Unified social credit code data quality control method based on traceability technology

Abstract

The invention discloses a unified social credit code data quality control method based on a traceability technology, and relates to the technical field of quality control, and the method comprises the steps: collecting an original data set with a unified social credit code as an index, calculating the activeness, and forming a data stream with an activeness label; constructing a tree-shaped credit primitive structure through a depth-first algorithm, and determining an authoritative data source in combination with a Gaussian integral minimization model; conflict resolution is completed by using dynamic weight distribution, and a consistent credit file is output; sequentially executing structured verification, semantic consistency verification and business logic comparison, identifying deviation conditions of the key fields, and generating a difference report; based on the difference report, consistency verification is completed through version number verification and block chain evidence storage; and if the problem is not solved overtime, turning to manual work order processing. According to the method, risks caused by multi-source data conflicts and deviations are effectively reduced, and an accurate and credible data basis is provided for credit information management, financial supervision and decision support.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of quality control, and particularly relates to a unified social credit code data quality control method based on traceability technology. BACKGROUND

[0002] In recent years, with the construction of social credit system and the development of big data technology, the unified social credit code has become an important index for enterprise and organization information management, promoting the research of various credit data collection, integration and application. In the prior art, through multi-source data collection and data cleaning method, the core information such as enterprise subject, equity structure, registered capital and business scope can be obtained, and the data is standardized processed, providing basic data support for credit evaluation, risk control and supervision. At the same time, the data quality control method is constantly optimized, for example, using the activity index to evaluate the data update frequency and reliability, and improving the data processing efficiency through hierarchical indexing and structured storage. However, how to realize the structured association, conflict resolution and credible verification of key fields in the multi-source and heterogeneous data environment is still the focus and difficulty of current research.

[0003] CN119648246A discloses a tracking and tracing system based on big data analysis, which cleans and verifies the original data through a data quality verification module, and uses a data comparison algorithm to verify the consistency of multi-source data, while supplemented by anonymization and permission control technology to ensure the confidentiality of data in the storage and transmission process. This technology can improve the data accuracy and supervision efficiency to a certain extent, and promote information sharing among participants. However, the technical solution mainly relies on data comparison and permission control means, and lacks fine-grained processing of structured association and activity difference analysis between key fields in multi-source data, which cannot fully solve the problems of heterogeneous data conflict, dynamic authoritative source determination and hierarchical identification of key field deviation.

[0004] CN120234211B discloses an intelligent data quality monitoring method and system, which realizes the detection and processing of abnormal data by monitoring the integrity, accuracy and consistency of data flow, and provides report generation and alarm mechanism. This method has certain effect in monitoring single data dimension or static rules, but still has some deficiencies in practical application: it cannot perform recursive structured analysis on the semantic dependency relationship of key fields, and it is also difficult to dynamically generate hierarchical difference groups combined with data activity, which leads to insufficient precision in data deviation identification, and limited risk visualization and decision assistance capability.

[0005] In summary, the existing big data tracking and quality monitoring technology has problems such as limited conflict resolution capability of multi-source data, insufficient structured and hierarchical analysis of key fields, etc. when processing unified social credit code related data. The present application provides a unified social credit code data quality control method based on traceability technology, taking the unified social credit code as the root node, and recursively mounting the key fields and activity according to the semantic dependency relationship to form a multi-level, logically clear tree structure, realizing the structured association of key fields; using statistical optimization method to determine the authoritative source in multi-source data, and solving the field conflict based on dynamic weight adjustment to realize consistent credit file output. SUMMARY

[0006] This part aims to summarize some aspects of the embodiments of the present application and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this part and the abstract and title of the specification to avoid obscuring the purpose of this part, abstract and title, and such simplifications or omissions cannot be used to limit the scope of the present application.

[0007] In view of the above existing problems, the present application is proposed.

[0008] Therefore, the technical problem solved by the present application is how to improve the consistency, reliability and traceability of multi-source data.

