Data processing method and system based on trusted metadata

Abstract

The invention relates to a data processing method and system based on trusted metadata, and belongs to the field of data processing. In the method, under the condition that a first service of a user is triggered to be executed, a processing device determines key metadata and non-key metadata in the metadata of the first service by performing mapping analysis on the metadata of the first service; the processing equipment encrypts the key metadata, and performs integrity protection on the key metadata and the non-key metadata to obtain metadata subjected to security protection; the processing device sends the security-protected metadata to the cloud, wherein the security-protected metadata is used for the cloud to execute anti-fraud analysis on the first service; and the processing device receives an anti-fraud analysis result returned by the cloud for the first service, wherein the anti-fraud analysis result indicates whether the first service has a fraud risk.

Description

Technical Field

[0001] This invention belongs to the field of data processing and relates to data processing methods and systems based on trusted metadata. Background Technology

[0002] With the deep integration of mobile internet and digital services, real-time risk control (such as fraud prevention) has become a crucial link in ensuring the security of various online businesses. This type of risk control typically relies on real-time analysis and judgment of multi-dimensional metadata generated from user business behavior. In this process, to reduce latency and protect user privacy, some computing tasks are beginning to migrate to the user's terminal (such as a smartphone), meaning that the collected metadata is initially processed or analyzed locally on the terminal.

[0003] However, performing metadata analysis locally on the terminal faces significant challenges. Terminal devices, especially mobile terminals, have limited computing resources (such as processors and memory) and battery life. Performing high-quality analysis of complex and diverse business metadata (such as feature extraction, pattern recognition, and anomaly detection) is inherently a computationally intensive task. Executing it on the terminal inevitably consumes a large amount of computing power, potentially leading to device overheating, lag, and excessive power consumption, thus harming the user experience. More importantly, the limited computing power of the terminal often makes it difficult to deploy or run sufficiently complex analytical models, resulting in low depth, accuracy, and precision of local analysis. This makes it difficult to reliably identify potentially high-risk behaviors, thereby limiting the effectiveness of the overall risk control system. Therefore, overcoming the terminal's computing power bottleneck and achieving efficient and accurate metadata analysis locally has become a prominent technical challenge. Summary of the Invention

[0004] In view of this, in order to solve the above problems, the present invention provides a data processing method and system based on trusted metadata.

[0005] To achieve the above objectives, the present invention provides the following technical solution: Firstly, a data processing method based on trusted metadata is provided, applied to a user-side processing device. The method includes: when a user's first service is triggered, the processing device performs mapping analysis on the metadata of the first service to determine key and non-key metadata within the metadata of the first service; the processing device encrypts the key metadata and performs integrity protection on both key and non-key metadata to obtain securely protected metadata; the processing device sends the securely protected metadata to the cloud, whereby the securely protected metadata is used by the cloud to perform anti-fraud analysis on the first service; the processing device receives the anti-fraud analysis results returned by the cloud for the first service, whereby the anti-fraud analysis results indicate whether the first service carries a fraud risk.

[0006] Therefore, this method optimizes security resource allocation while ensuring the effectiveness of anti-fraud analysis by performing intelligent hierarchical and differentiated security processing of metadata on the terminal device side. For example, upon business triggering, the importance of metadata is dynamically determined locally on the terminal, applying high-strength encryption protection only to the identified key parts, while implementing integrity protection for all data. This mechanism overcomes the dilemma in existing terminal solutions where "full encryption leads to resource overload" and "partial encryption or plaintext transmission leads to security risks." It ensures that data packets transmitted to the cloud retain complete analytical value, guarantee that core secrets are not leaked and remain tamper-proof, achieving metadata trustworthiness, while significantly reducing the terminal's computational overhead and transmission load. This makes it highly suitable for scenarios where resource-constrained mobile terminals participate in high-security anti-fraud collaboration.

[0007] Optionally, the processing device performs mapping analysis on the metadata of the first service to determine the key and non-key metadata in the metadata of the first service, including: the processing device maps the metadata of the first service to a point cloud in three-dimensional space with each complete character in the metadata as the granularity, and saves the mapping context; each point in the point cloud corresponds to a complete character in the metadata of the first service; the processing device determines the key and non-key metadata by analyzing the spatial coordinate relationship of the point cloud points.

[0008] Therefore, by transforming text or structured data into a set of points in space, the inherent correlation strength and structural importance between data units can be objectively analyzed using the principles of geometric topology. This eliminates the subjectivity and lag of traditional reliance on predefined sensitive field rules, enabling adaptive and more fundamental discovery of core data units that support the integrity of business logic, making the key data identification process more robust, covert, and resistant to rule evasion.

