Data processing

Abstract

A data processing method, comprising: on the basis of task information of a financial crime analysis task, determining M target data parties participating in the financial crime analysis task, wherein M is a positive integer, and the M target data parties each store financial feature data representing financial activities of a target entity; and coordinating the M target data parties to perform secure multi-party computation on the basis of the respective stored financial feature data and the task information, so as to obtain a financial crime analysis result indicating whether the financial activities of the target entity have a financial crime risk.

Description

Data processing Cross-reference of related applications

[0001] This application claims priority to Chinese Patent Application No. 202411648627.8, filed on November 18, 2024, the entire contents of which are incorporated herein by reference. Technical Field

[0002] This application relates to data processing technology. Background Technology

[0003] With the development of electronic technology, multi-party data processing has become increasingly common.

[0004] For example, in financial scenarios, financial institutions' understanding of users is primarily based on user behavior data collected by the institutions themselves. If there is limited user behavior data for some users, the financial institution's understanding of these users may be inaccurate, leading to potential risks. Therefore, multi-party data processing and data security during this process are crucial in financial scenarios. Invention Summary

[0005] This application provides a data processing method, comprising: determining M target data parties participating in the financial black market analysis task based on task information of the financial black market analysis task, wherein M is a positive integer, and each of the M target data parties stores financial characteristic data representing the financial activities of the target object; and coordinating the M target data parties to perform multi-party security calculations based on their respective stored financial characteristic data and the task information to obtain a financial black market analysis result indicating whether the financial activities of the target object are at risk of financial black market activity.

[0006] This application embodiment also provides a data processing apparatus, comprising: a determining module, configured to determine M target data parties participating in the financial black market analysis task based on task information of the financial black market analysis task, wherein M is a positive integer, and each of the M target data parties stores financial characteristic data representing the financial activities of the target object; and a coordination module, configured to coordinate the M target data parties to perform multi-party security calculations based on their respective stored financial characteristic data and the task information to obtain a financial black market analysis result indicating whether the financial activities of the target object are at risk of financial black market activity.

[0007] This application also provides a computer device including a processor and a memory. The memory stores instructions that can be executed by the processor to implement the data processing method described above.

[0008] This application also provides a computer program product, which includes computer instructions. The computer instructions can be executed by a processor to implement the data processing method described above.

[0009] This application also provides a non-transitory computer-readable storage medium. The computer-readable storage medium stores instructions that can be executed by a processor to implement the data processing method described above. Attached Figure Description

[0010] Figure 1 schematically illustrates the environment in which the data processing method according to an embodiment of this application can be applied.

[0011] Figure 2 is a schematic flowchart of a data processing method applied to a server according to an embodiment of this application.

[0012] Figure 3 schematically illustrates an example scenario in which the data processing method according to an embodiment of this application can be applied.

[0013] Figure 4 is a flowchart illustrating the task configuration operation performed by the server according to an embodiment of this application.

[0014] Figure 5 schematically illustrates an example task configuration interface according to an embodiment of this application.

[0015] Figure 6 is a flowchart of an example of a data processing method according to an embodiment of this application.

[0016] Figure 7 is a flowchart of another example of a data processing method according to an embodiment of this application.

[0017] Figure 8 schematically illustrates part of the operation of the data processing platform performing the data processing method according to an embodiment of this application.

[0018] Figure 9 schematically illustrates an exemplary process of a data processing method according to an embodiment of this application.

[0019] Figure 10 schematically illustrates an example scenario in which the data processing method according to an embodiment of this application can be applied.

[0020] Figure 11 is a schematic diagram of the structure of a data processing apparatus according to an embodiment of this application.

[0021] Figure 12 is a schematic diagram of the structure of a computer device according to an embodiment of this application. Detailed Implementation

[0022] Some embodiments of this application will now be described in detail with reference to the accompanying drawings. The described embodiments are for illustrative purposes only and are not intended to limit the scope of this application.

[0023] In the description of the embodiments of this application, the terms "first," "second," etc., are used only for distinguishing descriptions and should not be construed as indicating or implying relative importance. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise expressly defined.

[0024] Financial black market refers to industries related to the financial sector that use illegal means to profit. It typically includes illegal rights protection agencies, anti-collection schemes, organized debt evasion, malicious complaints, credit repair, illegal insurance cancellation, loan intermediary fraud, account theft, telecommunications fraud, setting up phishing websites, implanting Trojan viruses, and hacker extortion—all illegal activities conducted online. Financial black market not only infringes on consumer privacy and property security but also damages the financial ecosystem, endangers financial security, and even disrupts financial market order and social stability. Furthermore, financial black market can increase the operating costs of various institutions within the financial industry. For example, emerging financial institutions such as consumer finance and internet finance companies, limited by historical data accumulation and user scale, often only confirm the identity of suspected malicious users after losses have occurred. Financial institutions need to coordinate multiple departments to handle suspected malicious users, leading to significant costs. Even after confirming malicious users, activities such as evidence collection, reporting to the police, and litigation also incur substantial costs.

[0025] To prevent and combat financial illicit activities, it is necessary to analyze these activities (hereinafter referred to as "illicit activities"). The goal of financial illicit activity analysis is to identify them in order to prevent harm to the financial system and users. Through this analysis, we can better understand their operational mechanisms and dangers, thereby enabling us to take effective preventative and countermeasure measures to ensure the security and stability of the financial system. Specifically, financial illicit activity analysis can include the following aspects.

[0026] 1. Identify attack methods of the black market: The main attack methods of the financial black market include deepfakes, money laundering, mobile attacks, marketing campaign fraud, and large-scale model attacks; analyzing the evolution trend of these attack methods and identifying their technical characteristics and implementation methods are important foundations for prevention measures.

[0027] 2. Monitoring Data Breaches: The financial industry involves a large amount of high-value user data. If attacked by black market actors, it can lead to data breaches. Monitoring the black market activities on the internet, revealing the inside story of the data transaction chain, and discovering the source and path of data breaches are key to preventing data from being used by criminals.

[0028] 3. Risk Assessment: Conduct risk assessments on the financial system to identify potential threats from malicious actors, including but not limited to risks such as user data breaches, transaction fraud, and system security risks. By utilizing risk assessment models to promptly identify and address these risks, the harm caused by malicious actors to the financial system can be effectively prevented.