[0009] To solve the above technical problems, the present application provides the following technical solutions: collecting a raw data set indexed by a unified social credit code and calculating an activity to form a data stream with activity annotation; constructing a tree-shaped credit primitive structure by a depth-first algorithm for the data stream, and determining an authoritative data source by combining a Gaussian integral minimization model; completing conflict resolution by dynamic weight distribution and outputting a consistent credit file; based on the consistent credit file, sequentially performing structured verification, semantic consistency checking and business logic comparison to identify the deviation of key fields and generate a difference report; based on the difference report, completing consistency verification through version number checking and blockchain storage; if not resolved within a timeout, it is transferred to a manual work order processing.

[0010] As a preferred scheme of the present application, wherein: the calculation of the activity includes: extracting the query frequency, the update time difference value and the new data weight factor of each record in the raw data set to calculate the activity according to the following formula:

[0011]

[0012] wherein, is the activity; is the query frequency factor, which is the continuous value of the query frequency normalized to [0, 1]; 、 2、 3 is a weight coefficient; is a time decay rate, is a half-life period; is a time decay coefficient; is a difference between a current time and a last update time of data; is a maximum valid time window; if the activity is greater than the 80th percentile of the historical activity distribution, it is marked as high activity; otherwise, it is marked as low activity.

[0013] As a preferred scheme of the present application, wherein: the tree-shaped credit primitive structure constructed by the depth-first algorithm comprises: receiving the data stream, and taking the unified social credit code field as the root node of the tree-shaped structure, while binding the activity value; taking the key fields in the data stream as input, generating a semantic dependency matrix according to the inherent logical relationship between the key fields, wherein the matrix element is 1, indicating that there is a parent-child relationship or logical dependency, otherwise it is 0; starting from the root node, recursively expanding the key fields marked as dependent relationship in the semantic dependency matrix one by one, and hanging each key field and the corresponding activity together to form a child node and child node activity, and obtaining the tree-shaped credit primitive structure; the semantic dependency matrix is generated based on a predefined business rule library, and the business rule library defines the parent-child dependency relationship between the key fields.

[0014] As a preferred scheme of the present application, wherein: the authoritative data source determined by combining the Gaussian integral minimization model comprises: extracting the path field and the activity of each branch path in the tree-shaped credit primitive structure to form a path field set; calculating the path field deviation of each path field: if the field type is a numerical field, the absolute value of the difference between the current value of the field and the weighted mean of all candidate values of the corresponding field in the same data source set is calculated; if the field type is a date field, the absolute value of the day difference between the current value of the field and the weighted mean of all candidate values of the corresponding field in the same data source set is calculated; if the field type is a text field, the field deviation is calculated according to the edit distance or string similarity; for all path fields of each path, input each path field deviation into a normal distribution function to obtain a field Gaussian integral value; the mean of the normal distribution function takes the mean of the historical value, and the standard deviation takes the standard deviation of the historical value; the field Gaussian integral value and the activity in the path are calculated according to the weighted sum to obtain a path matching score; the path residual is calculated by weighting and summing the path matching score and the activity; the data source corresponding to the path with a path residual less than a preset residual threshold is included in the authoritative data source set.

[0015] As a preferred scheme of the present application, wherein: the conflict resolution using dynamic weight distribution and outputting a consistent credit file includes: obtaining path field deviation for the path field with conflict in the tree credit primitive structure; adjusting the path field weight in proportion according to the path field deviation and the path residual, and normalizing the path weight of the same path field after adjustment; the initial value of the path field weight is the activity value; using the adjusted path weight to weight and fuse the path field with conflict, selecting the path field value with the highest weighted score as the authoritative field value, and recording the corresponding data source; combining all path fields after dynamic weight distribution and conflict resolution to generate a consistent credit file, including the authoritative field value, the corresponding authoritative data source and the activity.

[0016] As a preferred scheme of the present application, wherein: the calculation formula for adjusting the path field weight in proportion is:

[0017] wherein, is the path deviation ratio, calculated by dividing the path field deviation by the path residual, is the initial value of the path field weight, is the adjusted path field weight.

