution can still quickly classify the resources to be classified based on their business types, and determine the resource registration book corresponding to the resources to be classified, thereby improving the classification efficiency of the resources to be classified and further improving the management efficiency of the transferred resources.

[0101] In an exemplary embodiment, the resource register includes a register of resources to be cleared, such as Figure 4 As shown, S400 includes the following steps:

[0102] S410: Identify the service type of the resource to be classified according to the resource transfer message.

[0103] S420: When the business type is the first business type, the resources to be classified are transferred to a register of resources to be cleared, and a resource transfer operation prompt message is pushed to the device logged in by the resource management account.

[0104] S430, when the service type is the second service type, taking the resource transfer message as input, calling the trained resource classification model, and obtaining a resource register corresponding to the resource to be classified.

[0105] Following the above embodiment, in the fund classification scenario, the first business type may be enterprise business, and the resources to be classified may be the enterprise's pending settlement funds. Enterprise business includes but is not limited to fund transfers between different enterprises and internal fund transfers between the same enterprise. The second business type may be personal business, and the resources to be classified may be personal funds. Personal business includes but is not limited to fund transfers from enterprises to individuals and fund transfers between individuals. Exemplarily, in the case where the business type is enterprise business, the entity account in the multi-level account structure may be a public current settlement account. At this time, since the enterprise business has high requirements for the accuracy of fund classification, manual classification can be used to classify funds. Specifically, the transferred funds to be settled can be registered in the register of resources to be cleared, and a resource transfer operation prompt message can be pushed to the background management end, that is, the device logged in by the resource management account, to remind the relevant resource management personnel to perform fund classification operations.

[0106] When the business type is personal business and the resources to be classified are personal funds, the entity account in the multi-level account structure can be a personal fund account. The individual user needs to classify the personal funds transferred into the personal fund account. At this time, since the accuracy requirement of fund classification is relatively low, in order to improve the efficiency of fund classification, the resource register corresponding to the resources to be classified can be obtained through the trained resource classification model. The training method of the resource classification model can be as follows:

[0107] Before training the model, it is necessary to collect a large amount of historical data, such as historical resources to be classified and resource register labels carried by historical resources to be classified, and then pre-process the historical data through data cleaning, feature extraction, format normalization, etc., and finally form the training data set of the model. The resource classification model can be a decision tree model, a convolutional neural network model, a recurrent neural network model, etc. Select an appropriate resource classification model according to actual needs. For example, select a convolutional neural network model to handle complex resource classification problems, train the selected model with the training data set, and continuously optimize the parameters of the model during the training process to obtain a trained resource classification model. Deploy the trained resource classification model to the relevant system or product, so that the resource classification model can receive and transfer messages according to the resources to be classified in real time, and output the resource register corresponding to the resources to be classified.

[0108] Furthermore, the trained resource classification model can be further evaluated and optimized. For example, a test set of the model can be constructed. If the classification accuracy of the resource classification model on the test set is 85%, the classification performance of the resource classification model can be improved by adjusting the learning rate parameter of the model or increasing the number of training rounds of the model until the classification accuracy reaches more than 90%. The resource classification model is then used in online resource classification.

