CN115860749A

CN115860749A - Data processing method, device and equipment

Info

Publication number: CN115860749A
Application number: CN202310111090.0A
Authority: CN
Inventors: 李怀松; 宋博文; 张天翼; 成鹏; 秦思嘉
Original assignee: Alipay Hangzhou Information Technology Co Ltd
Current assignee: Alipay Hangzhou Information Technology Co Ltd
Priority date: 2023-02-09
Filing date: 2023-02-09
Publication date: 2023-03-28
Anticipated expiration: 2043-02-09
Also published as: CN115860749B

Abstract

The embodiment of the specification discloses a data processing method, a device and equipment, wherein the method comprises the following steps: receiving an account restriction request for a target account; acquiring limitation removing data used for removing a limitation target account, determining whether the target account is in an account white list or an account black list based on the limitation removing data, if so, refusing to perform limitation removing processing on the target account, and if so, acquiring and executing a first limitation removing rule corresponding to the account white list so as to perform limitation removing processing on the target account; if the target account cannot be determined to be in the account white list or the account black list based on the limitation solving data, different types of data contained in the limitation solving data are respectively processed through preset data processing rules to obtain corresponding limitation solving auxiliary information, and the target account is subjected to limitation solving based on the limitation solving auxiliary information to improve the efficiency and accuracy of account limitation solving.

Description

Data processing method, device and equipment

Technical Field

The present disclosure relates to the field of computer technologies, and in particular, to a data processing method, apparatus, and device.

Background

In the field of risk prevention and control, a situation that misjudgment occurs inevitably to limit the use of one or more rights of an account causes great disturbance to user experience, so that when a user complains about the misjudgment or requests to remove the rights, the user's appeal needs to be processed at the first time, whether the account of the user has risks is reexamined, and whether the rights of the account of the user need to be removed is further determined, but in reexamination, different levels of trial requirements (such as high trial speed, accurate evaluation result, compliance and interpretability) often exist. Generally, a manual trial mode can be adopted, but the cost of the manual review is very high in the mode, the efficiency is very low, a large number of tasks cannot be processed simultaneously, although certain efficiency can be improved through preset rules, the accuracy rate is difficult to control, serious misjudgment events can easily occur, and therefore the trial requirements of different levels can be well met, the limit evaluation efficiency can be improved, and an account limit solving scheme for controlling the accuracy rate can be well provided.

Disclosure of Invention

The embodiment of the specification aims to provide an account limitation solving scheme which can well meet the examination requirements of different levels to improve the limitation solving review efficiency and well control the accuracy of the limitation solving review scheme.

In order to implement the above technical solution, the embodiments of the present specification are implemented as follows:

the data processing method provided by the embodiment of the specification comprises the following steps: an account restriction request for a target account is received. Obtaining limitation data used for limiting the target account, determining whether the target account is in an account white list or in an account blacklist based on the limitation data, if the target account is determined to be in the account blacklist, refusing to perform limitation removal on the target account, and if the target account is determined to be in the account whitelist, obtaining and executing a first limitation removal rule corresponding to the account whitelist so as to perform limitation removal on the target account. If the target account cannot be determined to be in the account white list or the account black list based on the limitation removal data, processing different types of data contained in the limitation removal data respectively through preset data processing rules to obtain corresponding limitation removal auxiliary information, wherein the limitation removal auxiliary information is used for triggering a limitation removal manager to perform limitation removal processing on the target account based on the limitation removal auxiliary information.

The data processing system provided by the embodiment of the present specification includes an atomic capability subsystem, an algorithm subsystem and an application subsystem, wherein: the atomic capability subsystem is configured to provide corresponding algorithmic support for the algorithmic system subsystem and the application subsystem. The application subsystem is configured to receive an account limitation solving request for a target account, acquire limitation solving data for limitation solving of the target account, and call the algorithm system subsystem to carry out limitation solving processing based on the limitation solving data. The algorithm system subsystem is configured to call the algorithm in the atomic capability subsystem to execute the following processing: determining whether the target account is in an account white list or an account blacklist or not based on the limitation removing data, if the target account is determined to be in the account blacklist, refusing to carry out limitation removing processing on the target account, and if the target account is determined to be in the account white list, acquiring and executing a first limitation removing rule corresponding to the account white list so as to carry out limitation removing processing on the target account; if the target account cannot be determined to be in the account white list or the account black list based on the limitation data, processing different types of data contained in the limitation data through preset data processing rules respectively to obtain corresponding limitation auxiliary information, wherein the limitation auxiliary information is used for triggering a limitation removal management party to carry out limitation removal processing on the target account based on the limitation removal auxiliary information.

An embodiment of this specification provides a data processing apparatus, the apparatus includes: and the limitation removal request module receives an account limitation removal request aiming at the target account. The first limitation removing module is used for obtaining limitation removing data used for removing the limitation of the target account, determining whether the target account is in an account white list or an account blacklist based on the limitation removing data, refusing to remove the limitation of the target account if the target account is determined to be in the account blacklist, and obtaining and executing a first limitation removing rule corresponding to the account white list to remove the limitation of the target account if the target account is determined to be in the account white list. And the second limitation removing module is used for processing different types of data contained in the limitation removing data respectively through preset data processing rules to obtain corresponding limitation removing auxiliary information if the target account cannot be determined to be in an account white list or an account black list based on the limitation removing data, wherein the limitation removing auxiliary information is used for triggering a limitation removing manager to remove the limitation of the target account based on the limitation removing auxiliary information.

An embodiment of the present specification provides a data processing apparatus, including: a processor; and a memory arranged to store computer executable instructions that, when executed, cause the processor to: an account restriction request for a target account is received. Obtaining limitation removing data used for removing the limitation of the target account, determining whether the target account is in an account white list or an account black list based on the limitation removing data, if the target account is determined to be in the account black list, refusing to perform limitation removing processing on the target account, and if the target account is determined to be in the account white list, obtaining and executing a first limitation removing rule corresponding to the account white list so as to perform limitation removing processing on the target account. If the target account cannot be determined to be in the account white list or the account black list based on the limitation removal data, processing different types of data contained in the limitation removal data respectively through preset data processing rules to obtain corresponding limitation removal auxiliary information, wherein the limitation removal auxiliary information is used for triggering a limitation removal manager to perform limitation removal processing on the target account based on the limitation removal auxiliary information.

Embodiments of the present specification also provide a storage medium for storing computer-executable instructions, which when executed by a processor implement the following processes: an account restriction request for a target account is received. Obtaining limitation data used for limiting the target account, determining whether the target account is in an account white list or in an account blacklist based on the limitation data, if the target account is determined to be in the account blacklist, refusing to perform limitation removal on the target account, and if the target account is determined to be in the account whitelist, obtaining and executing a first limitation removal rule corresponding to the account whitelist so as to perform limitation removal on the target account. If the target account cannot be determined to be in the account white list or the account black list based on the limitation removal data, processing different types of data contained in the limitation removal data respectively through preset data processing rules to obtain corresponding limitation removal auxiliary information, wherein the limitation removal auxiliary information is used for triggering a limitation removal manager to perform limitation removal processing on the target account based on the limitation removal auxiliary information.

Drawings

In order to more clearly illustrate the embodiments of the present specification or the technical solutions in the prior art, the drawings required to be used in the description of the embodiments or the prior art will be briefly introduced below, it is obvious that the drawings in the following description are only some embodiments described in the present specification, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts;

FIG. 1A illustrates an embodiment of a data processing method of the present disclosure;

FIG. 1B is a schematic diagram of a data processing process according to the present description;

FIG. 2 is a block diagram of a data processing system according to the present description;

FIG. 3 is a schematic diagram of another data processing process described herein;

FIG. 4 is a block diagram of another data processing system according to the present description;

FIG. 5 is a diagram of one embodiment of a data processing apparatus according to the present disclosure;

fig. 6 is a data processing apparatus embodiment of the present description.

Detailed Description

The embodiment of the specification provides a data processing method, a data processing device and data processing equipment.

In order to make those skilled in the art better understand the technical solutions in the present specification, the technical solutions in the embodiments of the present specification will be clearly and completely described below with reference to the drawings in the embodiments of the present specification, and it is obvious that the described embodiments are only a part of the embodiments of the present specification, and not all of the embodiments. All other embodiments obtained by a person skilled in the art based on the embodiments in the present specification without any inventive step should fall within the scope of protection of the present specification.

Example one

As shown in fig. 1A and 1B, an execution subject of the method may be a terminal device or a server, where the terminal device may be a certain terminal device such as a mobile phone and a tablet computer, a computer device such as a notebook computer or a desktop computer, or an IoT device (specifically, a smart watch, a vehicle-mounted device, etc.). The server may be an independent server, or a server cluster formed by a plurality of servers, and the server may be a background server of financial service or online shopping service, or a background server of an application program. In this embodiment, a server is taken as an example to describe in detail, and for the execution process of the terminal device, reference may be made to the following relevant contents, which are not described herein again. The method specifically comprises the following steps:

in step S102, an account opening request for the target account is received.