[0018] As a preferred scheme of the present application, wherein: the deviation condition of the key field includes: extracting each authoritative field value, corresponding path residual and corresponding activity in the consistent credit file to construct a deviation detection vector, the elements of the deviation detection vector are three-tuple <authoritative field name, path residual, activity>; dividing the deviation detection vector into deviation levels, marking the field with path residual lower than the first threshold value as low deviation, marking the field with path residual between the first threshold value and the second threshold value as medium deviation, and marking the field with path residual higher than the second threshold value as high deviation; generating a key field deviation distribution table according to the activity and the deviation level, marking the field with high activity and high deviation as a sensitive field, and marking the field with low activity and low deviation as a stable field.

[0019] As a preferred scheme of the present application, wherein: the difference report generated according to the hierarchical decision tree model includes: constructing a two-dimensional deviation projection matrix based on the key field deviation distribution table, wherein the rows represent the deviation levels and the columns represent the activity intervals; performing hierarchical clustering operation on the two-dimensional deviation projection matrix, and dividing the authoritative fields with the same deviation level and activity level into the same difference group; assigning a difference label to each difference group, and generating a difference report combined with the historical version record of the authoritative field in the difference group.

[0020] As a preferred scheme of the present application, wherein: the consistency verification through version number checking and blockchain storage includes: checking the sensitive fields in the difference report according to the version number in the authoritative data source set, if the version number sequence of the same unified social credit code field in different data source paths is discontinuous or the same version number corresponds to different authoritative field values, it is considered that version number conflict occurs; if version number conflict occurs, priority is given to the consistency comparison of the authoritative field values, if the authoritative field values are consistent, only the version number record is updated; if the authoritative field values are different, the authoritative field values and the corresponding authoritative data source are written into the blockchain to generate an irrevocable difference certificate.

[0021] In a second aspect, the present application provides a computer device, comprising a memory and a processor, the memory stores a computer program, wherein: the computer program instructions are executed by the processor to realize the steps of the unified social credit code data quality control method based on the traceability technology as described in the present application.

[0022] In a third aspect, the present application provides a computer readable storage medium having a computer program stored thereon, wherein: the computer program instructions are executed by the processor to realize the steps of the unified social credit code data quality control method based on the traceability technology as described in the present application.

[0023] The present application has the following beneficial effects: the present application can dynamically identify high-risk and sensitive fields according to data update frequency and historical activity, and at the same time distinguish low-activity stable fields, thereby enhancing data consistency and reliability; by classifying and clustering the deviation of key fields, the present application improves the discovery ability of abnormal and conflict data, which helps to timely handle potential problems and reduce the need for manual intervention; according to the traceable verification and storage mechanism, the data processing process is further ensured to be transparent and tamper-proof.

[0024] The present application effectively reduces the risk brought by multi-source data conflict and deviation, provides accurate and reliable data basis for credit information management, financial supervision and decision support, and at the same time improves the efficiency and transparency of overall data governance. BRIEF DESCRIPTION OF DRAWINGS

[0025] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor. Among them: Figure 1 The flowchart of the unified social credit code data quality control method based on the traceability technology shown in the present application. DETAILED DESCRIPTION

[0026] In order to make the above objectives, features and advantages of the present application more obvious and easy to understand, the specific embodiments of the present application are described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments.

[0027] Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should belong to the protection scope of the present application.

[0028] In the following description, many specific details are set forth in order to provide a thorough understanding of the present application. However, the present application can be practiced in other manners different from those described herein, and those skilled in the art can make similar generalizations without departing from the spirit and scope of the present application. Therefore, the present application is not limited to the specific embodiments disclosed below.

[0029] According to the embodiments of the present application, combined with the flowchart shown in the drawings, the unified social credit code data quality control method based on the traceability technology comprises: Figure 1 As shown in the flowchart, the unified social credit code data quality control method based on the traceability technology comprises: S1: Collecting a raw data set indexed by a unified social credit code, and calculating an activity degree to form a data stream with an activity degree label.

[0030] S1.1: Collecting a raw data set indexed by a unified social credit code, and the set covers 12 key fields.

[0031] Through parallel grabbing of data interfaces of official platforms, credit investigation institutions and local government departments, the contents collected mainly include 12 key fields. Among them, the 12 key fields include: subject name, unified code, legal person information, equity structure, registered capital, date of establishment, registered address, business scope, enterprise status, business period, registration authority, and main management personnel.