[0109] It should be noted that in addition to selecting a resource classification method based on the identified business type, the user can also select a resource classification method by himself. For example, in a fund classification scenario, the funds transferred to the user's current account can be classified through a resource classification product that contains a multi-level account structure. During the fund classification process, the user can select the resource classification method of the resource classification product. Specifically, during the process of signing a contract between the user's current account and the resource classification product, a resource classification method selection interface is pushed to the user, and in response to the touch operation of the user end, a suitable resource classification method is selected for the user in a targeted manner. For example, when the business type is personal business, the resources to be classified are personal funds, and the entity accounts in the multi-level account structure are personal fund accounts, some users' personal fund accounts have few channels for fund entry and relatively few personal funds that need to be classified. Users can also choose to register personal funds in different resource registers through simple resource classification strategies and manual classification, thereby saving computing resources consumed in the processing process and improving resource classification efficiency; when the business type is corporate business, the resources to be classified can be the company's pending settlement funds, and the entity accounts in the multi-level account structure are public current settlement accounts, since there are many channels for fund entry in public current settlement accounts, the efficiency of manual classification is relatively low. Therefore, enterprises can also choose to classify funds through resource classification models, and considering that the higher the classification accuracy of the resource classification model, the higher the computing resources consumed and the processing time may be, the enterprise can also choose different resource classification models by itself. For example, if the enterprise needs to obtain a classification result with a higher accuracy, it should choose a resource classification model with a higher precision, and it will also need to pay a longer processing time accordingly. In addition, the resource classification method selected by the user can have a higher priority than the resource classification method selected by the resource classification system or resource classification product based on the business type, thereby making the resource classification method more in line with the user's business needs and improving user satisfaction.

[0110] In this embodiment, taking into account that the data characteristics and business processing requirements of different business types may be different, different resource classification methods are selected in a targeted manner, so that the classified resources can be classified more accurately; further, the user can also select an appropriate resource classification method according to the business processing requirements. For simple and common business types, the manual classification method is selected, and for complex business types, the user can select an appropriate resource classification model for resource classification, so as to reasonably utilize computing resources and improve the efficiency of resource classification.

[0111] In an exemplary embodiment, Figure 5 As shown, S430 includes:

[0112] S431, push the resource transfer operation prompt message to the device logged in by the resource management account.

[0113] S433, receiving the resource register corresponding to the to-be-classified resources determined by the resource management personnel fed back by the device.

[0114] S435, comparing the resource register determined by the resource management personnel and the resource register determined by the resource classification model to see if they are consistent.

[0115] S437, obtaining the classification accuracy of the trained resource classification model, and generating and pushing alarm information to the device when the classification accuracy is less than a preset accuracy threshold.

[0116] Following the above embodiment, if the resource classification model is selected according to the business type or the user, the resources to be classified are classified to obtain the resource register determined by the resource classification model, and a manual verification link can be added later. Exemplarily, a resource transfer prompt message is pushed to the device logged in by the resource management account to notify the resource management personnel to manually classify the resources to be classified to determine the classification accuracy of the resource classification model.

[0117] Exemplarily, it may be a regular or irregular manner to randomly extract resources to be classified through a resource classification model, and push a resource transfer operation prompt message to the device logged in by the resource management account to remind the resource management personnel to manually classify the randomly extracted resources to be classified and determine the resource registration book corresponding to the resources to be classified.

[0118] Furthermore, the resource register obtained by manual classification is compared with the resource register determined by the resource classification model to determine whether the two are consistent. Generally speaking, the resource register obtained by manual classification is more in line with the user's resource classification needs. Therefore, in the case of inconsistency between the two, the resource register determined by the resource classification model is determined to be an erroneous resource register, and the classification error probability or classification accuracy of the resource classification model is further calculated. When the classification accuracy is lower than the preset accuracy, for example, the classification accuracy is lower than 90%, an alarm message can be generated and pushed to the device logged in by the resource management account to remind the resource management personnel to maintain and update the resource classification model.

[0119] In this embodiment, the reliability of the resource classification model is evaluated by randomly extracting resource registers determined by the resource classification model and manually verifying it. When the classification accuracy of the resource classification model is lower than the preset accuracy threshold, the resource management personnel are promptly notified to maintain and update the model, which can effectively improve the accuracy of resource classification and further improve the accuracy and reliability of subsequent resource management.

[0120] In an exemplary embodiment, Figure 5 As shown, after S435, the method further includes:

[0121] S436, when the resource register determined by the resource management personnel and the resource register determined by the resource classification model are inconsistent, the resource to be classified is determined as a negative example sample, and the resource classification model is updated based on the negative example sample.