[0009] Optionally, the processing device determines key metadata and non-key metadata by analyzing the spatial coordinate relationships of point cloud points in the point cloud. This includes: the processing device performing an aggregation operation on the point cloud points to obtain multiple discrete point cloud points in three-dimensional space, and the processing device also saving the context of the aggregation operation, wherein the spatial distance between any two aggregated point cloud points in the point cloud is less than a spatial distance threshold, and the spatial distance between any two discrete point cloud points among the multiple discrete point cloud points is greater than or equal to the spatial distance threshold; the processing device determines key metadata and non-key metadata by analyzing the spatial coordinate relationships of the multiple discrete point cloud points.

[0010] In this way, firstly, the analysis efficiency is greatly improved, simplifying the calculation of massive point-to-point relationships to the calculation of relationships between a few representative points, making the algorithm more suitable for the limited computing power of the terminal; secondly, through aggregation and abstraction, the cluster structure naturally formed in the original metadata is strengthened and highlighted, so that the subsequent judgment of key nodes can be based on a more stable and representative data skeleton, improving the accuracy and consistency of the recognition results.

[0011] Optionally, the multiple discrete point cloud points are N discrete point cloud points, where N is an integer greater than or equal to 3. The processing device analyzes the spatial coordinate relationships of the multiple discrete point cloud points to determine key metadata and non-key metadata, including: for the i-th discrete point cloud point among the N discrete point cloud points, where i is an integer from 1 to N; the processing device determines the spatial distances between the i-th discrete point cloud point and each of the N-1 discrete point cloud points, resulting in a total of N-1 spatial distances, where the N-1 discrete point cloud points are all the other discrete point cloud points besides the i-th discrete point cloud point; the processing device determines the sum of the N-1 spatial distances as the total spatial distance of the i-th discrete point cloud point. Distance: When i traverses from 1 to N, the total spatial distance of each of the N discrete point cloud points is obtained, resulting in a total of N total spatial distances. The processing device selects the M smallest total spatial distances from the N total spatial distances as the M total spatial distances, where M is an integer greater than or equal to 1 and less than N. The processing device determines key metadata based on the M discrete point cloud points that correspond one-to-one with the M total spatial distances from the N discrete point cloud points, and determines non-key metadata based on the NM discrete point cloud points that correspond one-to-one with the NM total spatial distances from the N discrete point cloud points. The NM total spatial distances are all the other total spatial distances from the N total spatial distances except for the M total spatial distances.

[0012] Therefore, by calculating the total spatial distance between each discrete point and all other points, and using the "minimum total distance" as a quantitative indicator of "centrality" or "coupling," it is possible to accurately pinpoint points located at the hub positions in the spatial topology. These points typically geometrically connect to or influence multiple data clusters, and their corresponding metadata characters thus possess stronger business relevance. This method transforms the abstract judgment of "importance" into rigorous mathematical calculations, making the entire identification process objective, transparent, and repeatable. It effectively eliminates the uncertainty of subjective rules and ensures the accuracy of key data screening.

[0013] Optionally, the processing device determines key metadata based on M discrete point cloud points that correspond one-to-one with M total spatial distances among N discrete point cloud points. This includes: the processing device performing the inverse operation of the aggregation operation on the M discrete point cloud points according to the context of the aggregation operation to obtain a first set of point cloud points corresponding to the M discrete point cloud points in the point cloud; and the processing device performing the inverse operation of the mapping on the first set of point cloud points according to the context of the mapping to obtain a first set of characters corresponding to the first set of point cloud points in the metadata of the first service, wherein the first set of characters is the key metadata.

[0014] Therefore, by utilizing pre-saved mapping and aggregation contexts, it is ensured that the analytical conclusions of the mathematical space can be accurately projected back into the business data domain. This technique achieves a closed-loop intelligent analysis chain, guaranteeing that the objects subjected to strong encryption protection are indeed the original data characters identified as crucial through topological analysis, thus avoiding security policy mismatches caused by information reconstruction errors.

[0015] Optionally, the processing device determines non-critical metadata based on NM discrete point cloud points that correspond one-to-one with NM total spatial distances. This includes: the processing device performing the inverse operation of the aggregation operation on the NM discrete point cloud points according to the context of the aggregation operation to obtain a second set of point cloud points corresponding to the NM discrete point cloud points in the point cloud; and the processing device performing the inverse operation of the mapping on the second set of point cloud points according to the context of the mapping to obtain a second set of characters corresponding to the second set of point cloud points in the metadata of the first service, wherein the second set of characters is non-critical metadata.

[0016] Therefore, by fully defining the inverse mapping process from non-critical point cloud points to original non-critical metadata, it is possible to complete the complete binary division of all metadata without omission. This provides a clear and accurate data partitioning basis for the subsequent implementation of a differentiated security strategy of "strong encryption of critical data and integrity protection only for non-critical data." This ensures that the hierarchical protection strategy can strictly cover all data to be transmitted.

[0017] Optionally, the processing device encrypts key metadata and performs integrity protection on both key and non-key metadata to obtain securely protected metadata. This includes: the processing device encrypts each complete character in the key metadata at the granular level, and performs integrity protection on each complete character in both key and non-key metadata; the processing device replaces a corresponding complete character in the metadata of the first service with each encrypted and integrity-protected complete character in the key metadata, and replaces a corresponding complete character in the metadata of the first service with each integrity-protected complete character in the non-key metadata, to obtain securely protected metadata.