[0032] Preferably, in order to balance data coverage and timeliness, an asynchronous submission mechanism is mainly used in the grabbing process, supplemented by a synchronous submission mechanism, and a priority queue is set for grabbing tasks from different sources, and high update frequency data sources are processed preferentially.

[0033] Specifically, high update frequency data source tasks are added to the priority queue, and multi-thread parallel scheduling grabbing processes are used, while the grabbing state and response time are monitored. After each grabbing is completed, the grabbing record is compared with the historical record to confirm whether there is repetition or conflict, and the field integrity, source identification and grabbing time of each record are ensured.

[0034] Finally, the raw data set is output.

[0035] S1.2: Calculate the activity degree label to form a data stream with an activity degree label.

[0036] Input the original data set, extract the query frequency, update time difference and new data weight factor for each record, and calculate the activity according to the following formula:

[0037]

[0038] wherein, is the activity, ranging from [0, 1]; is the query frequency factor, which is the query frequency normalized to [0, 1] continuous value; , 2, 3 is the weight coefficient, the recommended value is 0.4, 0.3, 0.3 respectively, and it is allowed to be adjusted according to business; is the time decay rate, is the half-life period; is the time decay coefficient; is the difference between the current time and the last update time of the data; is the maximum effective time window, used for linear freshness truncation, and the default is set to 720 hours (i.e. 30 days).

[0039] wherein, the query frequency refers to the number of times the corresponding field content is marked as suspected error or inconsistency by third parties or the system, and is normalized to [0, 1] continuous value as the query frequency factor.

[0040] If the activity is greater than the 80th percentile of the historical activity distribution, it is marked as high activity; otherwise, it is marked as low activity.

[0041] S2: pass the data stream through the depth-first algorithm to construct a tree-shaped credit primitive structure, and determine the authoritative data source by combining the Gaussian integral minimization model; use dynamic weight allocation to complete conflict resolution and output consistent credit archives.

[0042] S2.1: construct a tree-shaped credit primitive structure by a depth-first algorithm.

[0043] S2.1.1: receive the data stream, and take the unified social credit code field as the root node of the tree structure, while binding the activity value. The activity value is used as the initial weight of the node, and the setting of the root node ensures that each credit record has a unified index in the tree structure.

[0044] S2.1.2: take the key fields in the data stream as input, and generate a semantic dependency matrix according to the inherent logical relationship between the key fields, wherein the matrix element is 1 indicating the existence of parent-child relationship or logical dependence, otherwise 0.

[0045] It should be noted that the semantic dependency matrix is ​​generated based on a predefined business rule base. This base defines the parent-child dependencies, logical order, and constraints between key fields. For example, legal entity information depends on the entity name, and equity structure depends on legal entity information. When generating the matrix, the field activity level of each record needs to be recorded so that this activity level information can be synchronously attached to the corresponding child nodes during subsequent recursion, thus binding node activity level with the actual field activity level.

[0046] S2.1.3: Starting from the root node, recursively expand each key field marked as a dependency in the semantic dependency matrix, and attach each key field and its corresponding activity level to form child nodes and child node activity levels, thus obtaining a tree-like credit primitive structure.

[0047] For example, information about the legal entity and key management personnel depends on the entity name / unified code; the registered address depends on the registration authority; and the equity structure depends on the legal entity information.

[0048] During each expansion, the activity level of the current node is passed to its child nodes, and weighted in conjunction with the activity level of the child nodes' own fields to form the child node's activity level. During the recursive process, node paths, node activity levels, and node origins are recorded to form a complete tree-like credit primitive structure.

[0049] S2.2: Determine the authoritative data source by combining the Gaussian integral minimization model.

[0050] S2.2.1: Extract the path field and activity level for each branch path in the tree-like credit primitive structure to form a path field set. Each path field includes a field value, field type, and activity level.

[0051] S2.2.2: Calculate the path field deviation for each path field: If the field type is numeric, the arithmetic difference is used to calculate the absolute value of the difference between the current value of the field and the weighted mean of all candidate values ​​of the corresponding field in the same data source set; where the weighted mean is provided by activity level, and fields with higher activity level have a larger weight in the mean calculation.