[0122] Following the above embodiment, when the resource register determined by the resource management personnel and the resource register determined by the resource classification model are inconsistent, it can be considered that the resource to be classified is classified incorrectly. At this time, it is necessary to mark the resource to be classified, add negative example labels, and obtain negative example samples. Further, based on the negative example samples, the resource classification model is updated. During the update process, the relevant parameters of the resource classification model can be adjusted, the number of learning rounds can be increased, etc., to improve the classification accuracy of the resource classification model.

[0123] In this embodiment, when the resource registration book determined by the resource management personnel and the resource registration book determined by the resource classification model are inconsistent, remedial measures are taken in a timely manner. By taking the misclassified resources to be classified as negative examples, the resource classification model is updated in a targeted manner, thereby improving the resource classification efficiency and accuracy of the resources to be classified.

[0124] In an exemplary embodiment, the resource transfer message includes at least one of a resource register tag, a resource transferee identifier, and resource transfer remark information, such as Figure 6 As shown, S300 includes:

[0125] S310, determining whether the resource transfer message includes a resource register tag.

[0126] S320: If yes, determine the resource register corresponding to the resource register tag as the resource register corresponding to the resource to be classified.

[0127] S330: If not, determine the resource register corresponding to the resource to be classified based on the target resource classification strategy, the resource transferee identifier and / or the resource transfer remark information.

[0128] Taking the application of the solution of this embodiment in the fund classification scenario as an example, the resource to be classified is the transferred funds transferred to the entity account, wherein the resource register tag can be used to specify the resource register specified by the transferred funds. For example, when the transferred funds are recorded, the resource transfer message carried includes "resource register 0001", which indicates that the resource register specified by the transferred funds is numbered 0001. At this time, the resource register numbered 0001 can be directly used as the resource register corresponding to the transferred funds.

[0129] The resource transferee identifier may include, but is not limited to, the resource transferee's account name, account number, and other identifiers. For example, when the resource transfer-in message includes the resource transferee's account name, and the account name is "Zhang San", the target resource classification strategy that matches it may be "When Zhang San's account transfers money to the entity account, the transferred funds will be registered in the 0002 resource register", so the transferred funds can be registered in the resource register numbered 0002.

[0130] In another exemplary embodiment, when the resource transfer message includes the account number of the resource transferee, and the account number is "310xxxx", the target resource classification strategy that matches it may be "When account number 310xxxx transfers money to an entity account, the transferred funds are registered in the 0003 resource register", so the transferred funds are registered in the resource register numbered 0003.

[0131] In addition, the resource transfer message may also include the above account name and account number to improve the accuracy of resource classification. For example, the resource transfer message includes "Company A" and "Account 310xxxx", and the matching target resource classification strategy may be "When Company A and Account 310xxxx transfer funds to the entity account, the transferred funds are registered in the 0004 resource register", so the transferred funds are registered in the resource register numbered 0004.

[0132] When the resource transfer message contains resource transfer remark information, for example, the resource transfer remark information of the funds to be classified is "project payment", the corresponding target resource classification strategy can be "When the remark information of the funds transferred to the entity account is project payment, the transferred funds are registered in the 0005 resource registration book", so the transferred funds are registered in the resource registration book labeled 0005.

[0133] In this embodiment, for common resource classification services, resource classification is performed by setting resource classification strategies. This is because these services usually have fixed classification rules and clear classification paths. By reading the resource registration book label, resource transferee identifier, and resource transfer note information in the resource transfer message, resource classification processing can be performed directly based on the corresponding resource classification strategy. This method is fast and has low overhead, and can complete resource classification quickly and efficiently, thereby improving resource management efficiency.

[0134] In an exemplary embodiment, S410 includes: extracting an account type field from the resource transfer message, and identifying the business type of the resource to be classified based on the account type field.