[0018] This demonstrates that it achieves "on-demand allocation" of security protection strength. By applying computationally intensive encryption to a small amount of core data, it ensures a high level of confidentiality while controlling overall overhead; by implementing relatively lightweight integrity protection for all data, it constructs an overall defense against tampering. The use of in-situ replacement to generate the final data packet maintains the structure and order of the original data, greatly facilitating cloud reception and parsing, and improving the processing efficiency of end-to-cloud collaboration.

[0019] Optionally, each encrypted and integrity-protected complete character in the critical metadata has the same character length as each integrity-protected complete character in the non-critical metadata. The processing device sends the securely protected metadata to the cloud, including: the securely protected metadata and encrypted indication information. The indication information indicates the index of each encrypted and integrity-protected complete character in the critical metadata within the securely protected metadata. In other words, the cloud does not need to scan or perform complex parsing of the received data packet to locate the ciphertext; it can directly extract the encrypted portion for decryption based on the index information and the condition of the same character length. This significantly reduces the complexity of data preprocessing in the cloud, accelerates the startup speed of the anti-fraud analysis process, thereby reducing overall business processing latency and improving the response performance of the real-time anti-fraud system.

[0020] Secondly, a data analysis system is provided, characterized in that the data analysis system includes a user-side processing device and a cloud, wherein the processing device is configured to: when a user's first service is triggered for execution, the processing device performs mapping analysis on the metadata of the first service to determine key metadata and non-key metadata in the metadata of the first service; the processing device encrypts the key metadata and performs integrity protection on the key metadata and non-key metadata to obtain securely protected metadata; the processing device sends the securely protected metadata to the cloud, and the securely protected metadata is used by the cloud to perform anti-fraud analysis on the first service; the processing device receives the anti-fraud analysis results returned by the cloud for the first service, and the anti-fraud analysis results indicate whether the first service has a fraud risk.

[0021] It should be understood that the specific implementation of the second aspect can also refer to the method described in the first aspect, and will not be repeated here.

[0022] The objectives and other advantages of this invention can be realized and obtained through the following description. Attached Figure Description

[0023] To make the objectives, technical solutions, and advantages of the present invention clearer, the preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings, wherein: Figure 1 A schematic diagram of the architecture of a data analysis system provided by the present invention; Figure 2 A flowchart of a data processing method based on trusted metadata provided by the present invention; Figure 3 This is a schematic diagram of the structure of a processing device provided by the present invention. Detailed Implementation

[0024] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other. The accompanying drawings are for illustrative purposes only, representing only schematic diagrams and not actual physical objects, and should not be construed as limiting the present invention. To better illustrate the embodiments of the present invention, some parts in the drawings may be omitted, enlarged, or reduced, and do not represent the actual product form.

[0025] This application provides a data analysis system, which includes a user-side processing device and a cloud.

[0026] The processing device can be a terminal, or a terminal device. This terminal can also be referred to as a user unit, user station, mobile station, mobile station, remote station, remote terminal, mobile device, user terminal, terminal, wireless communication device, user agent, or user equipment. In the embodiments of this application, the terminal device can be a mobile phone, tablet computer, computer with wireless transceiver capabilities, virtual reality (VR) terminal device, augmented reality (AR) terminal device, wireless terminal in industrial control, wireless terminal in self-driving, wireless terminal in remote medical care, wireless terminal in smart grid, wireless terminal in transportation safety, wireless terminal in smart city, wireless terminal in smart home, vehicle-mounted terminal, RSU with terminal functionality, etc. The terminal device of this application may also be an on-board module, on-board component, on-board chip, or on-board unit that is built into a vehicle as one or more components or units. The vehicle can implement the method provided in this application through the built-in on-board module, on-board component, on-board chip, or on-board unit.

[0027] In this embodiment, the cloud is not a single node, but a layered, distributed, and highly available computing and service cluster. For example, the cloud can consist of a cluster of high-performance cloud servers deployed in a large data center. These servers are typically equipped with multi-core processors, large-capacity memory, and computing units or AI-dedicated accelerator cards optimized for machine learning workloads to efficiently run complex anti-fraud analysis models (including graph computing, causal inference models, etc.). To ensure data persistence and high-speed access, the backend uses distributed databases (such as time-series databases and graph databases) and object storage services. The entire cluster is orchestrated and managed through virtualization and containerization technologies (such as Kubernetes) to achieve elastic scaling and high availability of resources.