In this embodiment, the target account may be an account registered for executing a certain service or a plurality of different services, where the services may include a plurality of services, for example, the services may be any services for performing transactions, specifically, online shopping, physical transactions, marketing events, transfer services, payment services, and the like, and may be specifically set according to an actual situation, which is not limited in this description. The account opening request may be a request for opening the authority of a certain account, where the authority of the account may be defined for one or more different authorities possessed by the certain account, for example, a payment authority of the account or a transfer authority of the account, which may be specifically set according to actual circumstances, and this is not limited in the embodiments of this specification.

In the implementation, in the field of risk prevention and control, a situation that misjudgment occurs inevitably to limit the use of one or more rights of a certain account, which causes great disturbance to user experience, so when a user complains about the misjudgment or requests to remove the right of the account, the user's appeal needs to be processed at the first time, whether the account of the user has risk is reexamined, and whether the right of the account of the user needs to be removed is determined, but in reexamination, different levels of audits needs often exist (for example, the audits need to be fast, the audits need to be accurate in result, and the audits need to be compliant and interpretable). Generally, a manual auditing manner may be adopted, for example, information of key transactions related to the account is extracted through a manual or preset first rule, whether a limitation removal certificate submitted by a user is qualified or not is judged through a manual or preset second rule, whether the account is at risk or not may also be comprehensively judged through a manual or preset third rule, and whether limitation removal processing is performed on the account is further determined. The embodiment of the present specification provides an implementable technical solution, which may specifically include the following contents:

as shown in fig. 2, for some services (e.g., a transfer service), when a user executes the service, information related to the service executed by the user may be obtained, and the recorded information may include information related to the service (e.g., information of a transaction class), identity class information, time for executing the service, information generated by each operation of the user during the service execution process, location information, and the like, and based on the information, if it is determined that the user or an account of the user (i.e., a target account) is at risk, permission for the target account to execute the service may be defined, so that the target account cannot execute the service, and if the user considers that the target account is at risk, an account restriction removal request may be generated for the target account by a terminal device, and the account restriction removal request may be sent to a corresponding server, and the server may receive the account restriction removal request for the target account.

Or, as shown in fig. 2, the server may record related information that a plurality of different users execute a service through respective accounts, and each time a preset period (7 days, 1 month, or the like) is reached, may acquire the related information that the plurality of different users execute the service through respective accounts within the period, and may determine whether there is a risk in an account (i.e., a target account) of a different user or a user through the acquired information, if it is determined that a user (or a plurality of users) or an account (i.e., a target account) of the user has a risk, may limit an authority of the target account to execute the service, so that the target account cannot execute the service, if the corresponding user considers that the target account does not have a risk, an account restriction removal request may be generated for the target account through the terminal device, and may send the account restriction removal request to the corresponding server, which may receive the account restriction removal request for the target account.

It should be noted that, the above are only two optional processing manners, and in practical application, the implementation may further include a plurality of different processing manners, which may be specifically set according to practical situations, and this is not limited in this embodiment of the present specification.

In step S104, limitation removing data for removing the target account is obtained, whether the target account is in an account white list or in an account black list is determined based on the limitation removing data, if the target account is determined to be in the account black list, the limitation removing processing on the target account is rejected, and if the target account is determined to be in the account white list, a first limitation removing rule corresponding to the account white list is obtained and executed, and the first limitation removing rule is used for removing the limitation on the target account.

The limitation data may include a plurality of types, for example, one or more items of structured data, images, texts, audios, and the like, specifically, one or more items of an identifier of the target account, user operation behavior data corresponding to the target account, a limitation certificate image of the target account, and the like, which may be specifically set according to an actual situation, and this is not limited in this specification. The account white list may be a list constructed by information of preset accounts without specified risks, and the accounts recorded in the account white list have no specified risks. The account blacklist may be a list constructed by presetting information of accounts with specified risks, and the accounts recorded in the account blacklist have the specified risks. The first limitation-removing rule may include multiple types, and in practical applications, the first limitation-removing rule may be a relatively simple rule for removing a certain right of an account, for example, a user requesting a target account provides limitation-removing credentials, specifically, an image of a certificate of the user, which can prove the identity of the user, of the user requesting the target account, or an image of a business license of the user, of the user requesting the target account, or an image of transaction flow data of the user, and the like, if information of the image obtained by the request matches with corresponding reference information of a pre-stored target account, the target account is subjected to limitation-removing processing is performed, otherwise, the target account is rejected, and the like, which may be specifically set according to practical situations and is not limited in the embodiments of the present specification.

In implementation, after the account restriction request is obtained in the above manner, the identifier of the target account may be extracted from the account restriction request, and in addition, the account restriction request may further include one or more of images, texts and structured data uploaded by the user, so that the target account can be quickly restricted. In this way, the data included in the account restriction request may be obtained, and in addition, other related information of the target account may also be obtained through the identifier of the target account, for example, related data generated in the process that the target account executes some services (which may include user operation behavior data and specific transaction data in executing the services, etc.) may be obtained, and the obtained data may be used as restriction data for restricting the target account. Then, a plurality of different algorithms or models may be preset, and different types of data may be processed by using corresponding algorithms or models, for example, for an image included in the limitation data, the image may be identified by using a preset image identification algorithm, content included in the image is determined (for example, image information, included characters, and the like included in the image are determined), whether the target account is in an account white list or an account black list may be obtained based on the determined content, in addition, the obtained result may be verified by using other data, if the obtained result is different from the result based on other data, the different number and the same number may be counted, so as to determine whether the target account is in the account white list or the account black list, for example, it is determined that the target account is in the account white list by using the image, if the different number is greater than the same number, it is determined that the target account is in the account black list, and if the different number is less than the same number, it is determined that the target account is in the account white list.

If the target account is determined to be in the account blacklist, the limitation removal processing on the target account can be directly refused, at this time, a notification message for refusing the limitation removal can be sent to a user of the target account, and relevant data of the account limitation removal request of the target account can be provided for an administrator, so that the administrator can further process the target account. If the target account is determined to be in the account white list, a relatively simple limitation removing mode can be set to remove the limitation of the target account, and based on the limitation removing mode, a first limitation removing rule corresponding to the account white list can be preset, for example, the first limitation removing rule can provide an image of a certificate of a user, the identity of which can be proved, for the user requesting the target account, if the information of the image obtained by the request is matched with corresponding certificate reference information of the target account stored in advance, the limitation removing processing is performed on the target account, otherwise, the limitation removing processing is refused to be performed on the target account, and the like. Then, a first limitation removing rule corresponding to the account white list may be obtained, and the first limitation removing rule may be used to remove the limitation of the target account. If the account is successfully defrosted, a notification message of successful defrosted can be sent to the user of the target account.

In step S106, if it cannot be determined that the target account is in the account white list or the account black list based on the limitation data, different types of data included in the limitation data are respectively processed through a preset data processing rule to obtain corresponding limitation auxiliary information, where the limitation auxiliary information is used to trigger a limitation manager to perform limitation removal processing on the target account based on the limitation auxiliary information.

The data processing rule may include a plurality of rules, such as an extraction rule of information in the image, a search rule of similar data, a trade intention restoration rule, and the like, and may be specifically set according to an actual situation, which is not limited in the embodiment of the present specification.

In the implementation, for the case that it is difficult to determine whether the target account is in the account white list or the account black list, one or more different data processing rules may be set in advance, the limitation removal data may be analyzed and processed through the set data processing rules to obtain corresponding information that is easy to determine whether the target account is in the account white list or the account black list, or whether the target account has the relevant information specifying risk may be easily determined, the obtained information may be used as the limitation removal auxiliary information, and subsequently, the limitation removal manager may perform limitation removal processing on the target account based on the limitation removal auxiliary information, specifically, if it is not determined that the target account is in the account white list or the account black list based on the limitation removal data, the different types of data included in the limitation removal data may be processed through the preset data processing rules to obtain the corresponding limitation removal auxiliary information, for example, for an image included in the limitation removal data, the extraction rule of the information included in the image may be used as the limitation removal information, so that the efficiency is improved through manual processing, and thus the efficiency is improved. In addition, for the text data included in the limitation data, the text data may be analyzed by using a preset semantic analysis model or a classification algorithm to obtain semantics corresponding to the text data and a category to which the text data belongs (for example, frequent interaction with a certain account in an account black list), and the obtained information may be used as the limitation auxiliary information. For other data (such as audio data or transaction data) included in the limitation data, the data may be analyzed by using a preset corresponding data model or data processing algorithm to obtain a corresponding analysis result, and the obtained analysis result may be used as limitation auxiliary information, or the like.