[0029] In related technologies, a data processing platform is set up for data processing, which has the function of managing financial black market data. This data processing platform needs to communicate with electronic devices (data providers) of financial institutions that can provide data related to financial activities in order to integrate data from multiple data providers and obtain financial black market analysis results. Specifically, the data processing platform needs to obtain information such as the marking of black market activities by various financial institutions, the mobile phone numbers associated with black market activities as compiled by various financial institutions, and the recent activity time of black market activities. Typically, the data processing platform obtains this information by contacting data providers offline, which compromises data security and reduces data processing efficiency.

[0030] In view of this, embodiments of this application provide a data processing method based on privacy computing, which can securely exchange black market data without exposing user privacy data, while ensuring the accuracy and efficiency of black market analysis.

[0031] The data processing method according to the embodiments of this application can be executed by a terminal device or a server. For example, the data processing method can be implemented through a cloud interaction system that includes server and client devices.

[0032] Figure 1 schematically illustrates an environment in which the data processing method according to an embodiment of this application can be applied. This data processing method can be executed by an electronic device. For example, the electronic device can be a server 110 as shown in Figure 1, which can communicate with terminal devices 120 corresponding to multiple data parties via a network. The network can provide various types of communication links, such as wired communication links, wireless communication links, etc. Optionally, the electronic device can also be a smartphone, laptop, etc.

[0033] It should be understood that the server 110, network, and terminal device 120 in Figure 1 are merely exemplary. Any number of servers, networks, and terminal devices can be configured according to actual needs. For example, server 110 can be a physical server or a server cluster consisting of multiple servers, and terminal device 120 can be a mobile phone, tablet, desktop computer, laptop, etc. Furthermore, multiple terminal devices 120 can simultaneously access server 110.

[0034] As shown in Figure 2, the data processing method according to an embodiment of this application may include steps 201 to 202. In step 201, based on the task information of the financial black market analysis task, M target data parties participating in the financial black market analysis task are determined, where M is a positive integer, and each of the M target data parties stores financial characteristic data representing the financial activities of the target object. In step 202, the M target data parties collaborate to perform secure multi-party computation (MPC) based on their respective stored financial characteristic data and the task information to obtain a financial black market analysis result indicating whether the financial activities of the target object pose a risk of financial black market activity. MPC refers to multiple participating parties jointly computing a function based on a security protocol, where each participating party does not expose its original data to any other participating party.

[0035] Taking the application of this data processing method to a server as an example, the server can be equipped with a data processing platform. This platform can communicate with M target data parties to collaboratively perform financial black market analysis tasks and obtain the analysis results. When the server needs to collaborate with multiple parties for secure computation, M is greater than or equal to 2.

[0036] Financial black market analysis tasks can be initiated by the data processing platform itself. For example, the data processing platform may be set to periodically execute financial black market analysis tasks, which will be automatically triggered at the designated time. Financial black market analysis tasks can also be initiated by the data provider requesting the analysis. For example, data provider A, which has connected to the data processing platform, may detect frequent financial activities of a target entity. To analyze whether the target entity's financial activities pose a risk of financial black market activities, data provider A can send a request to the data processing platform to execute the financial black market analysis task. The target entity can be a single user or a financial group composed of multiple users.

[0037] In some embodiments, task information for financial black market analysis tasks can be pre-set. This task information may include a task identifier, task type, data type, and the identifier of the data party associated with the task. The server can send task invitations to each data party based on the task information, receive response information from each data party to the task invitations, and determine the target data parties participating in the financial black market analysis task based on the response information, and set the target task parameters and / or result presentation parameters.

[0038] For example, the server sends task invitations to each data party in the communication connection. Each data party determines whether to participate in the financial black market analysis task based on the information in the task invitation. If a data party decides to participate, it sends a response indicating agreement; if it decides not to participate, it sends a response indicating refusal. The server can designate data parties that have indicated agreement as target data parties and set target task parameters and / or result presentation parameters for these target data parties. This ensures task customization while mitigating the risk of data leakage during data display.

[0039] In some embodiments, when a data party participating in a financial black market analysis task sends a response indicating agreement to the server in response to the task invitation, it may also send the identifier of a shareable field in its stored financial feature data to the server. The server can set target task parameters for the target data party based on the task information and the identifier of the shareable field returned by the target data party. These target task parameters may indicate specific algorithms for the shareable fields, etc. The server can send the target task parameters set separately for each target data party to each target data party to coordinate multi-party secure computation, thereby completing the financial black market analysis task and obtaining the analysis results.

[0040] In some embodiments, the shareable field can be set according to the specific circumstances of the target data provider. For example, if the shareable field is set to a time field, the financial characteristic data corresponding to the shareable field is the time information of the target object's financial activities.

[0041] For example, to facilitate the querying of financial characteristic data, the data provider can locally store the financial characteristic data in a data table, where the category name of the financial characteristic data serves as the field name. During the execution of financial black market analysis tasks, the target data provider can access the data in the shareable fields of the stored financial characteristic data for multi-party secure computation.

[0042] In some embodiments, the financial characteristic data stored by the data provider may include at least one of the following: first characteristic data characterizing the behavior of the financial activities of the target object, second characteristic data characterizing the time / frequency of the financial activities of the target object, and third characteristic data characterizing the identity associated with the financial activities of the target object.

[0043] Specifically, the first feature data may include at least one of the following tags: agent intermediary mobile phone number tag, multi-device association tag, abnormal mobile phone number tag, complained mobile phone number tag, multiple voiceprint tag, non-personal voiceprint tag, black voiceprint tag, suspected black voiceprint tag, highly similar address tag, highly similar template tag, identity verification failure tag, and malicious complaint tag. The second feature data may include at least one of the time the target object conducts financial activities, the number of times the target object conducts financial activities, and the frequency of the target object's financial activities. The third feature data may include contact information associated with the target object's financial activities.

[0044] To make it easier to understand, as an example, the above labels in the first feature data are explained below.

[0045] The "Agent's Mobile Number" tag indicates that a pre-stored list of agent mobile numbers contains an agent mobile number that matches the target's mobile number. Therefore, the target with the agent mobile number may be a financial agent.