[0028] Furthermore, the cloud can also adopt a hybrid architecture of "edge access - central intelligence" to balance latency and computing power. Edge access points are widely deployed in regional operator data centers or public cloud edge computing nodes, and their equipment is mostly lightweight servers. They are mainly responsible for receiving massive amounts of data from terminals, performing preliminary verification, decryption, format standardization, and other preprocessing, and then aggregating the data to the central analysis cluster through a low-latency, high-bandwidth backbone network. The central cluster is located in the core data center, aggregating all data and performing complex global correlation analysis, model training, and policy calculation. This deployment method ensures both rapid response to terminal requests (through edge nodes) and the powerful computing power and global data perspective required for complex analysis (through the central cluster).

[0029] In this data analysis system, when a user's first service is triggered, the processing device performs mapping analysis on the metadata of the first service to determine the key and non-key metadata. The processing device encrypts the key metadata and performs integrity protection on both key and non-key metadata to obtain securely protected metadata. The processing device sends the securely protected metadata to the cloud, which is used by the cloud to perform anti-fraud analysis on the first service. The processing device receives the anti-fraud analysis results returned by the cloud for the first service, which indicate whether the first service has a fraud risk.

[0030] like Figure 2 As shown, a data processing method based on trusted metadata is provided. This method is applied to the aforementioned system, and the specific process of this method is as follows: S201, when the user's first service is triggered to execute, the processing device performs mapping analysis on the metadata of the first service to determine the key metadata and non-key metadata in the metadata of the first service.

[0031] The metadata for the first business can be key information describing the first business.

[0032] Example 1: When the first transaction is a payment transaction initiated by a user on an e-commerce platform, its metadata is a highly structured collection used to describe the characteristics and context of this transaction, including: 1. User and account identifiers, such as user ID and login session token; 2. Device and environment fingerprints, such as obfuscated device unique identifiers, device model and operating system, current IP address and geographical location (GPS / base station), and whether the device is jailbroken or in an emulator environment; 3. Core transaction information, such as order number, transaction amount, merchant code and type, and product or service category; 4. User interaction behavior sequence: a time-series log of a series of interface operations from entering the payment page to confirming payment, such as click coordinates, dwell time, input speed, and whether card number was copied and pasted; 5. Historical association information, such as the recent transaction frequency of the bank card used for this payment, a list of frequently used payment devices, and frequently used delivery addresses.

[0033] Example 2: When the primary business is logging into a bank's app or authorizing a large transfer, the metadata focuses on verifying the legitimacy of the identity and intent. Specifically, this may include: 1. Authentication credentials and methods, such as the entered account password, the hash value of the fingerprint template, or the SMS verification code; 2. A snapshot of the login environment, such as the device used for this login (compared to historically frequently used devices), the network environment (such as Wi-Fi SSID or cellular network operator), and the app's version and integrity signature; 3. Biometric characteristics of operational behavior: biometric indicators such as the pressure, area, and swipe trajectory characteristics of the touchscreen; 4. Temporal and contextual anomalies, such as the time of the login attempt (whether it occurred during an inactive period), the interval since the last successful login, and the foreground / background switching status.

[0034] Example 3: When the first business is for a user to change their mobile phone number, reset their password, or bind a new bank card / payment tool, the metadata focuses on identifying whether an abnormal transfer of account control is taking place, including: 1. The content of the modification itself, such as the location and operator information of the old and new mobile phone numbers; the BIN number and issuing bank of the old and new bank cards; 2. The operation path and verification, such as the channel through which the user initiated the modification (main APP, H5 link, customer service) and the verification steps completed (password, original SMS); 3. The associated environment and behavior, such as whether the device used during the operation is a new or unfamiliar device, whether the IP address has undergone a sudden geographical change, and the user's behavior path before the operation (whether they have just browsed fraudulent information or answered suspicious calls).

[0035] Furthermore, regarding S201, the processing device can perform the following operations: Step S1: The processing device maps the metadata of the first service into a point cloud in three-dimensional space at the granularity of each complete character in the metadata of the first service, and saves the mapping context. Each point in the point cloud corresponds to a complete character in the metadata of the first service.

[0036] Each complete character refers to a character with complete semantics. For example, in Example 1 above, the user ID, each token in this login session, and the device unique identifier are all complete characters. In Example 2 above, the account, password, and hash value of the fingerprint template are all complete characters. In Example 3 above, the location and operator information are all complete characters.

[0037] The algorithm used to map the metadata of the first business to a point cloud in three-dimensional space can be an existing character embedding algorithm. The mapping context can record the execution process of the algorithm, as well as the mapping relationship between each complete character and the spatial coordinates of the corresponding point cloud point.

[0038] Step S2: The processing device determines key metadata and non-key metadata by analyzing the spatial coordinate relationships of point cloud points in the point cloud.