One or more different kinds of limitation-removing auxiliary information can be obtained through the method, the obtained limitation-removing auxiliary information can be provided for a limitation-removing management party (such as a service operator and the like), the limitation-removing management party can judge whether to perform limitation-removing processing on a target account or not based on the limitation-removing auxiliary information, if the limitation-removing management party determines that the limitation-removing processing can be performed on the target account, the limitation-removing processing can be directly performed on the target account, a notification message that the limitation-removing processing is successful can be sent to a user of the target account when the limitation-removing processing of the account is successful, if the limitation-removing processing of the target account is determined to be impossible, the limitation-removing processing of the target account is directly refused, and at the moment, a notification message that the limitation-removing processing is refused can be sent to the user of the target account.

The embodiment of the specification provides a data processing method, which includes the steps of obtaining limitation removal data for limiting a target account by receiving an account limitation removal request for the target account, determining whether the target account is in an account white list or in an account black list based on the limitation removal data, refusing to perform limitation removal processing on the target account if the target account is determined to be in the account black list, obtaining and executing a first limitation removal rule corresponding to the account white list to perform limitation removal processing on the target account if the target account is determined to be in the account white list, and performing processing respectively through preset data processing rules for different types of data contained in the limitation removal data if the target account cannot be determined to be in the account white list or the account black list based on the limitation removal data, the corresponding limit removal auxiliary information is obtained and used for triggering a limit removal manager to carry out limit removal processing on a target account based on the limit removal auxiliary information, so that on the basis of atomic algorithm capability, different algorithm capabilities are correspondingly provided through automatic auditing and interpretable auxiliary limit removal auditing, the efficiency of the limit removal auditing is greatly improved, the accuracy and the user experience are also greatly improved, in addition, the structured data and the text information are provided in the limit removal scene, and related information submitted by a user is also provided, so that the limit removal development is carried out by simultaneously utilizing multiple data, the limit removal auditing in a fraud domain needs to be provided with multiple capabilities of forgery prevention, modification prevention, network diagram prevention, watermark prevention, KV extraction prevention and the like by an algorithm, and the limit removal auditing can be multiplexed into multiple scenes.

Example two

As shown in fig. 3, an execution subject of the method may be a terminal device or a server, where the terminal device may be a certain terminal device such as a mobile phone and a tablet computer, may also be a computer device such as a notebook computer or a desktop computer, or may also be an IoT device (specifically, a smart watch, a vehicle-mounted device, etc.). The server may be an independent server, or a server cluster formed by a plurality of servers, and the server may be a background server of financial service or online shopping service, or a background server of an application program. In this embodiment, a server is taken as an example to describe in detail, and for the execution process of the terminal device, reference may be made to the following relevant contents, which are not described herein again. The method may specifically comprise the steps of:

in step S302, an account opening request for a target account is received.

The account restriction request may be a request received after risk detection is performed on transaction data of different users in a preset risk detection period when the preset risk detection period is reached, the target account is determined to be an account with a preset risk, and the target account is subjected to the right restriction processing, or the account restriction request may be a request received after the target account is subjected to the designated transaction and the designated transaction is rejected and the target account is subjected to the right restriction processing when the preset risk is detected. The risk detection period can be set according to actual conditions, specifically, 24 hours or 7 days.

In step S304, limitation data for limiting the target account is acquired.

The data format and length specification are strictly followed, and the data format and length specification are mainly stored and managed through a relational database. Unstructured data may be data that is not suitable for representation by a database two-dimensional table, including office documents of all formats, XML, HTML, various types of reports, image and audio, video information, and the like. The database supporting the unstructured data adopts a multi-value field, sub-field and variable length field mechanism to create and manage data items.

In step S306, the limitation data is input into a first multi-modal limitation model trained in advance, and an output result of whether the target account is in an account white list or in an account black list is obtained, where the first multi-modal limitation model is obtained by performing model training based on first history data as structured data and second history data as unstructured data.

The first multi-modal limitation solving model may be a model that processes a plurality of different data, where the plurality of different data may be structured data or unstructured data, and the first multi-modal limitation solving model may be constructed by a plurality of different algorithms, such as a classification algorithm, a neural network model, and the like, and may be specifically set according to an actual situation, which is not limited in the embodiment of the present specification.

In implementation, a corresponding algorithm may be obtained, a first multi-modal threshold model may be constructed based on the algorithm, input data of the first multi-modal threshold model may be structured data and/or unstructured data, output data may be an output result of whether a target account is in an account white list or an account black list, then a training sample for training the first multi-modal threshold model (i.e., first history data as structured data and second history data as unstructured data) may be obtained, the training sample may be used to perform model training on the first multi-modal threshold model, in the process of performing model training, a target function may be set in advance in consideration of that the training sample is simply encoded under an actual service scene to obtain a corresponding vector feature, and model parameters in the first multi-modal threshold model may be optimized based on the target function, wherein the first multi-modal threshold model may be adjusted for the target function. Then, model training can be performed on the first multi-modal solution model by using the training samples, and meanwhile, model parameters are optimized through the objective function, so that the trained first multi-modal solution model is finally obtained. Then, the limitation data can be input into a first multi-modal limitation model trained in advance, and an output result of whether the target account is in an account white list or in an account black list is obtained.

In practical applications, the limitation data may include voiceprint data input by the user, and then it may be determined whether the target account is in the account white list based on the voiceprint data input by the user, which may be specifically referred to the following processing in step S308 and step S310.

In step S308, the voiceprint data input by the user is matched with the reference voiceprint data of the accounts in the account whitelist, and it is determined whether the target account is in the account whitelist based on the obtained matching result.

In step S310, if the target account is not in the account white list, matching the voiceprint data input by the user with the reference voiceprint data of the account in the account black list, and determining whether the target account is in the account black list based on the obtained matching result.

In step S312, if it is determined that the target account is in the account blacklist, the restriction processing on the target account is rejected.

In step S314, if it is determined that the target account is in the account white list, the unlimited credential information is acquired from the user of the target account.

In step S316, the aforementioned limitation certificate information is input into a certificate recognition model trained in advance, whether the limitation certificate information is valid is determined, and if the limitation certificate information is valid, the limitation processing is performed on the target account.

The credential identification model may be a model for processing the information of the delimitation credential, and the credential identification model may be constructed by a plurality of different algorithms, such as a classification algorithm, a neural network model, and the like, and may be specifically set according to an actual situation, which is not limited in the embodiments of the present specification.

In the implementation, a corresponding algorithm may be obtained, a credential recognition model may be constructed based on the algorithm, input data of the credential recognition model may be limitation credential information provided by a user, output data may be whether the limitation credential information is valid, then, a training sample for training the credential recognition model (i.e., historical limitation credential information provided by the user) may be obtained, the credential recognition model may be model-trained using the training sample, in the process of model training, in consideration of a practical business scenario, a simple encoding process is performed on the training sample to obtain a corresponding vector feature, an objective function may be set in advance, and a model parameter in the credential recognition model may be optimized based on the objective function, where the credential recognition model may be adjusted for the objective function. Then, model training can be performed on the certificate recognition model by using the training sample, and meanwhile, model parameters are optimized through the objective function, so that the trained certificate recognition model is finally obtained. Then, the above-mentioned limitation certificate information can be input into a certificate recognition model trained in advance, and whether the limitation certificate information is valid or not can be determined.

In practical applications, the information of the unlimited certificate may be an image, and the processing of step S316 may be various, and an alternative processing manner is provided below, which may specifically include the processing of step A2 and step A4 below.

In step A2, the restriction credential information is input to a pre-trained credential identification model, and it is determined by the credential identification model whether the image matches pre-stored reference credential information.

Wherein the credential identification model may be a model for analyzing the identification image.

In step A4, if the image information is matched with the image information, extracting the image information from the image, and judging whether the image has a preset risk or not based on the extracted image information, if not, determining that the limitation certificate information is valid.

The preset risk may include multiple types, for example, the preset risk may include a fraud risk, an illegal financial activity, and the like, and may be specifically set according to an actual situation, which is not limited in the embodiment of the present specification.

In practical applications, the above limitation data may include historical transaction data for the target account, and then the above limitation data may be processed by the following step S318.

In step S318, if it is not determined that the target account is in the account white list or the account black list based on the limitation data, the historical transaction data is restored based on a pre-trained meta learning model, so as to obtain scene information and transaction intention information corresponding to the historical transaction data, and text information serving as limitation assist information is determined based on the obtained scene information and transaction intention information, and the meta learning model is used for restoring the scene and transaction intention corresponding to the transaction data.