[0135] Among them, taking the solution in this embodiment as an example in the fund classification scenario, the resource to be classified is the transferred funds transferred to the entity account, and the account field types include but are not limited to the transaction ID, fund amount, currency type, fund transferor account, fund recipient account, transaction type, etc. Continuing with the above embodiment, the business type can be personal business and corporate business. Personal business includes but is not limited to company-to-person transfer and inter-personal transfer. Corporate business includes but is not limited to transfer between different companies and internal transfer within a company.

[0136] The account type field is extracted from the resource transfer message, and the business type of the resource to be classified is identified according to the account type field. Specifically, if the extracted account field type includes the fund transferor account and the fund receiver account, and the fund transferor account is a corporate account and the fund receiver account is an individual account, then the business type corresponding to the funds to be classified this time is company-to-individual transfer. If the fund transferor account and the fund receiver account are different corporate accounts, then the corresponding business type is transfer between different companies.

[0137] In this embodiment, by extracting the account type field from the resource transfer message, the business type of this resource transfer can be determined based on the account type field, thereby providing an accurate basis for subsequent selection of an appropriate resource classification method, thereby improving the efficiency and accuracy of resource classification.

[0138] In an exemplary embodiment, Figure 7 As shown, S200 includes:

[0139] S210, receiving a resource mapping relationship input by a resource transfer request object.

[0140] S220: Perform a validity check on the resource mapping relationship, and select resource mapping relationships that pass the validity check.

[0141] S230: Generate a resource classification strategy based on the screened resource mapping relationship.

[0142] S240: Search, according to the resource transfer message, whether there is a target resource classification policy matching the resource to be classified in the preset resource classification policies.

[0143] The resource mapping relationship refers to the mapping relationship between the resource to be classified and the resource register, which can be preset by the user. Taking the application of the solution in this embodiment in the fund classification scenario as an example, the resource to be classified is the transferred funds transferred to the entity account. When the user's entity account signs a contract with the fund classification product, the user can operate through the employee system interface of the bank counter or the relevant web interface to realize the setting of the resource mapping relationship.

[0144] Specifically, the resource mapping relationship may include payment types and payment conditions. On the relevant web page interface, the user can first define the payment type, which includes but is not limited to "payment by transfer account number", "payment by transfer account name", "payment by transfer account + account name", "payment by transfer note", etc. After defining the payment type, the user can also define the payment conditions on the relevant interface, which includes but is not limited to "when the transfer account number is 310xxxx, the transferred funds are registered in resource register A", "when the transfer account name is Zhang San, the transferred funds are registered in resource register B", "when Li Si transfers funds through account 310xxxx, the transferred funds are registered in resource register C", "when the note of the transferred funds is salary, the transferred funds are registered in resource register D", etc.

[0145] It should be noted that for the same resource register, multiple payment types and corresponding payment conditions can be set, but considering that each additional payment type will increase the system overhead when executing the corresponding payment conditions, in the best embodiment, the same resource register only has one payment type and the payment conditions corresponding to the payment type. In addition, the user can set, modify, delete, query, add, and other operations on the payment type and payment conditions through touch operations on the relevant web page interface. For example, the above operations can be performed by entering the payment conditions, resource register number, and the account of the resource classification product that has been signed, and the operation will take effect after submission.

[0146] Considering that there are usually multiple resource registers, in order to save the number of user operations, when inputting resource mapping relationships, users do not need to repeat the input and submission operations for each resource register, but can use the resource mapping relationship input template pushed by the background, such as a table file, to batch input the resource mapping relationships of each resource register and submit them to the server. In addition, if the user needs to modify the input resource mapping relationships, etc., they can also perform batch modification operations through the resource mapping relationship modification template.

[0147] Furthermore, when the server receives the resource mapping relationship submitted by the user, it will first perform a validity check on the resource mapping relationship, for example, verifying whether there is a resource to be classified in the resource mapping relationship submitted by the user, which may be registered in multiple different resource registers at the same time. Then, the server will split the resource mapping relationship that has passed the validity check from the resource mapping relationship input template one by one to allow the server to call it separately, generate a resource classification strategy, and return the resource classification strategy to the relevant web page interface for users to download and view. In addition, the resource mapping relationship that has not passed the validity check will also be fed back to the user, reminding the user to delete or modify it.