[0046] The multi-device association tag indicates that a target object is associated with multiple devices. Typically, one user operates one device. Therefore, if a user is associated with multiple devices, that user may pose a certain risk.

[0047] The abnormal phone number label indicates that the phone number itself is abnormal, the login address of the phone number is abnormal, or the login device of the phone number is abnormal. Therefore, the identity of the user corresponding to the phone number is abnormal, or the activity of the user corresponding to the phone number is abnormal.

[0048] The "complained phone number" tag indicates that there is a phone number in the pre-stored list of complained phone numbers that is the same as the target's phone number. Therefore, the target's identity with the complained phone number may be a risky user.

[0049] Multiple voiceprint tags indicate that a mobile phone number corresponds to multiple saved voiceprints. Therefore, the mobile phone number is used by multiple users, and the identity of the target with the mobile phone number may be a high-risk user group.

[0050] The "not the user's own voiceprint" tag indicates that the collected user's voiceprint does not match the pre-stored user's voiceprint, therefore the user's account may be at risk of being stolen.

[0051] A black voiceprint tag indicates that the collected user's voiceprint matches the voiceprint in the pre-stored black voiceprint list, therefore the user is a historically identified black market user.

[0052] The "suspected black voiceprint" tag indicates that the collected user's voiceprint has a high similarity to the voiceprints in the pre-stored black voiceprint list, so the user may be a black market user identified in the past.

[0053] The "highly similar address" tag indicates that the address where a user conducts financial activities is highly similar to an address in a pre-stored list of black market addresses, suggesting that the user may be from a black market region.

[0054] The "template highly similar" tag indicates that the user's financial activity information is highly similar to the information in the preset template, therefore the user may be at risk of being involved in illegal activities.

[0055] The "Identity Verification Failure" label indicates that the user's identity verification failed when logging in for financial activities or performing specific actions. Therefore, the user's identity may not match the pre-registered identity information, and the user poses a certain risk.

[0056] The "malicious complaint" label indicates that a user has a history of malicious complaint behavior, and therefore, the user's behavior may pose a certain risk.

[0057] In some embodiments, the target task parameters include instructions regarding multi-party secure computation, such as instructions for union processing, intersection processing, comparison and summation processing, etc. For example, union processing is suitable for first feature data characterizing the financial activities of the target object. Intersection processing is suitable for second feature data characterizing the time / frequency of the financial activities of the target object. Comparison and summation processing is suitable for third feature data characterizing the identity associated with the financial activities of the target object.

[0058] In some embodiments, the server can also determine the type of the target data party and set target task parameters for the target data party based on the type of the target data party and task information. For example, if target data party A is of the test type, the target task parameters can be set to test task parameters; if target data party B is of the important type, the target task parameters can be set to core task parameters.

[0059] By utilizing the data processing method according to the above embodiments, multi-party security calculations are performed by each target data party in collaboration based on their respective stored financial feature data and task information of the financial black market analysis task, thereby executing financial black market analysis. This makes the financial black market analysis results more accurate and ensures the security of data during the multi-party data processing.

[0060] As shown in Figure 3, the data processing platform and an unlimited number of data providers can connect via the internet. These data providers typically include various types of financial institutions such as banks, consumer finance companies, micro-loan companies, loan assistance companies, and fintech companies. In this business scenario, the data processing platform can configure task parameters for financial black market analysis tasks and distribute them to each data provider. Each data provider can then perform multi-party collaborative computations based on the received task parameters to obtain the financial black market analysis results, which are then presented on their respective management pages. It should be understood that both the data processing platform and each data provider can perform secure multi-party computations.

[0061] In some embodiments, the server can determine the first task parameters based on the task identifier and task type of the financial black market analysis task. For example, the server may pre-store a task parameter mapping table corresponding to different task identifiers and task types, allowing the server to obtain the first task parameters corresponding to the task identifier and task type of the financial black market analysis task by querying the task parameter mapping table. Then, the server can obtain the identifier of the shareable field from the target data source and add the identifier of the shareable field to the first task parameters to obtain the target task parameters. This allows for flexible setting of the target task parameters.

[0062] In some embodiments, based on the target task parameters received from the server, each target data party can participate in multi-party secure computation based on the financial characteristic data corresponding to their respective shareable fields to obtain their own computation results, and then send these results to the server. The server can perform statistical analysis based on the computation results fed back by each target data party to obtain the final financial black market analysis results.

[0063] In some embodiments, the server may first parse the target task parameters to determine the feature data processing instructions for the target data party. The feature data processing instructions may include at least one of data access instructions, data calculation instructions, and data transmission instructions. The server may send the feature data processing instructions to the target data party, enabling the target data party to participate in multi-party secure computation according to the instructions to obtain computation results, and then feed the computation results back to the server. The server can determine the financial black market analysis results based on the computation results fed back by each target data party. For example, the server may summarize, compare, or perform other operations on the computation results according to the target task parameters to obtain the financial black market analysis results.

[0064] In some embodiments, the server may perform task configuration operations in advance. As shown in FIG4, the task configuration operation may include steps 301 to 303.

[0065] In step 301, in response to a request for task configuration, a task configuration page is displayed. This page allows users to set the task identifier, task type, data types to be analyzed, and relevant institutions for the financial black market analysis task. Any one or more data parties with task requirements can request task configuration from the server to collaborate with other data parties in executing the financial black market analysis task.

[0066] Figure 5 shows an example of a task configuration page. In Figure 5, the task type is set to "Black Market Scan"; the task name is set to "Behavioral Feature Analysis"; the analysis data type is set to "Suspected Black Market Activities"; the minimum number of organizations hit by black market activities is set to "2", where the minimum number of organizations hit by black market activities refers to the minimum number of organizations containing the target object. For example, a minimum number of organizations hit by black market activities of 2 means that at least two organizations contain the target object; Organizations A, B, and C are selected as the organizations to be executed in the task.

[0067] In this embodiment, different financial black market analysis tasks can be set according to different situations. For example, users can use the task configuration page to set multiple data types to be analyzed, so that the data provider can determine the shareable fields to participate in the financial black market analysis task according to its own data sensitivity.