The meta-learning model can be a model which can enable the meta-learning model to obtain the capacity of adjusting the hyper-parameters, so that the meta-learning model can rapidly learn a new task on the basis of obtaining the existing knowledge. The meta-learning model may be constructed through a plurality of different algorithms, specifically, the meta-learning model may be constructed through a convolutional neural network model, or the meta-learning model may be constructed through a classification algorithm (specifically, a binary classification algorithm, etc.), which may be specifically set according to an actual situation, and this is not limited in this description embodiment.

In the implementation, a corresponding algorithm may be obtained, a meta-learning model may be constructed based on the algorithm, input data of the meta-learning model may be historical transaction data of a user, output data may be a scenario and a transaction intention corresponding to the restored historical transaction data, then, a training sample (i.e., historical transaction data of the user) for training the meta-learning model may be obtained, the meta-learning model may be model-trained using the training sample, in the process of performing model training, in consideration of a simple encoding process performed on the training sample in an actual business scenario, a corresponding vector feature may be obtained, an objective function may be set in advance, and model parameters in the meta-learning model may be optimized based on the objective function, where the meta-learning model may be adjusted for the objective function. Then, model training can be performed on the meta-learning model by using the training samples, and meanwhile, model parameters are optimized through the objective function, so that the trained meta-learning model is finally obtained.

By means of the method, it is determined that the target account cannot be determined to be in the account white list or the account black list based on the limitation data, historical transaction data can be restored based on a pre-trained meta-learning model, scene information and transaction intention information corresponding to the historical transaction data are obtained, and text information serving as limitation auxiliary information is determined based on the obtained scene information and transaction intention information.

In step S320, if it cannot be determined that the target account is in the account white list or the account black list based on the limitation data, historical limitation data with a similarity greater than a first preset similarity threshold with the limitation data is acquired from the limitation database, and the acquired historical limitation data is used as the limitation auxiliary information.

The first preset similarity threshold may be set according to an actual situation, specifically, 80% or 90%.

In practical applications, the above-mentioned limitation data may include voiceprint data input by the user, and the following processing of step S322 may be performed.

In step S322, if it cannot be determined that the target account is in the account white list or the account black list based on the above limitation data, historical voiceprint data whose similarity with the voiceprint data input by the user is greater than a second preset similarity threshold is obtained from historical voiceprint data input by different users.

The second preset similarity threshold may be set according to an actual situation, specifically, 80% or 90%.

In step S324, the acquired historical voiceprint data and the user information corresponding to the acquired historical voiceprint data are taken as the limitation assist information.

In step S326, behavior data generated by the user corresponding to the target account within a preset time period after the account opening request is provided is obtained.

The preset time period can be set according to actual conditions, specifically 1 month or 1 year and the like. The behavior data may be the number of times the user makes a call, an operation behavior track on the application program, limitation credential information submitted by the user, and the like, which are acquired after the authority of the target account is limited, and may be specifically set according to an actual situation, which is not limited in the embodiments of the present specification.

In step S328, the behavior data and the limitation data are input into a second multi-modal limitation model trained in advance to obtain an output result of whether the target account has a preset risk, where the second multi-modal limitation model is obtained by performing model training based on third history data as structured data and fourth history data as unstructured data, and historical behavior data of the user.

The second multi-modal solution model may be a model that processes a plurality of different data, where the plurality of different data may be structured data or may include unstructured data, and the second multi-modal solution model may be constructed by a plurality of different algorithms, such as a classification algorithm, a neural network model, and the like, and may be specifically set according to an actual situation, which is not limited in the embodiments of the present specification.

In implementation, a corresponding algorithm may be obtained, a second multi-modal solution limit model may be constructed based on the algorithm, input data of the second multi-modal solution limit model may be structured data and/or unstructured data, and output data may be an output result of whether a preset risk exists in a target account, then, a training sample for training the second multi-modal solution limit model (i.e., third history data serving as structured data, fourth history data serving as unstructured data, and historical behavior data of a user) may be obtained, and the training sample may be used to perform model training on the second multi-modal solution limit model, and in a process of performing model training, considering that in an actual business scene, the training sample is simply encoded to obtain a corresponding vector feature, a target function may be preset, and model parameters in the second multi-modal solution limit model may be optimized based on the target function, where the second multi-modal solution limit model may be adjusted for the target function. Then, model training can be performed on the second multi-modal solution model by using the training samples, and meanwhile, model parameters are optimized through the objective function, so that the trained second multi-modal solution model is finally obtained. And then, the behavior data and the limitation data can be input into a second multi-modal limitation solving model to obtain an output result of whether the target account has a preset risk or not.

In addition, the transaction reduction algorithm is used for replacing manual work or rules to extract key transactions, the limitation removal voucher identification algorithm is used for replacing manual work or rules to judge whether the limitation removal voucher submitted by the user is effective or not, in addition, the multi-mode limitation removal algorithm is used for replacing manual work or rules to comprehensively judge whether the user needs limitation removal or not, the accuracy is high, the coverage area is wide, besides the algorithm, information such as voiceprints and interpretable characteristics can be provided, and the fact that whether the account is limited or not can be determined accurately and quickly is helped.

EXAMPLE III

Based on the same idea, the data processing method provided by the embodiment of the present specification further provides a data processing system, and an execution main body in the embodiment may be provided with a data processing system, as shown in fig. 4, the data processing system includes an atomic capability subsystem 410, an algorithm subsystem 420, and an application subsystem 430, where:

the atomic capability subsystem 410 configured to provide corresponding algorithmic support for the algorithmic hierarchy subsystem 430 and the application subsystem 420;

the application subsystem 420 is configured to receive an account limitation solving request for a target account, acquire limitation solving data for solving the target account, and call the algorithm system subsystem 430 for limitation solving processing based on the limitation solving data;

the algorithm architecture subsystem 430 is configured to invoke the algorithm in the atomic capability subsystem to perform the following: determining whether the target account is in an account white list or in an account black list or not based on the limitation removing data, if so, refusing to perform limitation removing processing on the target account, and if so, acquiring and executing a first limitation removing rule corresponding to the account white list so as to perform limitation removing processing on the target account; if the target account cannot be determined to be in the account white list or the account black list based on the limitation removal data, processing different types of data contained in the limitation removal data respectively through preset data processing rules to obtain corresponding limitation removal auxiliary information, wherein the limitation removal auxiliary information is used for triggering a limitation removal manager to perform limitation removal processing on the target account based on the limitation removal auxiliary information.

In the embodiment of this specification, the algorithm subsystem 430 includes a first deadline authority layer 431, a second deadline authority layer 432, and a third deadline authority layer 433, where:

the first limitation removal examination layer 431 is configured to determine whether the target account is in an account white list or an account black list based on the limitation removal data, refuse to perform limitation removal processing on the target account if the target account is determined to be in the account black list, and acquire and execute a first limitation removal rule corresponding to the account white list to perform limitation removal processing on the target account if the target account is determined to be in the account white list;

the second limitation-removal examination layer 432 is configured to, if it cannot be determined that the target account is in an account white list or an account black list based on the limitation-removal data, perform processing on different types of data included in the limitation-removal data respectively through preset data processing rules to obtain corresponding limitation-removal auxiliary information;

the third unrestricted review layer 433 is configured to perform the restriction processing on the target account by:

performing limitation removal processing on the target account based on the limitation removal auxiliary information;

acquiring historical limitation data with the similarity between the limitation data and the limitation data larger than a first preset similarity threshold from a limitation database, and carrying out limitation removal processing on the target account based on the acquired historical limitation data;

and restoring historical transaction data in the limitation data based on a pre-trained meta-learning model to obtain scene information and transaction intention information corresponding to the historical transaction data, respectively converting the scene information and the transaction intention information into text information, and performing limitation processing on the target account based on the converted text information, wherein the meta-learning model is used for restoring the scene and the transaction intention corresponding to the transaction data.

In this embodiment of the present specification, the atomic capability subsystem 410 is provided with one or more of a first multi-modal limitation resolving model, a second multi-modal limitation resolving model, a credential recognition model, a first limitation resolving rule corresponding to an account white list, a meta-learning model, a transaction reduction algorithm, a data-to-text conversion algorithm, an image-oriented retrieval algorithm, a voiceprint recognition algorithm, and a voiceprint clustering algorithm, where the first multi-modal limitation resolving model is obtained by performing model training based on first history data serving as structured data and second history data serving as unstructured data, the second multi-modal limitation resolving model is obtained by performing model training based on third history data serving as structured data and fourth history data serving as unstructured data, and historical behavior data of a user.