[0148] In this embodiment, by receiving the resource mapping relationship input by the user and verifying the legality of the resource mapping relationship, the possibility of errors in the subsequent classification process is reduced, the accuracy of resource classification and resource management efficiency are improved, and for users, the intelligence and automation of resource classification are greatly improved; for banks, there is no need for additional modifications to the existing resource classification process, reducing development costs.

[0149] In order to make a clearer description of the resource classification method provided by the present application, the following is a specific embodiment and the attached Figure 8 To illustrate, this specific embodiment includes the following steps:

[0150] S100, receiving resources to be classified, the resources to be classified carrying a resource transfer message.

[0151] S200 , searching, according to the resource transfer message, whether there is a target resource classification policy matching the resource to be classified in the preset resource classification policies.

[0152] S310: When the target resource classification policy is found, it is determined whether the resource transfer message includes a resource register tag.

[0153] S320: If yes, determine the resource register corresponding to the resource register tag as the resource register corresponding to the resource to be classified.

[0154] S330: If not, determine the resource register corresponding to the resource to be classified based on the target resource classification strategy, the resource transferee identifier and / or the resource transfer remark information.

[0155] S410: When the target resource classification policy is not found, the service type of the resource to be classified is identified according to the resource transfer message.

[0156] S420: When the business type is the first business type, the resources to be classified are transferred to a register of resources to be cleared, and a resource transfer operation prompt message is pushed to the device logged in by the resource management account.

[0157] S430, when the service type is the second service type, taking the resource transfer message as input, calling the trained resource classification model, and obtaining a resource register corresponding to the resource to be classified.

[0158] Specifically, the specific process of searching whether there is a target resource classification policy that matches the resources to be classified in the preset resource classification policy can be: receiving the resource mapping relationship input by the resource transfer request object, and performing a legitimacy check on the resource mapping relationship, screening out the resource mapping relationships that pass the legitimacy check, and then generating a resource classification policy based on the screened resource mapping relationships, and finally searching whether there is a target resource classification policy that matches the resources to be classified in the preset resource classification policy according to the resource transfer message.

[0159] Exemplarily, when the business type is the second business type, the resource transfer message is used as input to call the trained resource classification model, and after obtaining the resource registration book corresponding to the resource to be classified, a resource transfer operation prompt message is pushed to the device logged in by the resource management account, and the resource registration book corresponding to the resource to be classified determined by the resource management personnel fed back by the receiving device is received. The resource registration book determined by the resource management personnel and the resource registration book determined by the resource classification model are further compared to determine whether they are consistent, and the classification accuracy of the trained resource classification model is obtained. When the classification accuracy is less than the preset accuracy threshold, an alarm message is generated and pushed to the device, and when the two are inconsistent, the resource to be classified is determined as a negative example sample, and based on the negative example sample, the resource classification model is updated to obtain an updated resource classification model.

[0160] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0161] Based on the same inventive concept, the embodiment of the present application also provides a resource classification device for implementing the resource classification method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more resource classification device embodiments provided below can refer to the limitations on the resource classification method above, and will not be repeated here.

[0162] In an exemplary embodiment, Fig. 9 As shown, a resource classification device 900 is provided, comprising: a data acquisition module 910, a policy matching module 920, a first resource classification module 930 and a second resource classification module 940, wherein:

[0163] The data acquisition module 910 is used to receive resources to be classified, and the resources to be classified carry resource transfer messages.

[0164] The policy matching module 920 is used to search, according to the resource transfer message, whether there is a target resource classification policy matching the resource to be classified in the preset resource classification policies.