[0039] The processing device performs point cloud aggregation operations on the point cloud, obtaining multiple discrete point cloud points in 3D space. The device also saves the context of the aggregation operation. Specifically, the spatial distance between any two aggregated point cloud points is less than a spatial distance threshold, and the spatial distance between any two discrete point cloud points is greater than or equal to the spatial distance threshold. In other words, the processing device can perform aggregation based on the spatial distance between point cloud points. If the spatial distance between any two point cloud points is less than the spatial distance threshold, these two points are aggregated into one point cloud point, such as the spatial coordinates of the aggregated point cloud point being the average of the spatial coordinates of the two aggregated points. This process continues until the spatial distance between the remaining point cloud points is greater than or equal to the spatial distance threshold, thus obtaining multiple discrete point cloud points. It should be understood that the spatial distance threshold can be set according to actual conditions, ensuring that the final number of discrete point cloud points is greater than or equal to 3.

[0040] The context of the aggregation operation can record the execution process of the aggregation, as well as which point cloud points each discrete point cloud point was aggregated from, such as recording the correspondence between the spatial coordinates of each point cloud point and the spatial coordinates of the discrete point cloud point.

[0041] The processing device determines key and non-key metadata by analyzing the spatial coordinate relationships of multiple discrete point cloud points.

[0042] For example, multiple discrete point cloud points are N discrete point cloud points, where N is an integer greater than or equal to 3. For the i-th discrete point cloud point among these N discrete point cloud points, where i is an integer from 1 to N, the processing device determines the spatial distances between the i-th discrete point cloud point and each of the other N-1 discrete point cloud points, resulting in N-1 spatial distances. These N-1 discrete point cloud points are all the other discrete point cloud points besides the i-th point. The processing device then determines the sum of these N-1 spatial distances as the total spatial distance to the i-th discrete point cloud point. Given that i iterates from 1 to N, this results in N discrete point cloud points. The processing device selects the M smallest total spatial distances from the N discrete point cloud points as the M total spatial distances, where M is an integer greater than or equal to 1 and less than N. Based on the M discrete point cloud points that correspond one-to-one with the M total spatial distances, the processing device determines key metadata. It also determines non-key metadata based on the NM discrete point cloud points that correspond one-to-one with the NM total spatial distances. The NM total spatial distances are all the other total spatial distances from the N total spatial distances except for the M total spatial distances. In other words, by calculating the total spatial distance between each discrete point and all other points, and using the "minimum total distance" as a quantifiable indicator of "centrality" or "coupling," the device can accurately pinpoint points located at hub positions in the spatial topology. These points typically geometrically connect to or influence multiple data clusters, and their corresponding metadata characters therefore possess stronger business relevance. This method transforms the abstract judgment of "importance" into rigorous mathematical calculations, making the entire identification process objective, transparent, and repeatable. It effectively eliminates the uncertainty of subjective rules and ensures the accuracy of key data screening.

[0043] For example, if N=5, for the first discrete point cloud point, it is necessary to calculate the spatial distance between it and the remaining 4 discrete point cloud points, that is, the spatial distance of 4 points in total. The sum of these 4 spatial distances is the total spatial distance of the first discrete point cloud point. The same applies to the second discrete point cloud point, etc.

[0044] In one possible implementation, the processing device can perform the inverse operation of the aggregation operation on M discrete point cloud points according to the context of the aggregation operation, to obtain the first set of point cloud points corresponding to the M discrete point cloud points in the point cloud. That is, based on the correspondence between the spatial coordinates of the aggregated point cloud points and the spatial coordinates of the discrete point cloud points obtained from the aggregation, the aggregated point cloud points, i.e., the first set of point cloud points, are found. The processing device then performs the inverse operation of the mapping on the first set of point cloud points according to the context of the mapping, to obtain the first set of characters corresponding to the first set of point cloud points in the metadata of the first business. The first set of characters is key metadata; that is, based on the mapping relationship between each complete character and the spatial coordinates of the corresponding point cloud point, a complete character corresponding to each point cloud point in the first set of point cloud points is found, i.e., the first set of characters is obtained. Therefore, by utilizing the pre-saved mapping and aggregation context, it is ensured that the analysis conclusions of the mathematical space can be accurately projected back to the business data domain. The effectiveness of this technology lies in achieving a closed-loop intelligent analysis link, ensuring that the object protected by strong encryption is the original data character that is identified as crucial through topology analysis, thus avoiding security policy mismatch caused by information reconstruction deviations.

[0045] Similarly, the processing device can also perform the inverse operation of the aggregation operation on NM discrete point cloud points according to the context of the aggregation operation, to obtain the second set of point cloud points corresponding to the NM discrete point cloud points in the point cloud; the processing device can then perform the inverse operation of the mapping on the second set of point cloud points according to the context of the mapping, to obtain the second set of characters corresponding to the second set of point cloud points in the metadata of the first service. The second set of characters is non-critical metadata, and the principle is similar to the above, which can be understood by referring to it. Therefore, by fully defining the inverse mapping process from non-critical point cloud points to the original non-critical metadata, it can complete the complete binary division of all metadata without omission, providing a clear and accurate data partitioning basis for the subsequent implementation of a differentiated security strategy of "strong encryption of critical data and only integrity protection of non-critical data." This ensures that the hierarchical protection strategy can strictly cover all data to be transmitted.