The first multimodal solution limit model can identify whether the user is at risk or not by using various data forms (such as structured data, pictures, texts, and the like), and is directly used for performing automatic processing in the first solution limit examination layer 431 in the data processing system. When the meta-learning model is used for restoring transaction data, the restoration categories of multiple small sample transaction data need to be judged, and the meta-learning model has good application in the learning of meta-learning small samples. The transaction recovery algorithm can recover each transaction, so as to identify which intention the transaction belongs to (multi-classification processing can be performed), and can classify the transaction by using a meta-learning model besides a mode of using a specified strategy. The input data of the data-to-text conversion algorithm may be a characteristic of the user, and the output data may be a piece of interpretable text information describing the behavior of the user, and if the user or the account of the user has a risk, it may describe which risk behaviors of the user or the account of the user include. If the user or the account of the user has no risk, the normal behavior of the user, such as normal life consumption, enterprise management and the like, is described by combining the transaction recovery algorithm. The retrieval algorithm for the image may retrieve a plurality of solution limit cases similar to the current solution limit data from the solution limit database as the solution limit auxiliary information. The voiceprint recognition algorithm determines whether the target account is in the account blacklist by matching the voiceprint data of the user with the user voiceprint data of the account in the account blacklist. The voiceprint clustering algorithm can be based on the voiceprint data of the current user, find out whether the voiceprint data of other users who ask for solving the limit is similar to the voiceprint data of the current user so as to obtain whether the users are the same person, if the users are the same person, the users are clustered into a category, and therefore whether the users ask for solving the limit for a plurality of users in batches is judged.

Besides, the atomic capability subsystem 410 may further include a plurality of different algorithms, such as an image algorithm and a certificate recognition algorithm, where the image algorithm is that the user submits the limitation-free certificate information during the limitation-free process, and the limitation-free certificate information is generally an image, so that the image algorithm is required to perform the automated process in the first limitation-free examination layer 431. For the certificate recognition algorithm, if the information of the unlimited certificate submitted by the user is an image (specifically, a certificate capable of proving the identity of the user, an image of a business license, and an image of a bank flow), the recognition process mainly includes multiple functions, such as anti-counterfeiting, anti-modification, anti-duplication, and whether the judgment is clear or not. In addition, it can also have an important function: the information in the information of the restriction voucher is extracted, such as identification, date and sex in the certificate which can prove the identity of the user, time, amount, number of strokes in bank water and the like, the information needs to spend a great deal of energy if being processed manually, and if the algorithm is used, great manpower resources can be saved.

The first limitation removal examination layer 431 can use the first multi-modal limitation removal model to identify the account in the high-precision account blacklist or the account in the account whitelist, and then automatically process the account without manual intervention, so that the time is greatly saved, and the examination efficiency and the throughput are improved, wherein: and refusing limitation removal processing on the determined account in the account blacklist, or actively removing the limitation on the determined account in the account whitelist. In addition, the account in the account white list only needs to be subjected to simple limitation removing processing.

The second limit-based auditing layer 432, namely the layer of algorithm, provides a plurality of interpretable results, and forms the interpretable results into a certain structural form, so as to prove that the target account belongs to the account white list, or prove that the target account belongs to the account white list, and an auditor can finish fast auditing according to the auxiliary information, so that auditing efficiency is improved, wherein:

white users (i.e., users corresponding to accounts in the account white list): the user with the accuracy rate of 80% can preferentially perform the limitation removing process, and only simple limitation removing voucher information needs to be provided, so that the user experience can be effectively improved. Transaction recovery & Data2Text: whether risk exists or not is judged for each transaction, and scene information and transaction intention information of the transaction are restored, for example, a user purchases things mainly at a certain E-commerce merchant in multiple transactions in the near term, and detailed logistics information and receiving time exist, so that the multiple transactions can be considered as normal shopping consumption, and a limit-removing management party can rapidly limit the account based on the information. A second multi-modal solution model (i.e., a post-hoc multi-modal solution model): after the authority of the user is limited, behavior data such as the number of times of making calls of the user, operation behavior tracks on an application program, limitation removal certificate information submitted by the user and the like can be acquired, whether risks exist in the account can be identified more accurately by combining the after-event information (namely the behavior data), and the limitation removal manager can rapidly limit the account based on the behavior data.

The third solution examination layer 433 may be some solution examination cases with very little acquired information, and needs to be examined in combination with interpretable information provided by an algorithm: feature interpretable & Data2Text algorithm: providing main judgment characteristics for judging whether risks exist or not, and summarizing the risk behaviors or normal transaction behaviors of the user by using texts so as to assist in judgment; and (3) a retrieval algorithm: for cases which are difficult to be qualified, the case closest to the case can be found from the limitation database, so that the limitation management party is helped to carry out limitation processing.

The application subsystem 420 is configured to, when a designated transaction is performed on the target account and a preset risk is detected to exist in the designated transaction, reject the designated transaction and perform a right-limiting process on the target account, and receive an account limitation-removing request for the target account, where the process may be a process executed in a real-time transaction process; or when a preset risk detection period is reached, performing risk detection on transaction data of different users in the risk detection period, and if the target account is determined to be an account with a preset risk, performing right-limiting processing on the target account, and receiving an account limitation-removing request for the target account, where the processing may be processing executed in an offline transaction.

In this embodiment of the present specification, the limitation data includes structured data and/or unstructured data, and the first limitation removal examination layer 431 is configured to input the limitation removal data into a pre-trained first multi-modal limitation removal model, and obtain an output result of whether the target account is in an account white list or in an account black list, where the first multi-modal limitation removal model is obtained by performing model training based on first historical data that is structured data and second historical data that is unstructured data.

In this embodiment of the present specification, the limitation removal data includes voiceprint data input by a user, and the first limitation removal examination layer 431 is configured to match the voiceprint data input by the user with reference voiceprint data of an account in the account whitelist, and determine whether the target account is in the account whitelist based on an obtained matching result; and if the target account is not in the account white list, matching the voiceprint data input by the user with the reference voiceprint data of the account in the account black list, and determining whether the target account is in the account black list or not based on the obtained matching result.

In an embodiment of this specification, the limitation data includes historical transaction data for the target account, the second limitation-limiting examination layer 432 is configured to restore the historical transaction data based on a pre-trained meta-learning model, obtain scene information and transaction intention information corresponding to the historical transaction data, and determine text information serving as limitation-limiting auxiliary information based on the obtained scene information and transaction intention information, where the meta-learning model is used to restore the scene and transaction intention corresponding to the transaction data.

In this embodiment of the present specification, the first limitation-removal auditing layer 431 is configured to, if it is determined that the target account is in the account white list, obtain limitation-removal credential information from a user of the target account; inputting the information of the limitation-free voucher into a pre-trained voucher identification model, determining whether the information of the limitation-free voucher is effective, and if the information of the limitation-free voucher is effective, carrying out limitation-free processing on the target account.

In this embodiment of the present specification, the limitation-removed credential information is an image, and the first limitation-removed examination layer 431 is configured to input the limitation-removed credential information into a pre-trained credential identification model, and determine, by using the credential identification model, whether the image matches pre-stored reference credential information; and if the image information is matched with the image information, extracting the image information from the image, judging whether the image has a preset risk or not based on the extracted image information, and if not, determining that the limitation voucher information is effective.

In this embodiment of the present specification, the second limitation limiting examination layer 432 is configured to acquire, from a limitation data base, historical limitation limiting data whose similarity with the limitation limiting data is greater than a first preset similarity threshold, and use the acquired historical limitation limiting data as the limitation limiting auxiliary information.

In this embodiment of the present specification, the limitation data includes voiceprint data input by a user, and the second limitation-solving examining layer 432 is configured to obtain, from historical voiceprint data input by different users, historical voiceprint data whose similarity with the voiceprint data input by the user is greater than a second preset similarity threshold; and taking the acquired historical voiceprint data and the user information corresponding to the acquired historical voiceprint data as the auxiliary information for the limitation.

In this embodiment of the present specification, the second limitation-removing examining layer 432 is configured to obtain behavior data generated by a user corresponding to the target account within a preset time after the user provides the account limitation-removing request; and inputting the behavior data and the limitation data into a pre-trained second multi-modal limitation model to obtain an output result of whether the target account has a preset risk, wherein the second multi-modal limitation model is obtained by performing model training on the basis of third history data serving as structured data, fourth history data serving as unstructured data and historical behavior data of the user.

For the specific processing procedures of the above parts, reference may be made to relevant contents in the above first embodiment and second embodiment, which are not described herein again.