[0052] If the field type is a date field, calculate the absolute value of the difference in days between the current value of the field and the weighted average of all candidate values ​​of the corresponding field in the same data source set.

[0053] If the field type is text, the field bias is calculated based on edit distance or string similarity. Specifically, edit distance calculation includes character-by-character comparison, insertion, deletion, and replacement operations, ultimately obtaining a quantified similarity value between the current field value and the weighted average string in the candidate value set. Text bias calculation allows for authoritative evaluation of unstructured text fields such as company names and registered addresses.

[0054] For all path fields of each path, the field deviation of each path field is input into a normal distribution function to obtain a field Gaussian integral value; the mean of the normal distribution function is the mean of the field historical value, and the standard deviation is the standard deviation of the historical value.

[0055] S2.2.3: The path field Gaussian integral value and the activity in the path are calculated according to weighted summation to obtain a path matching score.

[0056] Specifically, the path matching score is calculated for all field deviations in each path. The Gaussian integral value of each field represents the probability density of the deviation of the weighted mean of the current value and the candidate value set of the corresponding field from the normal distribution. The field weight is provided by the field activity, which is calculated by multiplying the Gaussian integral value, the activity, the path level attenuation coefficient and the field category weight. The path matching score is the cumulative sum of the weighted values of all fields on the path. Among them, the root node field contributes most to the path matching, and the contribution gradually decreases with the increase of the level of the child node, the value range of the path level attenuation coefficient is [0, 1], the closer to 1, the slower the attenuation, and the more attention to the leaf node field; the smaller the path level attenuation coefficient, the faster the attenuation, and the importance of the high-level node field is emphasized. The default value is recommended to be 0.9, which can be adjusted according to the actual business needs. The field category weight is used to distinguish the contribution of different types of fields to the authoritative determination in path matching, which can be predefined by the business rule library.

[0057] S2.2.4: Calculate the weighted sum of the path matching score and the activity to obtain the path residual, which represents the difference between the overall deviation of the path and the priori activity.

[0058] Specifically, the path residual is calculated by the absolute value of the difference between the path matching score and the cumulative value of the activity. By taking the absolute value, the degree of deviation of the path matching score from the path activity can be clearly quantified.

[0059] S2.2.5: The data source corresponding to the path with a path residual less than a preset residual threshold is included in the authoritative data source set.

[0060] S2.3: Conflict resolution is completed by using dynamic weight distribution, and a consistent credit file is output.

[0061] For the path field with conflict in the tree-shaped credit primitive structure, the path field deviation is obtained; that is, there are inconsistent values on the same field in multiple data sources corresponding to the same unified social credit code.

[0062] According to the path field deviation and the path residual, the path field weight is adjusted in proportion, and the path weight of the same path field after adjustment is normalized; The calculation formula for adjusting the path field weight in proportion is:

[0063] wherein, is a path deviation ratio, calculated by dividing the path residual by the path field deviation, is an initial value of the path field weight, is an adjusted path field weight, and the initial value of the path field weight is the activity value.

[0064] The path fields with conflicts are weighted and fused using the adjusted path weight, the path field value with the highest weighted score is selected as the authoritative field value, and the corresponding data source is recorded. It can be seen that the application not only solves the multi-source field conflict problem, but also ensures that the source of the authoritative field value is traceable, and takes into account the field activity and historical deviation, forming a highly reliable consistent credit file.

[0065] All path fields after dynamic weight allocation and conflict resolution are combined to generate a consistent credit file, including authoritative field values, corresponding authoritative data sources and activity.

[0066] S3: Based on the consistent credit file, sequentially perform structured verification, semantic consistency checking and business logic comparison to identify the deviation of the key field, and generate a difference report.

[0067] S3.1: Identify the deviation of the key field.

[0068] Extract each authoritative field value, corresponding path residual and corresponding activity in the consistent credit file to construct a deviation detection vector, and the elements of the deviation detection vector are triplets <authoritative field name, path residual, activity>.

[0069] It should be noted that the deviation detection vector as a quantitative tool can accurately describe the deviation level of each field and its activity weight in the data stream. By associating the path residual and the activity, it can be identified in subsequent analysis which field deviation may affect the reliability of the overall credit file.