[0046] In this way, firstly, the analysis efficiency is greatly improved, simplifying the calculation of massive point-to-point relationships to the calculation of relationships between a few representative points, making the algorithm more suitable for the limited computing power of the terminal; secondly, through aggregation and abstraction, the cluster structure naturally formed in the original metadata is strengthened and highlighted, so that the subsequent judgment of key nodes can be based on a more stable and representative data skeleton, improving the accuracy and consistency of the recognition results.

[0047] Therefore, S201, by transforming text or structured data into a set of points in space, can objectively analyze the inherent correlation strength and structural importance between data units using the principles of geometric topology. This eliminates the subjectivity and lag of traditional methods that rely on predefined sensitive field rules, enabling adaptive and more fundamental discovery of core data units that support the integrity of business logic, making the key data identification process more robust, covert, and resistant to rule evasion.

[0048] S202, the processing device encrypts the critical metadata and performs integrity protection on both critical and non-critical metadata to obtain securely protected metadata.

[0049] For example, the processing device encrypts each complete character in critical metadata at the granular level, and performs integrity protection on each complete character in both critical and non-critical metadata. The confidentiality key used for encryption of each complete character can be the same, and the integrity key used for integrity protection can also be the same. The processing device replaces a corresponding complete character in the metadata of the first service with each encrypted and integrity-protected complete character in the critical metadata, and replaces a corresponding complete character in the metadata of the first service with each integrity-protected complete character in the non-critical metadata, thus obtaining securely protected metadata. For example, the first complete character in the metadata of the first service is encrypted and integrity-protected, resulting in encrypted and integrity-protected character #1, and the first complete character in the metadata of the first service is replaced with character #1. The second complete character in the metadata of the first service is integrity-protected, resulting in integrity-protected character #2, and the second complete character in the metadata of the first service is replaced with character #2, and so on. Therefore, it achieves "on-demand allocation" of security protection strength. By applying computationally intensive encryption to a small amount of core data, a high level of confidentiality is ensured while controlling overall overhead; by implementing relatively lightweight integrity protection for all data, a comprehensive defense against tampering is constructed. The final data packet is generated using an in-situ replacement method, maintaining the original data structure and order, greatly facilitating cloud reception and parsing, and improving the processing efficiency of end-to-cloud collaboration.

[0050] S203, the processing device sends securely protected metadata to the cloud, and the securely protected metadata is used by the cloud to perform anti-fraud analysis on the first business.

[0051] In this context, each encrypted and integrity-protected complete character in the critical metadata has the same character length as each integrity-protected complete character in the non-critical metadata, such as a preset fixed length. This fixed length is greater than or equal to the maximum character length of the encrypted and integrity-protected complete character, and it is also greater than or equal to the maximum character length of the integrity-protected complete character. In other words, the processing device can pad each encrypted and integrity-protected complete character to a fixed length, and also pad each integrity-protected complete character to a fixed length. The processing device then sends the securely protected metadata and encrypted indication information to the cloud. The indication information indicates the index of each encrypted and integrity-protected complete character in the critical metadata within the securely protected metadata. In other words, the cloud does not need to scan or perform complex parsing of the received data packet to locate the ciphertext; it can directly extract the encrypted portion for decryption based on the index information and the condition of the same character length. This significantly reduces the complexity of data preprocessing in the cloud, accelerates the startup speed of the anti-fraud analysis process, thereby reducing overall business processing latency and improving the response performance of the real-time anti-fraud system.

[0052] Therefore, the cloud can extract each complete character from the securely protected metadata based on a fixed length. The cloud can perform security checks on the encrypted indication information, such as decryption and integrity checks, to obtain the plaintext indication information. Then, based on the plaintext indication information, it can determine which complete characters are encrypted. The cloud can use an integrity key to perform integrity checks on each integrity-protected complete character. If all checks pass, the cloud can use a confidentiality key to decrypt each encrypted complete character, thus obtaining the plaintext metadata. The cloud can input the plaintext metadata into a large model for analysis to obtain anti-fraud analysis results, which indicate whether the first business has a fraud risk. Furthermore, the large model in this application embodiment can be existing technology and is not specifically limited.

[0053] S204, The processing device receives the anti-fraud analysis results returned by the cloud for the first business, and the anti-fraud analysis results indicate whether there is a risk of fraud in the first business.

[0054] In summary, this method optimizes security resource allocation while ensuring the effectiveness of anti-fraud analysis by performing intelligent hierarchical and differentiated security processing of metadata on the terminal device side. For example, upon business triggering, the importance of metadata is dynamically determined locally on the terminal, applying high-strength encryption protection only to the identified critical parts, while implementing integrity protection for all data. This mechanism overcomes the dilemma in existing terminal solutions where "full encryption leads to resource overload" and "partial encryption or plaintext transmission leads to security risks." It ensures that data packets transmitted to the cloud retain complete analytical value, guarantee that core secrets are not leaked and remain tamper-proof, achieving metadata trustworthiness, while significantly reducing the terminal's computational overhead and transmission load. This makes it highly suitable for scenarios where resource-constrained mobile terminals participate in high-security anti-fraud collaboration.