The embodiment of the specification provides a data processing system, which obtains limitation removal data for limiting a target account by receiving an account limitation removal request for the target account, determines whether the target account is in an account white list or in an account black list based on the limitation removal data, refuses to perform limitation removal processing on the target account if the target account is determined to be in the account black list, obtains and executes a first limitation removal rule corresponding to the account white list to perform limitation removal processing on the target account if the target account is determined to be in the account white list, and respectively performs processing through preset data processing rules for different types of data contained in the limitation removal data if the target account cannot be determined to be in the account white list or the account black list based on the limitation removal data, the corresponding limit removal auxiliary information is obtained and used for triggering a limit removal manager to carry out limit removal processing on a target account based on the limit removal auxiliary information, so that on the basis of atomic algorithm capability, different algorithm capabilities are correspondingly provided through automatic auditing and interpretable auxiliary limit removal auditing, the efficiency of the limit removal auditing is greatly improved, the accuracy and the user experience are also greatly improved, in addition, the structured data and the text information are provided in the limit removal scene, and related information submitted by a user is also provided, so that the limit removal development is carried out by simultaneously utilizing multiple data, the limit removal auditing in a fraud domain needs to be provided with multiple capabilities of forgery prevention, modification prevention, network diagram prevention, watermark prevention, KV extraction prevention and the like by an algorithm, and the limit removal auditing can be multiplexed into multiple scenes.

Example four

Based on the same idea, the data processing system provided in the embodiment of the present specification further provides a data processing apparatus, as shown in fig. 5.

The data processing apparatus includes: a delimitation request module 501, a first delimitation module 502, and a second delimitation module 503, wherein:

a limitation removal request module 501, which receives an account limitation removal request for a target account;

the first limitation removing module 502 is configured to obtain limitation removing data for removing the target account, determine whether the target account is in an account white list or an account blacklist based on the limitation removing data, reject limitation removing processing on the target account if it is determined that the target account is in the account blacklist, and obtain and execute a first limitation removing rule corresponding to the account white list to remove the limitation on the target account if it is determined that the target account is in the account white list;

if the target account cannot be determined to be in the account white list or the account black list based on the limitation data, the second limitation removing module 503 processes different types of data included in the limitation removing data respectively through a preset data processing rule to obtain corresponding limitation removing auxiliary information, where the limitation removing auxiliary information is used to trigger a limitation removing manager to remove the limitation from the target account based on the limitation removing auxiliary information.

In an embodiment of the present specification, the limitation data includes structured data and/or unstructured data, and the first limitation removing module 502 inputs the limitation removing data into a first multi-modal limitation removing model trained in advance to obtain an output result of whether the target account is in an account white list or in an account black list, where the first multi-modal limitation removing model is obtained by performing model training based on first historical data serving as structured data and second historical data serving as unstructured data.

In this embodiment of the present specification, the limitation data includes voiceprint data input by a user, and the first limitation resolving module 502 includes:

the first matching unit is used for matching the voiceprint data input by the user with the reference voiceprint data of the account in the account white list and determining whether the target account is in the account white list or not based on the obtained matching result;

and the second matching unit is used for matching the voiceprint data input by the user with the reference voiceprint data of the account in the account blacklist if the target account is not in the account whitelist, and determining whether the target account is in the account blacklist or not based on the obtained matching result.

In an embodiment of this specification, the limitation data includes historical transaction data for the target account, the second limitation module 503 performs reduction processing on the historical transaction data based on a pre-trained meta-learning model, obtains scene information and transaction intention information corresponding to the historical transaction data, and determines text information serving as limitation-limiting auxiliary information based on the obtained scene information and transaction intention information, where the meta-learning model is used to perform reduction processing on a scene and a transaction intention corresponding to the transaction data.

In this embodiment, the first de-limiting module 502 includes:

the information acquisition unit is used for acquiring the information of the limitation-free certificate from the user of the target account if the target account is determined to be in the account white list;

the first limitation removing unit inputs the limitation removing certificate information into a certificate identification model trained in advance, determines whether the limitation removing certificate information is valid, and if the limitation removing certificate information is valid, carries out limitation removing processing on the target account.

In an embodiment of the present specification, the credential information is an image, and the first limitation unit inputs the credential information into a pre-trained credential recognition model, and determines whether the image matches pre-stored reference credential information or not through the credential recognition model; if the image information is matched with the image information, extracting the image information from the image, judging whether the image has a preset risk or not based on the extracted image information, and if not, determining that the information of the limitation certificate is effective.

In this embodiment of the present specification, the second limitation limiting module 502 obtains, from a limitation data base, historical limitation data whose similarity with the limitation data is greater than a first preset similarity threshold, and uses the obtained historical limitation data as the limitation auxiliary information.

In this embodiment of the present specification, the limitation data includes voiceprint data input by a user, and the second limitation module 502 includes:

the data acquisition unit is used for acquiring historical voiceprint data, the similarity of which with the voiceprint data input by the user is greater than a second preset similarity threshold value, from historical voiceprint data input by different users;

and the auxiliary information determining unit is used for taking the acquired historical voiceprint data and the user information corresponding to the acquired historical voiceprint data as the limitation auxiliary information.

In an embodiment of this specification, the apparatus further includes:

the behavior data acquisition module is used for acquiring behavior data which are generated by a user corresponding to the target account within a preset time after the account limitation solving request is provided;

and the risk determining module is used for inputting the behavior data and the limitation data into a pre-trained second multi-modal limitation model to obtain an output result of whether the target account has a preset risk or not, wherein the second multi-modal limitation model is obtained by performing model training on the basis of third history data serving as structured data, fourth history data serving as unstructured data and historical behavior data of a user.

In this embodiment of the present specification, the account restriction request is a request that is received after performing risk detection on transaction data of different users in a preset risk detection period when the preset risk detection period is reached, and determining that the target account is an account with a preset risk, and performing a right restriction process on the target account, or the account restriction request is a request that is received after performing an appointed transaction on the target account, and detecting that the preset risk exists in the appointed transaction, the appointed transaction is rejected, and performing the right restriction process on the target account.

The embodiment of the specification provides a data processing device, which obtains limitation removal data for limiting a target account by receiving an account limitation removal request for the target account, determines whether the target account is in an account white list or in an account black list based on the limitation removal data, refuses to perform limitation removal processing on the target account if the target account is determined to be in the account black list, obtains and executes a first limitation removal rule corresponding to the account white list to perform limitation removal processing on the target account if the target account is determined to be in the account white list, and respectively performs processing through preset data processing rules for different types of data contained in the limitation removal data if the target account cannot be determined to be in the account white list or the account black list based on the limitation removal data, the corresponding limit removal auxiliary information is obtained and used for triggering a limit removal manager to carry out limit removal processing on a target account based on the limit removal auxiliary information, so that on the basis of atomic algorithm capability, different algorithm capabilities are correspondingly provided through automatic auditing and interpretable auxiliary limit removal auditing, the efficiency of the limit removal auditing is greatly improved, the accuracy and the user experience are also greatly improved, in addition, the structured data and the text information are provided in the limit removal scene, and related information submitted by a user is also provided, so that the limit removal development is carried out by simultaneously utilizing multiple data, the limit removal auditing in a fraud domain needs to be provided with multiple capabilities of forgery prevention, modification prevention, network diagram prevention, watermark prevention, KV extraction prevention and the like by an algorithm, and the limit removal auditing can be multiplexed into multiple scenes.

EXAMPLE five

Based on the same idea, the data processing system provided in the embodiment of the present specification further provides a data processing apparatus, as shown in fig. 6.

The data processing device may provide a terminal device or a server, etc. for the above embodiments.

The data processing apparatus may have a large difference due to different configurations or performances, and may include one or more processors 601 and a memory 602, and one or more stored applications or data may be stored in the memory 602. Wherein the memory 602 may be transient or persistent storage. The application program stored in memory 602 may include one or more modules (not shown), each of which may include a series of computer-executable instructions for the data processing device. Still further, the processor 601 may be arranged in communication with the memory 602 to execute a series of computer executable instructions in the memory 602 on a data processing device. The data processing apparatus may also include one or more power supplies 603, one or more wired or wireless network interfaces 604, one or more input-output interfaces 605, one or more keyboards 606.

In particular, in this embodiment, the data processing apparatus includes a memory, and one or more programs, wherein the one or more programs are stored in the memory, and the one or more programs may include one or more modules, and each module may include a series of computer-executable instructions for the data processing apparatus, and the one or more programs configured to be executed by the one or more processors include computer-executable instructions for:

receiving an account restriction request for a target account;

acquiring limitation removing data used for removing the limitation of the target account, determining whether the target account is in an account white list or an account black list based on the limitation removing data, if the target account is determined to be in the account black list, refusing to perform limitation removing processing on the target account, and if the target account is determined to be in the account white list, acquiring and executing a first limitation removing rule corresponding to the account white list so as to perform limitation removing processing on the target account;

if the target account cannot be determined to be in the account white list or the account black list based on the limitation removal data, processing different types of data contained in the limitation removal data respectively through preset data processing rules to obtain corresponding limitation removal auxiliary information, wherein the limitation removal auxiliary information is used for triggering a limitation removal manager to perform limitation removal processing on the target account based on the limitation removal auxiliary information.