[0068] In step 302, the task information of the financial black market analysis task to be executed, configured based on the task configuration page, and the information of the candidate data parties associated with the financial black market analysis task are obtained, and the task information of the financial black market analysis task is sent to the candidate data parties so that the candidate data parties can respond.

[0069] In step 303, the response information sent by the candidate data party to the financial black market analysis task is received, and the candidate data party that agrees to participate in the financial black market analysis task is set as the target data party.

[0070] For example, a server-side data processing platform can send invitations to various institutions to participate in a financial black market analysis task. Each institution can respond to the invitation by providing feedback to the data processing platform, indicating whether they agree or refuse to participate in the calculation, as well as their data details. As an example, if the data processing platform receives responses from institutions A, B, and C indicating agreement to participate, while institution D indicates refusal, then the data processing platform does not need to send the calculation task parameters and result presentation parameters to institution D's calculation engine. Furthermore, institutions A and B provide the shareable fields required for the calculation, including behaviorTime, phoneNumberAmt, and features, while institution C provides null, meaning there are no shareable fields. After receiving the responses from institutions A, B, and C, the data processing platform can configure the shareable fields required for the calculation to be behaviorTime, phoneNumberAmt, and features.

[0071] As another example, the data processing platform invites 10 organizations. Some organizations' response messages indicate that they can provide three shareable fields, while the response messages of other organizations indicate that they can provide at most one shareable field. In this case, the data processing platform can configure the number of shareable fields required for computation to be either three or one.

[0072] In some embodiments, result presentation parameters can indicate whether the computation results are visible to each institution. For example, result presentation parameters can be set such that intersection results are visible to all institutions, while computation results for shareable fields are only visible to institutions that provided valid shareable fields.

[0073] In some embodiments, the data processing platform can directly generate calculation task parameters and result presentation parameters based on the administrator's configuration and distribute them to various institutions. In response, each institution can send feedback information to the data processing platform indicating whether to participate in the calculation and the information on available fields, and then immediately begin executing the financial black market analysis task.

[0074] In some embodiments, the M target data parties include a first target data party and a second target data party. The first target data party stores financial characteristic data corresponding to the target object with L fields, and the second target data party stores financial characteristic data corresponding to the target object with N fields, where L and N are both positive integers, and L may be equal to N or not equal to N. When L is not equal to N, the data processing platform can set corresponding target task parameters according to the financial characteristic data of each target data party, thereby achieving flexibility in multi-party data processing.

[0075] In some embodiments, the data processing platform can determine whether the M target data parties all have the same set of shareable fields. If the M target data parties all have the same set of shareable fields, the data processing platform can determine the same target task parameters for the M target data parties at once. If the M target data parties do not all have the same set of shareable fields, the data processing platform can determine the target task parameters for each target data party based on the task information and the shareable fields of that target data party. In this way, each target data party can obtain its own target task parameters to perform its own data processing, thereby achieving flexibility in multi-party data processing.

[0076] In some embodiments, before performing a financial black market analysis task, the server can pre-determine the target object to be analyzed. The target object can be a single target user in a financial scenario, or a group of target users.

[0077] In some embodiments, the server can parse the target task parameters to obtain shareable fields and algorithm instructions from the target task parameters. The shareable fields include at least one of a time field, a communication field, and a feature field. The algorithm instructions include at least one of instructions for intersection processing, comparison and summation processing, and union processing.

[0078] If the shareable field is a time field, the server can determine that the financial black market analysis task is an active time analysis task and generate a feature data processing instruction that indicates the intersection processing, so that the target data party can execute the feature data processing instruction to obtain the target user's active time information.

[0079] If the shareable field is a communication field, the server can determine that the financial black market analysis task is a communication method analysis task and generate characteristic data processing instructions for comparison and summation. This allows the target data provider to execute the instructions and obtain the target user's contact information across different data providers. This reduces the number of virtual users and lowers user risk.

[0080] If the shareable field is a feature field, the server can determine that the financial black market analysis task is a behavioral feature analysis task and generate a feature data processing instruction that instructs the target data provider to perform a union operation, thereby obtaining the behavioral characteristics of the target user on different data providers. This allows the server to determine whether a user meets the conditions for a risk behavior label, achieving user behavior security analysis.

[0081] As shown in Figure 6, an example of a data processing method according to an embodiment of this application includes the following steps: sending a computation task invitation, which includes data sharing computation parameters, according to the type of each participant; receiving responses from each participant, setting computation task parameters and result presentation parameters; running a multi-party threshold intersection algorithm (this algorithm is a privacy computation algorithm that allows all parties involved in the intersection to find the intersection that meets the preset threshold conditions without exposing information about the non-intersection parts); calculating the shareable fields one by one according to the data sharing computation parameters for the intersection in the intersection result, wherein different algorithms can be applied to different shareable fields; running a multi-party extremum function to obtain the maximum value of the behavior time of the records of the parties associated with the intersection; running a multi-party counting function to calculate the number of associated mobile phone numbers; running a multi-party union function to calculate the set of malicious behavior features and the number of features in the entire industry; the task initiator calculating the content to be presented to each participant according to the result presentation parameters and returning the data to the corresponding participant.

[0082] For example, the participating parties in the response can run a multi-party threshold intersection algorithm based on user identification information to determine the common user among the participating parties. If participant A contains user identification information of Zhang xx, Wang xx, Li xx, and Zhao xx, participant B contains user identification information of Zhang xx, Wu xx, and Sun xx, and participant C contains user identification information of Zhang xx and Zhou xx, then by running the multi-party threshold intersection algorithm, the intersection result of participants A, B, and C can be determined to be Zhang xx.

[0083] In some embodiments, the calculation result is a time comparison result, and the financial black market analysis result includes target time information. In this case, the server can receive the time comparison result from each object data party in the target data parties. The time comparison result is obtained by finding the intersection of the behavior time corresponding to the time field in the financial characteristic data of the object data party and the behavior time corresponding to the time field in the financial characteristic data of the reference data party. The reference data party is the data party other than the object data party in the target data parties. Then, the server can statistically analyze the time comparison results reported by each object data party to obtain the target time information. The target time information can indicate the user's most recent active time, and thus can be used for risk analysis.