[0055] Figure 3 This is a schematic diagram of the structure of a processing device provided in an embodiment of this application. Exemplarily, the processing device may be a terminal, or a chip (system) or other component or assembly that can be disposed on the terminal. Figure 3 As shown, the processing device 300 may include a processor 301. Optionally, the processing device 300 may also include a memory 302 and / or a transceiver 303. The processor 301 is coupled to the memory 302 and the transceiver 303, for example, via a communication bus.

[0056] The following is combined Figure 3 A detailed description of each component of the processing equipment 300 is provided below: The processor 301 is the control center of the processing device 300. It can be a single processor or a collective term for multiple processing elements. For example, the processor 301 can be one or more central processing units (CPUs), application-specific integrated circuits (ASICs), or one or more integrated circuits configured to implement the embodiments of this application, such as one or more digital signal processors (DSPs), or one or more field-programmable gate arrays (FPGAs).

[0057] Optionally, the processor 301 can perform various functions of the processing device 300 by running or executing software programs stored in the memory 302 and by calling data stored in the memory 302, such as performing the above-mentioned functions. Figure 2 The method shown.

[0058] In a specific implementation, as one example, processor 301 may include one or more CPUs, for example... Figure 3 CPU0 and CPU1 are shown in the diagram.

[0059] In a specific implementation, as one example, the processing device 300 may also include multiple processors, for example... Figure 3 The processor 301 shown is an example. Each of the processors 301 can be a single-core processor (single-CPU) or a multi-core processor (multi-CPU). Here, a processor can refer to one or more devices, circuits, and / or processing cores used to process data (e.g., computer program instructions).

[0060] The memory 302 is used to store the software program that executes the solution of this application, and is controlled by the processor 301 to execute it. The specific implementation method can be referred to the above method embodiment, and will not be repeated here.

[0061] Optionally, the memory 302 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto. The memory 302 may be integrated with the processor 301 or exist independently, and may be connected via the interface circuit of the processing device 300. Figure 3 (Not shown in the image) is coupled to processor 301, but this embodiment does not specifically limit this.

[0062] Transceiver 303 is used for communication with other processing devices. For example, if processing device 300 is a terminal, transceiver 303 can be used to communicate with a network device or with another terminal device. As another example, if processing device 300 is a network device, transceiver 303 can be used to communicate with a terminal or with another network device.

[0063] Alternatively, transceiver 303 may include a receiver and a transmitter. Figure 3 (Not shown separately). The receiver is used to implement the receiving function, and the transmitter is used to implement the transmitting function.

[0064] Optionally, the transceiver 303 can be integrated with the processor 301, or it can exist independently and be connected via the interface circuit of the processing device 300. Figure 3 (Not shown in the image) is coupled to processor 301, but this embodiment does not specifically limit this.

[0065] Understandable, Figure 3 The structure of the processing device 300 shown does not constitute a limitation on the processing device. The actual processing device may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0066] Furthermore, the technical effects of the processing device 300 can be referred to the technical effects of the method described in the above method embodiments, and will not be repeated here.

[0067] It should be understood that the processor in the embodiments of this application can be a central processing unit (CPU), or it can be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor, etc.

[0068] It should also be understood that the memory in the embodiments of this application can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDR SDRAM), enhanced synchronous DRAM (ESDRAM), synchronous linked DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0069] The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0070] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0071] In this application, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of a single item or a plurality of items. For example, at least one of a, b, or c can mean: a, b, c, ab, ac, bc, or abc, where a, b, and c can be a single item or multiple items.

[0072] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0073] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0074] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0075] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0076] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0077] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0078] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0079] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A data processing method based on trusted metadata, characterized in that, A processing device applied on the user side, the method comprising: When the user's first service is triggered, the processing device performs mapping analysis on the metadata of the first service to determine the key and non-key metadata in the metadata of the first service. The processing device encrypts the key metadata and performs integrity protection on both key and non-key metadata to obtain securely protected metadata. The processing device sends the securely protected metadata to the cloud, and the securely protected metadata is used by the cloud to perform anti-fraud analysis on the first service; The processing device receives the anti-fraud analysis results returned by the cloud for the first service, and the anti-fraud analysis results indicate whether the first service is at risk of fraud.

2. The method according to claim 1, characterized in that, The processing device performs mapping analysis on the metadata of the first service to determine key and non-key metadata within the metadata of the first service, including: The processing device maps the metadata of the first service into a point cloud in three-dimensional space at the granularity of each complete character in the metadata of the first service, and saves the context of the mapping; each point in the point cloud corresponds to a complete character in the metadata of the first service. The processing device determines the critical metadata and the non-critical metadata by analyzing the spatial coordinate relationships of the point cloud points.