[0165] The first resource classification module 930 is used to determine the resource register corresponding to the resource to be classified according to the resource transfer message and the target resource classification policy when the target resource classification policy is found.

[0166] The second resource classification module 940 is used to identify the business type of the resource to be classified according to the resource transfer message when the target resource classification strategy is not found, and determine the resource registration book corresponding to the resource to be classified based on the business type.

[0167] In one embodiment, the resource register includes a register of resources to be cleared, such as Fig.10 As shown, the second resource classification module 940 includes a first classification unit 941 and a second classification unit 942 .

[0168] The first classification unit 941 is used to transfer the resources to be classified into a register of resources to be cleared when the business type is the first business type, and push a resource transfer operation prompt message to the device logged in by the resource management account. The resource transfer operation prompt message is used to instruct the resource management personnel to determine the resource register corresponding to the resources to be classified.

[0169] The second classification unit 942 is used to, when the service type is the second service type, take the resource transfer message as input, call the trained resource classification model, and obtain the resource register corresponding to the resource to be classified.

[0170] Among them, the trained resource classification model is trained based on historical resources to be classified, and the historical resources to be classified carry resource registration book labels.

[0171] In one embodiment, Fig.10 As shown, the second resource classification module 940 also includes a message pushing unit 943 , a data receiving unit 944 , a comparing unit 945 and an alarm unit 946 .

[0172] The message push unit 943 is used to push a resource transfer operation prompt message to the device logged in by the resource management account. The resource transfer operation prompt message is used to instruct the resource management personnel to determine the resource registration book corresponding to the resource to be classified.

[0173] The data receiving unit 944 is used to receive the resource register corresponding to the to-be-classified resource determined by the resource management personnel fed back by the device.

[0174] The comparison unit 945 is used to compare the resource register determined by the resource management personnel and the resource register determined by the resource classification model to see if they are consistent, and obtain the classification accuracy of the trained resource classification model.

[0175] The alarm unit 946 is used to generate and push alarm information to the device when the classification accuracy is less than a preset accuracy threshold.

[0176] In one embodiment, Fig.10 As shown, the second resource classification module 940 also includes a sample determination unit 947 and a model updating unit 948 .

[0177] The sample determination unit 947 is used to determine the resource to be classified as a negative sample when it is determined that the resource register determined by the resource management personnel is inconsistent with the resource register determined by the resource classification model.

[0178] The model updating unit 948 is used to update the resource classification model based on the negative example samples.

[0179] In one embodiment, the resource transfer message includes at least one of a resource register tag, a resource transferee identifier, and resource transfer remark information, such as Fig.11 As shown, the first resource classification module 930 includes a tag classification unit 931 and an information classification unit 932 .

[0180] The label classification unit 931 is used to determine the resource register corresponding to the resource register tag as the resource register corresponding to the resource to be classified when the resource transfer message includes the resource register tag.

[0181] The information classification unit 932 is used to determine the resource register corresponding to the resource to be classified based on the target resource classification strategy, the resource transferee identifier and / or the resource transfer remark information when the resource transfer-in message does not include a resource register tag.

[0182] In one embodiment, the second resource classification module further includes a field extraction unit and a business type identification unit. The field extraction unit is used to extract the account type field from the resource transfer message, and the business type identification unit is used to identify the business type of the resource to be classified based on the account type field.

[0183] In one embodiment, Fig.12 As shown, the resource classification device 900 further includes a mapping relationship acquisition module 950 , a legality verification module 960 and a resource classification strategy determination module 970 .

[0184] The mapping relationship acquisition module 950 is used to receive the resource mapping relationship input by the resource transfer request object, and the resource mapping relationship is used to represent the mapping relationship between the resource to be classified and the resource register.

[0185] The legality verification module 960 is used to perform legality verification on the resource mapping relationship and screen out the resource mapping relationship that passes the legality verification.

[0186] The resource classification strategy determination module 970 is used to generate a resource classification strategy based on the screened resource mapping relationship.