[0084] Continuing with the example of three parties, A, B, and C, targeting the black market operator Zhang xx, if organization A marks the time of Zhang xx's malicious behavior as 2024-01-15 14:09:08, organization B marks it as 2023-12-15 14:09:08, and organization C marks it as 2024-01-12 10:09:08, then by running a multi-party extremum function, the maximum value among the three times is 2024-01-15 14:09:08, indicating that the most recent malicious behavior by black market operator Zhang xx occurred on 2024-01-15 14:09:08. This indicator helps organizations determine whether the black market operator is active.

[0085] In some embodiments, the calculation result is a communication comparison result, and the financial black market analysis result includes the number of communication identifiers. In this case, the server can receive the communication comparison result fed back by each object data party in the target data party. The communication comparison result is obtained by comparing the communication identifiers corresponding to the communication fields in the object data party with the communication identifiers corresponding to the communication fields in the reference data party. The reference data party is the data party in the target data party other than the object data party. Then, the server can sum the communication comparison results fed back by each object data party to obtain the number of communication identifiers. The number of communication identifiers can indicate the number of communication methods associated with the user, and thus can be used for risk analysis.

[0086] For example, referring to Figure 6, after running the multi-party threshold intersection algorithm, for the intersection in the intersection result, if the intersection data is a communication identifier, then the multi-party counting function is run to calculate the number of associated phone numbers. Continuing with the example of parties A, B, and C targeting Zhang xx in the black market, if organization A marks Zhang xx with phone numbers p1 and p2, organization B marks Zhang xx with phone numbers p1, p2, p4, and p5, and organization C marks Zhang xx with phone number p3, then a multi-party comparison can be performed first. When comparing A and B, it can be found that all the phone numbers recorded by A are included in the phone number set recorded by B, thus determining that the number of valid phone numbers recorded by A is 0, and the number of valid phone numbers recorded by B is 4. When comparing B and C, it can be determined that the number of valid phone numbers recorded by C is 1, and the number of valid phone numbers recorded by B is 4. Then, the sum can be obtained, yielding the number of valid phone numbers for Zhang xx as 0 + 4 + 1 = 5. This indicator reflects the likelihood that a user is suspected of involvement in black market activities. Generally, users with a larger number of phone numbers are more likely to be involved in black market activities.

[0087] In some embodiments, the calculation result is a feature comparison result, and the financial black market analysis result includes the number of feature tags. In this case, the server can receive the feature comparison result fed back by each object data party in the target data party. The feature comparison result is obtained by taking the union of the feature tags corresponding to the feature fields in the object data party and the feature tags corresponding to the feature fields in the reference data party. The reference data party is the data party in the target data party other than the object data party. Then, the server can statistically analyze the feature comparison results fed back by each object data party to obtain the number of feature tags. The number of feature tags can indicate the number of behavioral tags matched by the user, and can therefore be used for risk analysis.

[0088] For example, in some embodiments, the server can pre-set the following 12 feature tags: A1, indicating the agent's intermediary mobile phone number; A2, indicating multi-device association; A3, indicating an abnormal mobile phone number; A4, indicating the mobile phone number being complained about; B1, indicating multiple voiceprints; B2, indicating a voiceprint not belonging to the individual; B3, indicating a black voiceprint; B4, indicating a suspected black voiceprint; C1, indicating highly similar addresses; C2, indicating highly similar templates; D1, indicating identity verification failure; and D2, indicating malicious complaints. Each organization can record consumer behavior feature tags. Referring to Figure 6, after running the multi-threshold intersection algorithm, for the intersection in the intersection result, if the intersection data is user behavior features, then the multi-union function is run to calculate the industry's full set of malicious behavior features and the number of features. Continuing with the example of parties A, B, and C targeting Zhang XX in the black market, organization A records Zhang XX's behavioral characteristic tags as A1|C2|D1, organization B records them as A1|A2|C1|D1, and organization C records them as D2. By running a multi-party union function, we can obtain Zhang XX's complete set of malicious behavior characteristics for the entire industry as {A1|A2|C1|C2|D1|D2}. The total number of Zhang XX's malicious behavior characteristics is 6, meaning Zhang XX has 6 characteristic tags. Therefore, the server can determine Zhang XX's involvement in black market activities based on the number of his characteristic tags.

[0089] It is understood that, in this embodiment of the application, the server may also perform the above calculations for other fields, such as the amount of reduction requested by the consumer, the method of compensation requested, the cumulative number of complaints to the regulator, the number of complaint channels, etc.

[0090] In some embodiments of this application, the financial feature data further includes: label weights of the target object's financial activities, whereby the label weights characterize the frequency of the behavior. Besides counting the total number of malicious behavior feature labels of the target object, counting the number of times the same label is reported also has certain business significance. The server can return the counted number of times a certain label is reported to the target data provider.

[0091] As shown in Figure 7, another example of a data processing method according to an embodiment of this application includes the following steps: sending a calculation task invitation based on the type of each participant, which includes data sharing calculation parameters; receiving responses from each participant, setting calculation task parameters and result presentation parameters; running a multi-party threshold intersection algorithm; calculating the shareable fields one by one according to the data sharing calculation parameters for the intersection in the intersection result, wherein different algorithms can be applied to different shareable fields; running a multi-party extremum function to obtain the maximum value of the behavior time of the records of each party associated with the intersection; running a multi-party counting function to calculate the number of associated mobile phone numbers; running a multi-party union function to calculate the set of malicious behavior features and the number of features in the entire industry; the task initiator calculating the content to be presented to each participant according to the result presentation parameters and returning the data to the corresponding participant; the task initiator comparing the result of this calculation task with the result of the previous calculation task and cleaning up invalid tags and records; each participant receiving data synchronized by the operation center and comparing it with the result of the previous task stored locally, deleting invalid tags and invalid black market records.

[0092] The difference between the data processing method example shown in Figure 7 and the data processing method example shown in Figure 6 is that, after the task initiator calculates the content to be presented to each participant based on the result presentation parameters and returns the data to the corresponding participants, the task initiator compares the results of the current calculation task with the results of the previous calculation task, and cleans up invalid tags and records. For example, if the historical calculation task results indicate that the target user Zhang xx has 5 malicious behavior characteristics, meeting the risk user criteria, while the current calculation task results indicate that the target user Zhang xx has only 1 malicious behavior characteristic, not meeting the risk user criteria, then, based on the results of the current calculation task, invalid tags can be deleted and the risk user list updated to remove Zhang xx from the risk user list. Furthermore, the results of the current calculation task are synchronized to other participants, allowing each participant to compare the received synchronized data with the previously saved task results locally, and delete invalid tags and invalid black market records.