[0070] Divide the deviation detection vector into deviation levels, mark the fields with path residuals below the first threshold as low deviation, mark the fields with path residuals between the first threshold and the second threshold as medium deviation, and mark the fields with path residuals above the second threshold as high deviation. The first threshold and the second threshold are determined by business strategy or historical data statistical quantile, and can be adjusted according to different business scenarios, and the embodiments of the application are not limited to only one.

[0071] According to the activity and the deviation level, a key field deviation distribution table is generated, and the fields with high activity and high deviation are marked as sensitive fields, and the fields with low activity and low deviation are marked as stable fields.

[0072] Specifically, a two-dimensional mapping is generated according to the activity and bias level, and the sensitive field indicates that the field has high influence and large bias in multi-source data, and can have a significant impact on business decision-making and credit assessment; the stable field indicates that the data has high reliability and stable contribution to the overall credit profile.

[0073] Through the bias distribution table, the technician can intuitively master the bias situation of the key field, and provide basic data for difference group division and risk control.

[0074] Further, in the bias identification process, the historical authoritative data source information of each field is also recorded to trace the field source during bias analysis. By combining historical version data and path residual, bias detection can not only reflect the current state of the field, but also provide field trend analysis, providing a reference for enterprise credit monitoring and auditing.

[0075] S3.2: Generate a difference report according to the hierarchical decision tree model.

[0076] A, based on the key field bias distribution table, construct a two-dimensional bias projection matrix of bias level and activity, wherein the row represents the bias level and the list represents the activity interval.

[0077] The matrix element is the corresponding authoritative field set, and the matrix can intuitively present the distribution of field bias and activity. The construction of the two-dimensional matrix allows the technician to quickly identify high-risk fields through matrix visualization.

[0078] Perform hierarchical clustering operation on the two-dimensional bias projection matrix to divide the authoritative fields with the same bias level and activity level into the same difference group.

[0079] In an embodiment of the present application, performing hierarchical clustering operation includes: B, the field set located in the same cell in the two-dimensional bias projection matrix forms a preliminary candidate difference group, and the number of fields in the group, the average activity, the bias level distribution and the field parent node information (root node or upper key field) are recorded; traverse each field in the candidate difference group, and judge the parent-child relationship according to the semantic dependence matrix: If there is a parent node in the field in the group in the two-dimensional bias projection matrix, the corresponding field is removed from the current group and hung to the child level of the group where the parent node is located, forming a child node hierarchy; if there is no parent node in the field in the group, it is kept in the top difference group of the current group; For the adjusted candidate difference group, recursive aggregation is performed according to the deviation level and the average activity level: the high deviation and high activity field is preferentially formed into a top-level difference group; if the low deviation or low activity field has a parent-child dependency with the field of the top-level difference group, the low deviation or low activity field is regarded as a child node and is included in the top-level group; otherwise, a middle-low level difference group is formed; and the average activity and the deviation distribution of the field in each layer of aggregation are updated.

[0080] Optionally, consistency checking is performed on the generated hierarchical difference group: the activity difference of the field in the group is within a reasonable range (based on historical activity fluctuation); and the dependency relationship between the field in the group and the parent node conforms to the semantic dependency matrix definition (1 represents dependency, and 0 represents no dependency).

[0081] C. A difference label is assigned to each difference group, and a difference report is generated in combination with the historical version record of the authoritative field in the difference group.

[0082] The difference report includes the difference label, the set of authoritative field names in the difference group, the path residual interval and the activity distribution of the field in the difference group, and the historical authoritative data source identifier of each field in the difference group.

[0083] It should be noted that the generation of the difference report also includes traceability and auditability functions. Each difference group record field source, path residual and activity change trajectory, ensuring that the field abnormality root cause and change process can be traced in subsequent credit archive optimization, manual intervention or regulatory review.

[0084] S4: Based on the difference report, consistency verification is completed through version number checking and blockchain storage; if it is not solved within a timeout, it is transferred to manual work order processing.

[0085] In the embodiment of the application, the consistency verification through the version number checking and the blockchain storage includes: For sensitive fields in the difference report, the version number in the set of authoritative data sources is checked, and if the version number sequence of the same unified social credit code field in different data source paths is discontinuous or the same version number corresponds to different authoritative field values, it is considered that a version number conflict occurs.