In this embodiment, the determining whether the target account is in the white list or the black list based on the limitation data includes:

and inputting the limitation solving data into a pre-trained first multi-modal limitation solving model to obtain an output result of whether the target account is in an account white list or in an account black list, wherein the first multi-modal limitation solving model is obtained by performing model training on the basis of first historical data serving as structured data and second historical data serving as unstructured data.

In an embodiment of this specification, the determining, based on the limitation data, whether the target account is in an account white list or in an account black list includes:

matching voice print data input by a user with reference voice print data of an account in the account white list, and determining whether the target account is in the account white list or not based on an obtained matching result;

and if the target account is not in the account white list, matching the voiceprint data input by the user with the reference voiceprint data of the account in the account black list, and determining whether the target account is in the account black list or not based on the obtained matching result.

In an embodiment of this specification, the limitation data includes historical transaction data for the target account, and the different types of data included in the limitation data are respectively processed by preset data processing rules to obtain corresponding limitation-solving auxiliary information, where the limitation-solving auxiliary information includes:

and restoring the historical transaction data based on a pre-trained meta-learning model to obtain scene information and transaction intention information corresponding to the historical transaction data, and determining text information serving as limitation-solving auxiliary information based on the obtained scene information and transaction intention information, wherein the meta-learning model is used for restoring the scene and the transaction intention corresponding to the transaction data.

In an embodiment of this specification, if it is determined that the target account is in the account white list, acquiring and executing a first limitation removing rule corresponding to the account white list to perform limitation removing processing on the target account, where the method includes:

if the target account is determined to be in the account white list, acquiring limitation-free certificate information from a user of the target account;

inputting the information of the limitation-free voucher into a pre-trained voucher identification model, determining whether the information of the limitation-free voucher is effective, and if the information of the limitation-free voucher is effective, carrying out limitation-free processing on the target account.

In an embodiment of this specification, the information of the solution certificate is an image, and the inputting the information of the solution certificate into a certificate recognition model trained in advance to determine whether the information of the solution certificate is valid includes:

inputting the information of the limitation certificate into a certificate recognition model trained in advance, and judging whether the image is matched with pre-stored reference certificate information or not through the certificate recognition model;

and if the image information is matched with the image information, extracting the image information from the image, judging whether the image has a preset risk or not based on the extracted image information, and if not, determining that the limitation voucher information is effective.

In an embodiment of this specification, the processing different types of data included in the limitation removal data through a preset data processing rule respectively to obtain corresponding limitation removal auxiliary information includes:

and acquiring historical limitation data with the similarity between the historical limitation data and the limitation data larger than a first preset similarity threshold from a limitation database, and taking the acquired historical limitation data as the limitation auxiliary information.

In an embodiment of this specification, the limitation data includes voiceprint data input by a user, and the processing is performed on different types of data included in the limitation data respectively according to preset data processing rules to obtain corresponding limitation-solving auxiliary information, where the processing includes:

acquiring historical voiceprint data with the similarity between the voiceprint data input by different users and the voiceprint data input by the users larger than a second preset similarity threshold value from the historical voiceprint data input by different users;

and taking the acquired historical voiceprint data and the user information corresponding to the acquired historical voiceprint data as the auxiliary information for the limitation.

In the embodiment of this specification, the method further includes:

acquiring behavior data generated by a user corresponding to the target account within a preset time after the account limitation removing request is provided;

and inputting the behavior data and the limitation data into a pre-trained second multi-modal limitation model to obtain an output result of whether the target account has a preset risk, wherein the second multi-modal limitation model is obtained by performing model training on the basis of third historical data serving as structured data, fourth historical data serving as unstructured data and historical behavior data of the user.

In addition, the transaction reduction algorithm is used for replacing manual or rule extraction key transactions, the limit removal voucher identification algorithm is used for replacing manual or rule judgment whether the limit removal voucher submitted by the user is effective, in addition, the multi-mode limit removal algorithm is used for replacing manual or rule comprehensive judgment whether the user needs limit removal, the accuracy is high, the coverage is wide, besides the algorithm, voiceprint, interpretable characteristics and other information can be provided, and the user can be helped to determine whether the account is limited more accurately and more quickly.

EXAMPLE six

Further, based on the methods shown in fig. 1A to fig. 3, one or more embodiments of the present specification further provide a storage medium for storing computer-executable instruction information, in a specific embodiment, the storage medium may be a usb disk, an optical disk, a hard disk, and the like, and when the storage medium stores the computer-executable instruction information, the storage medium implements the following processes:

receiving an account restriction request for a target account;

if the target account is determined to be in the account white list, acquiring limitation-removing credential information from a user of the target account;

In the embodiment of this specification, the method further includes:

and inputting the behavior data and the limitation data into a pre-trained second multi-modal limitation model to obtain an output result of whether the target account has a preset risk, wherein the second multi-modal limitation model is obtained by performing model training on the basis of third history data serving as structured data, fourth history data serving as unstructured data and historical behavior data of the user.

The embodiment of the specification provides a storage medium, which obtains limitation removal data for limiting a target account by receiving an account limitation removal request for the target account, determines whether the target account is in an account white list or in an account black list based on the limitation removal data, refuses to perform limitation removal processing on the target account if the target account is determined to be in the account black list, obtains and executes a first limitation removal rule corresponding to the account white list to perform limitation removal processing on the target account if the target account is determined to be in the account white list, and respectively processes different types of data contained in the limitation removal data through preset data processing rules if the target account cannot be determined to be in the account white list or the account black list based on the limitation removal data, the corresponding limit removal auxiliary information is obtained and used for triggering a limit removal manager to carry out limit removal processing on a target account based on the limit removal auxiliary information, so that on the basis of atomic algorithm capability, different algorithm capabilities are correspondingly provided through automatic auditing and interpretable auxiliary limit removal auditing, the efficiency of the limit removal auditing is greatly improved, the accuracy and the user experience are also greatly improved, in addition, the structured data and the text information are provided in the limit removal scene, and related information submitted by a user is also provided, so that the limit removal development is carried out by simultaneously utilizing multiple data, the limit removal auditing in a fraud domain needs to be provided with multiple capabilities of forgery prevention, modification prevention, network diagram prevention, watermark prevention, KV extraction prevention and the like by an algorithm, and the limit removal auditing can be multiplexed into multiple scenes.

The foregoing description has been directed to specific embodiments of this disclosure. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.

In the 90's of the 20 th century, improvements to a technology could clearly distinguish between improvements in hardware (e.g., improvements to circuit structures such as diodes, transistors, switches, etc.) and improvements in software (improvements to process flow). However, as technology advances, many of today's process flow improvements have been seen as direct improvements in hardware circuit architecture. Designers almost always obtain the corresponding hardware circuit structure by programming an improved method flow into the hardware circuit. Thus, it cannot be said that an improvement in the process flow cannot be realized by hardware physical blocks. For example, a Programmable Logic Device (PLD), such as a Field Programmable Gate Array (FPGA), is an integrated circuit whose Logic functions are determined by programming the Device by a user. A digital system is "integrated" on a PLD by the designer's own programming without requiring the chip manufacturer to design and fabricate application-specific integrated circuit chips. Furthermore, nowadays, instead of manually manufacturing an integrated Circuit chip, such Programming is often implemented by "logic compiler" software, which is similar to a software compiler used in program development and writing, but the original code before compiling is also written by a specific Programming Language, which is called Hardware Description Language (HDL), and HDL is not only one but many, such as abll (advanced desktop Expression Language), AHDL (alternate Hardware Description Language), traffic, CUPL (computer unified Programming Language), HDCal, jhddl (Java Description Language), lava, lola, HDL, PALASM, rhyd (Hardware Description Language), and the like, which are currently used in the field-Hardware Language. It will also be apparent to those skilled in the art that hardware circuitry that implements the logical method flows can be readily obtained by merely slightly programming the method flows into an integrated circuit using the hardware description languages described above.

The controller may be implemented in any suitable manner, for example, the controller may take the form of, for example, a microprocessor or processor and a computer-readable medium storing computer-readable program code (e.g., software or firmware) executable by the (micro) processor, logic gates, switches, an Application Specific Integrated Circuit (ASIC), a programmable logic controller, and an embedded microcontroller, examples of which include, but are not limited to, the following microcontrollers: ARC 625D, atmel at91SAM, microchip PIC18F26K20, and Silicone Labs C8051F320, the memory controller may also be implemented as part of the control logic of the memory. Those skilled in the art will also appreciate that, in addition to implementing the controller as pure computer readable program code, the same functionality can be implemented by logically programming method steps such that the controller is in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers and the like. Such a controller may thus be considered a hardware component, and the means included therein for performing the various functions may also be considered as a structure within the hardware component. Or even means for performing the functions may be regarded as being both a software module for performing the method and a structure within a hardware component.