[0093] As mentioned above, various organizations can record data such as the target's phone number, the time of malicious activity, and malicious tags, and perform MPC (Multi-Channel Computation) to participate in financial black market analysis tasks. In some cases, operational errors may lead to incorrect labeling, resulting in a target being incorrectly tagged, which may necessitate the removal of the incorrect label.

[0094] In some embodiments, for the target object, if the number of feature tags contained in its financial black market analysis results is greater than a predetermined threshold, it is determined that it has the risk of black market activities; if the number of feature tags contained in its financial black market analysis results is less than or equal to the predetermined threshold, it is determined that it does not have the risk of black market activities, and the server does not share all the tags of the target object with each data party.

[0095] In some embodiments, the server may update the financial black market analysis results of the target object after receiving an update event of financial characteristic data of the target object from at least one target data party, so as to ensure the real-time nature of the data.

[0096] In some embodiments, after obtaining the analysis results of financial black market activities, the data processing platform can send these results to the data provider, enabling the data provider to understand the analysis results. For example, the data processing platform can send the analysis results to a first target data provider with the authority to obtain them. The data provider's authority can be determined based on the financial characteristic data provided by the data provider. For example, if the data provider provides more than a predetermined amount of financial characteristic data, then the data provider has the authority to obtain the analysis results of financial black market activities.

[0097] In some embodiments, the server can analyze the results of financial black market analysis, determine the target financial characteristic data corresponding to the analysis results, and send the analysis results to the target data provider that provided the target financial characteristic data. This allows the financial black market analysis results to be displayed to the corresponding data provider, preventing data leakage.

[0098] In some embodiments, the server locally stores the financial characteristic data of the target object. The accuracy of the financial black market analysis result can be improved by performing multi-party secure computation based on the financial characteristic data stored by each of the M target data parties, the financial characteristic data stored locally on the server, and the task information. For example, each of the M target data parties stores financial characteristic data with X shareable fields, while the financial characteristic data stored locally on the server has Y shareable fields, where X and Y are both positive integers. These Y shareable fields can be the same as or different from these X shareable fields.

[0099] In some embodiments, the server can set result presentation parameters for the target data party based on its type to prevent data leakage. Result presentation parameters can indicate whether a participating institution is legally entitled to the computation results or only has access to the number of intersections. Result presentation parameters can also be understood as characterizing whether the shared computation results of a shareable field are visible to a particular data party. Result presentation parameters can be set to control whether certain shared computation results are invisible to a particular data party.

[0100] Continuing with the example of parties A, B, and C targeting Zhang XX in the black market, if organization C is one of the participating institutions in the test, the data processing platform can set result presentation parameters for organization C. This allows the platform to only notify organization C that Zhang XX has financial activities with two other institutions simultaneously, without displaying the number of phone numbers, behavioral characteristic sets, etc., to organization C. This avoids the over-dissemination of calculation results.

[0101] In some embodiments, each data provider can trigger task configuration operations according to its own needs to obtain the required financial black market analysis results.

[0102] For example, the business logic module of the data processing platform in the server can obtain information on whether organizations A, B, and C participate in the calculation of shareable fields, and generate calculation target task parameters for A, B, and C, as well as calculation target task parameters for the data processing platform itself. The business logic device of the data processing platform sends a request to the calculation engine module, and then the calculation engine module of the data processing platform sends the target task parameters to the calculation engine modules of organizations A, B, and C.

[0103] In some embodiments of this application, the target task parameters may also be referred to as computation task configuration parameters, data sharing computation parameters, task parameters, computation task parameters, task configuration parameters, task execution parameters, task management parameters, task logic parameters, etc.

[0104] In some embodiments of this application, the calculation of shareable fields can also be referred to as shared tag calculation or shared calculation. A data party's participation in the calculation of shareable fields can be understood as follows: during the MPC process, the data party can provide the value of a specific shareable field to other data parties, enabling the task initiator to determine whether the target object poses a risk of financial malpractice based on the value of the shareable field. In the embodiments of this application, the value of the shareable field provided by one data party to other data parties is encrypted, which can prevent other data parties from knowing the plaintext of the shareable field value and reduce the security risk caused by the leakage of the shareable field value.

[0105] Taking A and B as examples where shared computing is supported, while C does not, the business logic module of the data processing platform can parse the task information of the financial black market analysis task to obtain the calculation task parameters for the financial black market analysis. Then, it can send the calculation task parameters to the calculation engine module via the following command (JSON code):

[0106] In the above instructions, the `algtype` field indicates the algorithm to be executed. The value of this field is determined by the type of financial black market analysis task. For example, `blackSampleMatch` indicates that the algorithm being executed is a black market clue collision algorithm. The `parties` field indicates the participating institutions and whether the institutions support the calculation of financial feature data corresponding to the shareable fields. 10001-10003 are the identifiers of institutions A and C in the system, respectively. For example, 10001 corresponds to 1, indicating that institution A supports shared computation; 10003 corresponds to 0, indicating that institution C does not support shared computation. `sharedFields` indicates the shareable fields involved in the computation and the algorithm for each shareable field. For example, `behaviorTime` represents the time information of the behavior, and `latest` represents the most recent behavior time.

[0107] For example, the fields involved in the calculation can be configured through the data processing platform's page.

[0108] As shown in Figure 8, after the business engine module of the data processing platform sends the above JSON code or code containing the above information to the computing engine module of the data processing platform, the computing engine module can parse the instruction, and based on the instruction, confirm that it needs to coordinate with the three organizations A, B, and C to carry out intersection and shared field calculations, and thus generate and issue task instructions to the three organizations A, B, and C.

[0109] For example, after receiving the target task parameters, the computing engine module generates its own computing instructions and sends computing instructions to A, B, and C as follows.