3. The method according to claim 2, characterized in that, The processing device determines the critical metadata and the non-critical metadata by analyzing the spatial coordinate relationships of the point cloud points, including: The processing device performs a point cloud point aggregation operation on the point cloud to obtain multiple discrete point cloud points in the three-dimensional space, and the processing device also saves the context of the aggregation operation, wherein the spatial distance between any two aggregated point cloud points in the point cloud is less than a spatial distance threshold, and the spatial distance between any two discrete point cloud points among the multiple discrete point cloud points is greater than or equal to the spatial distance threshold. The processing device determines the key metadata and the non-key metadata by analyzing the spatial coordinate relationships of the multiple discrete point cloud points.

4. The method according to claim 3, characterized in that, The plurality of discrete point cloud points are N discrete point cloud points, where N is an integer greater than or equal to 3. The processing device determines the key metadata and the non-key metadata by analyzing the spatial coordinate relationship of the plurality of discrete point cloud points, including: For the i-th discrete point in the N discrete point cloud points, where i is an integer from 1 to N; The processing device determines the spatial distances between the i-th discrete point cloud point and N-1 discrete point cloud points, obtaining a total of N-1 spatial distances. The N-1 discrete point cloud points are all the other discrete point cloud points among the N discrete point cloud points except for the i-th discrete point cloud point. The processing device determines the sum of the N-1 spatial distances as the total spatial distance of the i-th discrete point cloud point. When i traverses from 1 to N, it obtains the total spatial distance of each of the N discrete point cloud points, for a total of N total spatial distances. The processing device selects the M smallest total spatial distances from the N total spatial distances as the M total spatial distances, where M is an integer greater than or equal to 1 and less than N; The processing device determines the key metadata based on M discrete point cloud points that correspond one-to-one with the M total spatial distances from the N discrete point cloud points, and determines the non-key metadata based on NM discrete point cloud points that correspond one-to-one with the NM total spatial distances from the N discrete point cloud points, wherein the NM total spatial distances are all other total spatial distances from the N total spatial distances except for the M total spatial distances.

5. The method according to claim 4, characterized in that, The processing device determines the key metadata based on M discrete point cloud points from the N discrete point cloud points that correspond one-to-one with the M total spatial distances, including: The processing device performs the inverse operation of the aggregation operation on the M discrete point cloud points according to the context of the aggregation operation, to obtain the first set of point cloud points corresponding to the M discrete point cloud points in the point cloud. The processing device performs the inverse operation of the mapping on the first point cloud point set according to the context of the mapping, to obtain the first character set corresponding to the first point cloud point set in the metadata of the first service, and the first character set is the key metadata.

6. The method according to claim 4, characterized in that, The processing device determines the non-critical metadata based on the NM discrete point cloud points that correspond one-to-one with the NM total spatial distances, including: The processing device performs the inverse operation of the aggregation operation on the NM discrete point cloud points according to the context of the aggregation operation, to obtain a second set of point cloud points corresponding to the NM discrete point cloud points in the point cloud. The processing device performs the inverse operation of the mapping on the second point cloud point set according to the context of the mapping, to obtain the second character set corresponding to the second point cloud point set in the metadata of the first service, and the second character set is the non-critical metadata.

7. The method according to any one of claims 1-6, characterized in that, The processing device encrypts the critical metadata and performs integrity protection on both critical and non-critical metadata to obtain securely protected metadata, including: The processing device encrypts each complete character in the key metadata at the granular level, and performs integrity protection on each complete character in both the key metadata and non-key metadata. The processing device replaces a corresponding complete character in the metadata of the first service with each encrypted and integrity-protected complete character in the critical metadata, and replaces a corresponding complete character in the metadata of the first service with each integrity-protected complete character in the non-critical metadata, to obtain the securely protected metadata.

8. The method according to claim 7, characterized in that, Each encrypted and integrity-protected complete character in the critical metadata has the same character length as each integrity-protected complete character in the non-critical metadata.

9. The method according to claim 8, characterized in that, The processing device sends the securely protected metadata to the cloud, including: The processing device sends the securely protected metadata and encrypted indication information to the cloud. The indication information is used to indicate the index of each encrypted and integrity-protected complete character in the key metadata within the securely protected metadata.

10. A data analysis system, characterized in that, The data analysis system includes user-side processing equipment and a cloud platform, wherein the processing equipment is configured as follows: When the user's first service is triggered, the processing device performs mapping analysis on the metadata of the first service to determine the key and non-key metadata in the metadata of the first service. The processing device encrypts the key metadata and performs integrity protection on both key and non-key metadata to obtain securely protected metadata. The processing device sends the securely protected metadata to the cloud, and the securely protected metadata is used by the cloud to perform anti-fraud analysis on the first service; The processing device receives the anti-fraud analysis results returned by the cloud for the first service, and the anti-fraud analysis results indicate whether the first service is at risk of fraud.

Images