The systems, apparatuses, modules or units described in the above embodiments may be specifically implemented by a computer chip or an entity, or implemented by a product with certain functions. One typical implementation device is a computer. In particular, the computer may be, for example, a personal computer, a laptop computer, a cellular telephone, a camera phone, a smartphone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

For convenience of description, the above devices are described as being divided into various units by function, and are described separately. Of course, the functionality of the various elements may be implemented in the same one or more pieces of software and/or hardware in implementing one or more embodiments of the present description.

As will be appreciated by one skilled in the art, embodiments of the present description may be provided as a method, system, or computer program product. Accordingly, one or more embodiments of the present description may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, one or more embodiments of the present description may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

Embodiments of the present description are described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the description. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable fraud case serial-parallel apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable fraud case serial-parallel apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable fraud case to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable fraud case serial-parallel apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

In a typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.

The memory may include forms of volatile memory in a computer readable medium, random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). Memory is an example of a computer-readable medium.

Computer-readable media, including both permanent and non-permanent, removable and non-removable media, may implement the information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer readable medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.

It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrases "comprising one of 8230; \8230;" 8230; "does not exclude the presence of additional like elements in a process, method, article, or apparatus that comprises that element.

One or more embodiments of the present description may be described in the general context of computer-executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. One or more embodiments of the specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including memory storage devices.

The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, as for the system embodiment, since it is substantially similar to the method embodiment, the description is relatively simple, and reference may be made to the partial description of the method embodiment for relevant points.

The above description is only an example of the present specification, and is not intended to limit the present application. Various modifications and alterations to this description will become apparent to those skilled in the art. Any modification, equivalent replacement, improvement or the like made within the spirit and principle of the present specification should be included in the scope of the claims of the present specification.

Claims

1. A method of data processing, the method comprising:

receiving an account restriction request for a target account;

2. The method of claim 1, the threshold data comprising structured data and/or unstructured data, the determining whether the target account is in an account whitelist or in an account blacklist based on the threshold data comprising:

3. The method of claim 1, the decommissioning data comprising user-entered voiceprint data, the determining whether the target account is in an account whitelist or in an account blacklist based on the decommissioning data comprising:

matching the voiceprint data input by the user with the reference voiceprint data of the accounts in the account white list, and determining whether the target account is in the account white list or not based on the obtained matching result;

4. The method according to claim 1, wherein the limitation data includes historical transaction data for the target account, and different types of data included in the limitation data are respectively processed by preset data processing rules to obtain corresponding limitation-solving auxiliary information, including:

5. The method of claim 1, wherein if it is determined that the target account is in the account white list, acquiring and executing a first limitation removing rule corresponding to the account white list to perform limitation removing processing on the target account, includes:

6. The method of claim 5, wherein the decommissioning credential information is an image, and wherein inputting the decommissioning credential information into a pre-trained credential recognition model to determine whether the decommissioning credential information is valid comprises:

inputting the information of the restriction voucher into a pre-trained voucher identification model, and judging whether the image is matched with pre-stored reference voucher information or not through the voucher identification model;

7. The method according to claim 1, wherein the processing different types of data included in the limitation data by preset data processing rules respectively to obtain corresponding limitation auxiliary information includes:

8. The method according to claim 1, wherein the limitation data includes voiceprint data input by a user, and the processing is performed on different types of data included in the limitation data through preset data processing rules respectively to obtain corresponding limitation auxiliary information, including:

acquiring historical voiceprint data, the similarity of which with the voiceprint data input by the user is greater than a second preset similarity threshold value, from historical voiceprint data input by different users;

9. The method of claim 1, further comprising:

10. The method according to any one of claims 1 to 9, wherein the account restriction request is a request received after risk detection is performed on transaction data of different users in a preset risk detection period when the preset risk detection period is reached, the target account is determined to be an account with a preset risk, and the target account is subjected to the right restriction processing, or the account restriction request is a request received after the target account performs a designated transaction and the designated transaction is detected to have a preset risk, the designated transaction is rejected and the target account is subjected to the right restriction processing.

11. A data processing system, the system comprising an atomic capability subsystem, an algorithmic system subsystem, and an application subsystem, wherein:

the atomic capability subsystem is configured to provide corresponding algorithm support for the algorithm system subsystem and the application subsystem;

the application subsystem is configured to receive an account limitation solving request for a target account, acquire limitation solving data for solving the target account, and call the algorithm system subsystem to carry out limitation solving processing based on the limitation solving data;

the algorithm system subsystem is configured to call the algorithm in the atomic capability subsystem to execute the following processing: determining whether the target account is in an account white list or in an account black list or not based on the limitation removing data, if so, refusing to perform limitation removing processing on the target account, and if so, acquiring and executing a first limitation removing rule corresponding to the account white list so as to perform limitation removing processing on the target account; if the target account cannot be determined to be in the account white list or the account black list based on the limitation removal data, processing different types of data contained in the limitation removal data respectively through preset data processing rules to obtain corresponding limitation removal auxiliary information, wherein the limitation removal auxiliary information is used for triggering a limitation removal manager to perform limitation removal processing on the target account based on the limitation removal auxiliary information.

12. The system of claim 11, wherein the atomic capability subsystem is provided with one or more of a first multi-modal limitation-removing model, a second multi-modal limitation-removing model, a credential recognition model, a first limitation-removing rule corresponding to an account white list, a meta-learning model, a transaction reduction algorithm, a data-to-text conversion algorithm, a retrieval algorithm for images, a voiceprint recognition algorithm, and a voiceprint clustering algorithm, wherein the first multi-modal limitation-removing model is obtained by model training based on first historical data serving as structured data and second historical data serving as unstructured data, the second multi-modal limitation-removing model is obtained by model training based on third historical data serving as structured data and fourth historical data serving as unstructured data, and historical behavior data of a user.

13. The system of claim 11, the algorithmic hierarchy subsystem comprising a first deadline authority layer, a second deadline authority layer, and a third deadline authority layer, wherein:

the first limitation removal examination layer is configured to determine whether the target account is in an account white list or an account black list based on the limitation removal data, refuse to perform limitation removal processing on the target account if the target account is determined to be in the account black list, and acquire and execute a first limitation removal rule corresponding to the account white list to perform limitation removal processing on the target account if the target account is determined to be in the account white list;

the second limitation-removal examination layer is configured to, if it cannot be determined that the target account is in an account white list or an account black list based on the limitation-removal data, process different types of data included in the limitation-removal data respectively through preset data processing rules to obtain corresponding limitation-removal auxiliary information;

the third limitation-clearance layer is configured to perform limitation clearance processing on the target account by:

obtaining historical limitation data with the similarity between the obtained historical limitation data and the limitation data being larger than a first preset similarity threshold from a limitation database, and carrying out limitation solving processing on the target account based on the obtained historical limitation data;

14. The system of claim 11, the application subsystem configured to, when a designated transaction is conducted on the target account and a preset risk of the designated transaction is detected, reject the designated transaction and perform a limited right processing on the target account, receive an account limitation removing request for the target account; or when a preset risk detection period is reached, carrying out risk detection on transaction data of different users in the risk detection period, if the target account is determined to be an account with a preset risk, carrying out right limiting processing on the target account, and receiving an account limitation removing request aiming at the target account.

15. A data processing apparatus, the apparatus comprising:

the limitation removal request module receives an account limitation removal request aiming at the target account;

the first limitation removal module is used for obtaining limitation removal data used for removing the limitation of the target account, determining whether the target account is in an account white list or an account blacklist based on the limitation removal data, if the target account is determined to be in the account blacklist, refusing to remove the limitation of the target account, and if the target account is determined to be in the account white list, obtaining and executing a first limitation removal rule corresponding to the account white list so as to remove the limitation of the target account;

and the second limitation removing module is used for processing different types of data contained in the limitation removing data respectively through preset data processing rules to obtain corresponding limitation removing auxiliary information if the target account cannot be determined to be in an account white list or an account black list based on the limitation removing data, wherein the limitation removing auxiliary information is used for triggering a limitation removing manager to remove the limitation of the target account based on the limitation removing auxiliary information.

16. A data processing apparatus, the data processing apparatus comprising:

a processor; and

a memory arranged to store computer executable instructions that, when executed, cause the processor to:

receiving an account restriction request for a target account;

17. A storage medium for storing computer-executable instructions, which when executed by a processor implement the following:

receiving an account restriction request for a target account;