[0110] The computation instructions for the computation engine module of the data processing platform may include:

[0111] According to the above instructions, the dataset aifasset-1 used by the computing engine module of the data processing platform only needs to contain the field used for intersection calculation, with only one row of data, to ensure that the dataset used for this privacy calculation is not empty. 10000 corresponds to 0, indicating that the data processing platform does not support shared computation, meaning that the data processing platform does not participate in the calculation of shareable fields. In the sharedField field, "behaviorTime" corresponds to "latest," indicating that the processing method (or algorithm) for the behaviorTime shareable field is latest, for example, calculating the latest malicious behavior time through intersection calculation. "phoneNumberAmt" corresponds to "sum," indicating that the processing method for the phoneNumberAmt shareable field is sum, for example, calculating the number of valid phone numbers of the target object (e.g., Zhang San) through comparison and summation. "features" corresponds to "union," indicating that the processing method for the features shareable field is union, for example, calculating the tag set through union calculation. In the result field, the value of the intersection field is 1, indicating that the data processing platform needs to participate in or perform intersection calculation. A value of "0" for behaviorTime indicates that the data processing platform does not participate in the shared calculation of the behaviorTime shareable field. Similarly, the data processing platform does not participate in the shared calculation of the phoneNumberAmt and features shareable fields.

[0112] The computation instructions sent by the computation engine module of the data processing platform to the computation engines of organizations A and B may include:

[0113] The above instructions instruct institutions A and B to participate in intersection calculations and the calculation of shareable fields. The results, including the number of intersections and calculations related to the time of the action, the total number of phone numbers, and malicious behavior characteristics, are all visible. For example, if institutions A and B both perform intersection processing, and the server performs intersection calculations on the financial feature data corresponding to institutions A and B, the server can simultaneously send the intersection results to both institutions, allowing them to view the final results.

[0114] For example, the datasets used for calculations by A and B need to include fields for intersection and shared calculations for behavior time, total number of phone numbers, and behavioral characteristics. If the dataset used does not contain the required fields, or if a field in a data record is empty during the calculation, the relevant results for that data will not be visible. This can be controlled by an algorithm. In this way, computational fairness can be achieved through the above control, preventing organizations from using datasets containing a large number of empty fields to participate in calculations and obtain the calculation results of other organizations.

[0115] The computation instructions sent from the computation engine module of the data processing platform to the computation engine of organization C may include:

[0116] The above instructions direct the computation engine of Organization C to only participate in intersection calculations and not in calculations of shareable fields. Only the intersection count is visible in the results; calculations related to shareable fields are not visible. For example, when Organization C executes the above instructions, its computation engine may ignore whether its data file contains the behaviorTime, phoneNumerAmt, and features fields.

[0117] This application also provides a data processing method. Using this method, multiple data parties can collaboratively perform MPC based on their respective financial characteristic data, so that each data party can obtain financial black market analysis results for a target object without exposing its own financial characteristic data to other data parties. As shown in Figure 9, the method may include: in step S501, multiple target data parties (e.g., computing nodes) respectively send multiple first financial black market analysis results obtained by performing MPC based on their respective financial characteristic data of the target object to the task initiator; in step S502, the task initiator obtains a second financial black market analysis result based on the received multiple first financial black market analysis results. The financial characteristic data may include at least one of the following: behavioral data related to the malicious financial activities of the target object (e.g., data corresponding to the features field), data characterizing the activity level of the target object in financial activities (e.g., data corresponding to the behaviorTime field), and data characterizing whether the identity of the target object is genuine (e.g., data corresponding to the phoneNumberAmt field).

[0118] As shown in Figure 10, taking the collaborative implementation of financial black market analysis by three computing nodes as an example, the first computing node can perform MPC based on local behavioral data (features), activity level data (behaviorTime), and real identity data (phoneNumberAmt) to obtain a first financial black market analysis result. Similarly, the second computing node can perform MPC based on local behavioral data (features), activity level data (behaviorTime), and real identity data (phoneNumberAmt) to obtain another first financial black market analysis result. The task initiator can obtain these two first financial black market analysis results and derive a second financial black market analysis result based on them.

[0119] The above embodiments primarily use the architecture of a data processing platform and data providers as examples. However, a decentralized architecture can also be adopted. For instance, each computing node is equal, with no central role. Furthermore, the data processing platform itself is also a computing node, possessing similar functions to other computing nodes (or data providers). Any data provider can act as a task initiator.

[0120] It should be understood that although each step in the flowcharts of the embodiments described above is shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but may be executed at different times. The execution order of these steps or stages is not necessarily sequential, but may be performed alternately or in turn with other steps or at least a portion of the steps or stages of other steps.

[0121] This application also provides a data processing apparatus for implementing any of the above-described data processing methods. This data processing apparatus can, for example, be integrated into a server. As shown in FIG11, the data processing apparatus may include: a determining module 401, used to determine M target data parties participating in the financial black market analysis task based on task information of the financial black market analysis task, wherein M is a positive integer, and each of the M target data parties stores financial characteristic data representing the financial activities of the target object; and a coordination module 402, used to coordinate the M target data parties to perform multi-party security calculations based on their respective stored financial characteristic data and the task information to obtain a financial black market analysis result indicating whether the financial activities of the target object carry a risk of financial black market activity.

[0122] Embodiments of this application also provide a computer device. As shown in FIG12, the computer device may include a processor, a memory, an input / output interface, a communication interface, a display unit, and an input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are connected to the system bus via the input / output interface. The processor of the computer device provides computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and computer instructions. The internal memory provides an environment for the operation of the operating system and computer instructions in the non-volatile storage medium. The input / output interface of the computer device is used for exchanging information between the processor and external devices. The communication interface of the computer device is used for wired or wireless communication with external terminals; wireless communication can be implemented through Wi-Fi, mobile cellular networks, NFC (Near Field Communication), or other technologies. The computer instructions can be executed by the processor to implement any of the above-described data processing methods. The display unit of the computer device is used to form a visually visible image. It can be a display screen, a projection device, or a virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the casing of the computer device, or external keyboards, touchpads, or mice, etc.

[0123] This application also provides a non-transitory computer-readable storage medium. This computer-readable storage medium may include: read-only memory (ROM), random access memory (RAM), a magnetic disk, or an optical disk, etc. The computer-readable storage medium stores computer instructions that can be executed by a processor to implement any of the data processing methods described above.