[0187] Each module in the resource classification device can be implemented in whole or in part by software, hardware, or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each module.

[0188] In an exemplary embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. Fig.13As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data such as resource classification strategies. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a resource classification method is implemented.

[0189] Those skilled in the art will understand that Fig.13 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0190] In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.

[0191] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0192] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0193] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., but are not limited to this.

[0194] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0195] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A resource classification method, characterized in that: The method comprises: receiving a resource to be classified, wherein the resource to be classified carries a resource transfer message; According to the resource transfer message, searching for a target resource classification strategy matching the resource to be classified in the preset resource classification strategies; When the target resource classification policy is found, determining a resource register corresponding to the resource to be classified according to the resource transfer message and the target resource classification policy; In the case that the target resource classification strategy is not found, the service type of the resource to be classified is identified according to the resource transfer message, and the resource register corresponding to the resource to be classified is determined based on the service type.

2. The method according to claim 1, characterized in that The resource register includes a register of resources to be cleared; The determining, based on the business type, a resource register corresponding to the resource to be classified includes: In the case where the business type is the first business type, the resources to be classified are transferred into a register of resources to be cleared, and a resource transfer operation prompt message is pushed to the device logged in by the resource management account, wherein the resource transfer operation prompt message is used to instruct the resource management personnel to determine the resource register corresponding to the resources to be classified; In the case where the service type is the second service type, taking the resource transfer message as input, calling the trained resource classification model, and obtaining a resource register corresponding to the resource to be classified; The trained resource classification model is obtained by training based on historical resources to be classified, and the historical resources to be classified carry resource register tags.

3. The method according to claim 2, characterized in that After taking the resource transfer message as input, calling the trained resource classification model, and obtaining the resource register corresponding to the resource to be classified, the method further includes: Pushing a resource transfer operation prompt message to the device logged in by the resource management account, wherein the resource transfer operation prompt message is used to instruct the resource management personnel to determine the resource register corresponding to the resource to be classified; receiving a resource register corresponding to the resource to be classified determined by the resource management personnel fed back by the device; Comparing the resource register determined by the resource management personnel and the resource register determined by the resource classification model to determine whether they are consistent, and obtaining the classification accuracy of the trained resource classification model; When the classification accuracy is less than a preset accuracy threshold, an alarm message is generated and pushed to the device.

4. The method according to claim 3, characterized in that After comparing the resource register determined by the resource management personnel and the resource register determined by the resource classification model to see whether they are consistent, the method further includes: In the case where it is determined that the resource register determined by the resource management personnel is inconsistent with the resource register determined by the resource classification model, determining the resource to be classified as a negative sample; Based on the negative example samples, the resource classification model is updated.

5. The method according to any one of claims 1 to 4, characterized in that: The resource transfer message includes at least one of a resource register tag, a resource transferee identifier, and resource transfer remark information; Determining the resource register corresponding to the resource to be classified according to the resource transfer message and the target resource classification strategy includes: In the case where the resource transfer-in message includes the resource register book tag, determining the resource register book corresponding to the resource register book tag as the resource register book corresponding to the resource to be classified; In the case that the resource transfer-in message does not include the resource register tag, the resource register corresponding to the resource to be classified is determined based on the target resource classification strategy, the resource transfer-in party identifier and / or resource transfer remark information.

6. The method according to any one of claims 1 to 4, characterized in that: The identifying the service type of the resource to be classified according to the resource transfer message includes: Extracting an account type field from the resource transfer-in message; Based on the account type field, the business type of the resource to be classified is identified.

7. The method according to any one of claims 1 to 4, characterized in that: The preset resource classification strategy is obtained based on the following method: Receiving a resource mapping relationship input by a resource transfer request object, wherein the resource mapping relationship is used to represent a mapping relationship between the resource to be classified and a resource register; Performing a validity check on the resource mapping relationship, and screening out the resource mapping relationship that passes the validity check; Generate a resource classification strategy based on the filtered resource mapping relationship.