[0124] This application also provides a computer program product. This computer program product includes computer instructions that can be executed by a processor to implement any of the data processing methods described above.

[0125] It should be noted that the object data (including but not limited to user device information, user personal information, etc.) and dialogue data involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0126] The memory in the embodiments of this application may include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory may include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0127] The processors in the embodiments of this application may include general-purpose processors, central processing units, graphics processors, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc.

[0128] For the sake of brevity, the specific working process and beneficial effects of the data processing apparatus, computer-readable storage medium, computer program product, computer equipment and their corresponding units described above can be found in the descriptions in the embodiments of the data processing method above.

[0129] The foregoing has described some embodiments of this application, which are not intended to limit the scope of this application. Various modifications or equivalent substitutions can be made to these embodiments by those skilled in the art. Such modifications or equivalent substitutions should be included within the scope of this application.

Claims

1. A data processing method, comprising: Based on the task information of the financial black market analysis task, M target data parties are identified to participate in the task, where M is a positive integer. Each of the M target data parties stores financial characteristic data representing the financial activities of the target object; and The M target data parties collaborate to perform multi-party security calculations based on their respective stored financial feature data and task information to obtain financial black market analysis results indicating whether the financial activities of the target object pose a risk of financial black market activities.

2. The method of claim 1, wherein, The financial characteristic data includes at least one of the following: first characteristic data characterizing the financial activities of the target object, second characteristic data characterizing the time or frequency of the financial activities of the target object, and third characteristic data characterizing the identity associated with the financial activities of the target object.

3. The method as described in claim 2, wherein, The first feature data includes at least one of the following tags: agent intermediary mobile phone number tag, multi-device association tag, abnormal mobile phone number tag, complained mobile phone number tag, multiple voiceprint tag, non-personal voiceprint tag, black voiceprint tag, suspected black voiceprint tag, highly similar address tag, highly similar template tag, identity verification failure tag, and malicious complaint tag; The second feature data includes at least one of the following: the time the target object conducts financial activities, the number of times the target object conducts financial activities, and the frequency of the target object's financial activities; and / or The third feature data includes contact information associated with the target object.

4. The method of claim 2, wherein, Determining the M target data includes: Based on the task information, task invitations are sent to multiple data parties; Receive multiple response messages from the multiple data parties in response to the task invitation; and For each of the response messages, in response to determining that the response message indicates agreement, one of the multiple data parties that responded with the response message is identified as one of the M target data parties.

5. The method of claim 4, wherein, The multi-party security computation, performed in coordination with the M target data parties, includes: Based on the task information, M target task parameters are set for the M target data parties, and each target task parameter includes instructions regarding the multi-party secure computation; The M target task parameters are respectively sent to the M target data parties; and The M target data parties collaborate to perform the multi-party security calculation based on their respective stored financial feature data and the M target task parameters.

6. The method of claim 5, wherein, Setting the M target task parameters for the M target data sources respectively includes: For each of the M target data parties, based on the task information and the identifier of the shareable field in the financial feature data stored by the target data party, a task parameter is set for the target data party as one of the M target task parameters. The identifier is fed back by the target data party along with one of the multiple response information. The target task parameter indicates the algorithm for performing the multi-party secure computation on the shareable field.

7. The method of claim 6, wherein, The shareable field corresponds to the first feature data, and the target task parameter indicates the union processing.

8. The method of claim 6, wherein, The shareable field corresponds to the second feature data, and the target task parameter indicates the intersection processing.

9. The method of claim 6, wherein, The shareable field corresponds to the third feature data, and the target task parameter indicates comparison and summation processing.

10. The method of claim 5, wherein, Setting the M target task parameters for the M target data sources respectively includes: For each of the M target data parties, determine the type of the target data party, and set task parameters for the target data party as one of the M target task parameters based on the type of the target data party and the task information.

11. The method of any one of claims 1-10, wherein, The M target data parties include a first target data party and a second target data party; and The financial feature data stored by the first target data provider has L fields, and the financial feature data stored by the second target data provider has N fields, where L and N are both positive integers.

12. The method of claim 11, wherein, L is not equal to N.

13. The method according to any one of claims 1-10, further comprising: The financial black market analysis results are sent to the target data parties among the M target data parties that have the authority to obtain the financial black market analysis results.

14. The method according to any one of claims 1-10, further comprising: Determine the target financial characteristic data corresponding to the analysis results of the financial black market; as well as The financial black market analysis results are sent to the target data parties that provided the target financial characteristic data among the M target data parties.

15. The method according to any one of claims 1-10, further comprising: In response to receiving an update event for the financial feature data from at least one of the M target data sources, the financial black market analysis result is updated.

16. The method of any one of claims 1-10, wherein, The multi-party security computation, performed in coordination with the M target data parties, includes: The M target data parties collaborate to perform the multi-party secure computation based on their respective stored financial feature data, locally stored financial feature data representing the financial activities of the target object, and the task information.

17. The method of claim 16, wherein, Each of the M target data sources stores X shared fields for the financial feature data, and the locally stored financial feature data has Y shared fields, where X and Y are both positive integers.

18. The method of any one of claims 1-10, wherein, The financial characteristic data includes: label weights representing the frequency of the financial activities of the target object.

19. The method of any one of claims 1-10, wherein, The analysis results of the financial black market include the number of feature tags. The method further includes: In response to determining that the number of feature tags is greater than a predetermined threshold, it is determined that the financial activities of the target object pose a risk of financial black market activities.

20. A data processing apparatus, comprising: The determination module is used to determine M target data parties participating in the financial black market analysis task based on the task information of the task, where M is a positive integer, and each of the M target data parties stores financial characteristic data representing the financial activities of the target object; and The collaboration module is used to coordinate the M target data parties to perform multi-party security calculations based on their respective stored financial feature data and task information, so as to obtain a financial black market analysis result indicating whether the financial activities of the target object have the risk of financial black market activities.

21. A computer device comprising a processor and a memory, the memory storing instructions executable by the processor to implement the data processing method as claimed in any one of claims 1 to 19.

22. A computer program product comprising computer instructions executable by a processor to implement the data processing method as described in any one of claims 1 to 19.

23. A non-transitory computer-readable storage medium storing instructions executable by a processor to implement the data processing method as claimed in any one of claims 1 to 19.

Images