8. A resource classification device, characterized in that: The device comprises: A data acquisition module, used for receiving resources to be classified, wherein the resources to be classified carry resource transfer messages; A policy matching module is used to search, according to the resource transfer message, whether there is a target resource classification policy that matches the resource to be classified in the preset resource classification policy; A first resource classification module is used to determine the resource register corresponding to the resource to be classified according to the resource transfer message and the target resource classification strategy when the target resource classification strategy is found; The second resource classification module is used to identify the business type of the resource to be classified according to the resource transfer message when the target resource classification strategy is not found, and determine the resource registration book corresponding to the resource to be classified based on the business type.

9. The device according to claim 8, characterized in that The resource register includes a register of resources to be cleared, and the second resource classification module includes a first classification unit and a second classification unit; The first classification unit is used to transfer the resources to be classified into a register of resources to be cleared when the business type is the first business type, and push a resource transfer operation prompt message to the device logged in by the resource management account, wherein the resource transfer operation prompt message is used to instruct the resource management personnel to determine the resource register corresponding to the resources to be classified; The second classification unit is used to, when the service type is the second service type, take the resource transfer message as input, call the trained resource classification model, and obtain a resource register corresponding to the resource to be classified; The trained resource classification model is obtained by training based on historical resources to be classified, and the historical resources to be classified carry resource register tags.

10. The device according to claim 9, characterized in that The second resource classification module also includes a message pushing unit, a data receiving unit, a comparison unit and an alarm unit; The message push unit is used to push a resource transfer operation prompt message to the device logged in by the resource management account, and the resource transfer operation prompt message is used to instruct the resource management personnel to determine the resource registration book corresponding to the resource to be classified; The data receiving unit is used to receive the resource register corresponding to the resource to be classified determined by the resource management personnel fed back by the device; The comparison unit is used to compare the resource register determined by the resource management personnel and the resource register determined by the resource classification model to see whether they are consistent, and obtain the classification accuracy of the trained resource classification model; The alarm unit is used to generate and push alarm information to the device when the classification accuracy is less than a preset accuracy threshold.

11. The device according to claim 10, characterized in that The second resource classification module also includes a sample determination unit and a model updating unit; The sample determination unit is used to determine the resource to be classified as a negative sample when it is determined that the resource register determined by the resource management personnel is inconsistent with the resource register determined by the resource classification model; The model updating unit is used to update the resource classification model based on the negative example samples.

12. The device according to any one of claims 8 to 11, characterized in that The resource transfer message includes at least one of a resource register tag, a resource transferee identifier, and resource transfer remark information, and the first resource classification module includes a tag classification unit and an information classification unit; The label classification unit is used to determine the resource register corresponding to the resource register tag as the resource register corresponding to the resource to be classified when the resource transfer message includes the resource register tag; The information classification unit is used to determine the resource register corresponding to the resource to be classified based on the target resource classification strategy, the resource transferee identifier and / or resource transfer note information when the resource transfer-in message does not include the resource register tag.

13. The device according to any one of claims 8 to 11, characterized in that The second resource classification module also includes a field extraction unit and a business type identification unit; The field extraction unit is used to extract the account type field from the resource transfer message; The business type identification unit is used to identify the business type of the resource to be classified based on the account type field.

14. The device according to any one of claims 8 to 11, characterized in that The device also includes a mapping relationship acquisition module, a legality verification module and a resource classification strategy determination module; The mapping relationship acquisition module is used to receive the resource mapping relationship input by the resource transfer request object, and the resource mapping relationship is used to represent the mapping relationship between the resource to be classified and the resource register; The legality verification module is used to perform legality verification on the resource mapping relationship and select the resource mapping relationship that passes the legality verification; The resource classification strategy determination module is used to generate a resource classification strategy based on the screened resource mapping relationship.

15. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

16. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

17. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images