[0086] If a version number conflict occurs, the consistency of the authoritative field value is preferentially compared, and if the authoritative field values are consistent, only the version number record is updated; if the authoritative field values are different, the authoritative field value and the corresponding authoritative data source are written into the blockchain to generate an unalterable difference voucher.

[0087] Further, the updating of the version number record comprises: creating a latest version number record entry for each field, recording the field value, the latest version number, the source data source path and the generation timestamp. The new entry is inserted into the original version number sequence to ensure the sequence is complete, continuous and traceable. The above operation does not modify the actual value of the field, ensures the stability of the consistent credit archive data, and forms a complete version number history record chain, facilitating future audit and abnormal source tracing.

[0088] By setting the timeout threshold of the sensitive field error correction and the blockchain storage, if the consistency verification cannot be completed within the predefined time or there is a version number conflict that cannot be automatically repaired, an artificial work order is automatically generated. The work order includes field information, conflict type, involved data source path, activity and path residual. The work order is sent to the data management personnel or audit personnel through a unified interface to ensure that the exception that cannot be automatically processed can be manually intervened and finally corrected. The artificial work order processing flow comprises work order receiving, field value verification, conflict reason analysis and authoritative field value confirmation. The technician modifies and confirms the abnormal field according to the difference report and historical data record, and returns the modification result.

[0089] The embodiments of the present application also disclose another aspect which comprises one or more processors and a memory.

[0090] The memory is used to store instructions which can be operated, and when executed by the one or more processors, the instructions make the one or more processors execute operations, and the operations comprise the flow of the data quality control method of the unified social credit code based on the traceability technology of the foregoing embodiments, especially Figure 1 the flow of the method shown.

[0091] The embodiments of the present application also disclose another aspect which comprises one or more processors and a memory. Figure 1 the flow of the method shown.

[0092] It should be recognized that the embodiments of the present application can be realized or implemented by computer hardware, a combination of hardware and software, or through computer instructions stored in a non-transitory computer readable memory.

[0093] The method can be implemented in a computer program configured with a computer program by using standard programming techniques, wherein the storage medium configured in this way makes the computer operate in a specific and predefined manner.

[0094] Each program can be implemented in a high level procedural or object oriented programming language to communicate with a computer system, however, the programs can be implemented in assembly or machine language, if desired.

[0095] In any case, the language can be a compiled or interpreted language.

[0096] Also, for this purpose, the program can be run on a programmed special purpose computer.

[0097] The processes described herein (or variants and / or combinations thereof) can be performed under the control of one or more computer systems configured with executable instructions, and can be implemented as code (e.g., executable instructions, one or more computer programs or one or more applications) running on one or more processors, by hardware or combinations thereof. The computer programs include processor-executable instructions that are

[0098] Further, the methods can be implemented in any type of computing platform operably connected to a suitable computing platform, including but not limited to a personal computer, a mini-computer, a mainframe, a workstation, a network or distributed computing environment, a separate or integrated computer platform, or in communication with a charged particle tool or other imaging device.

[0099] Aspects of the present application can be implemented in machine readable code stored on a non-transitory storage medium or device, whether removable or integrated to a computing platform, such as a hard disk, an optical read and / or write storage medium, RAM, ROM, etc., such that it can be read by a programmable computer and used to configure and operate the computer to perform the processes described herein when the storage medium or device is read by the computer.

[0100] Further, the machine readable code, or portions thereof, can be transmitted over a wired or wireless network.

[0101] The application described herein includes these and other different types of non-transitory computer readable storage media when such media include instructions or programs to implement the steps described above in conjunction with a microprocessor or other data processor.

[0102] It should be noted that the above-mentioned embodiments are only used to illustrate the technical solutions of the present application but not to limit the present application, and although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or equivalent replaced without departing from the spirit and scope of the present application, and all of them should be covered in the scope of the claims of the present application.

Claims

1. A unified social credit code data quality control method based on traceability technology, characterized in that, The method comprises the following steps: Collecting a raw data set indexed by a unified social credit code, and calculating an activity degree to form a data stream with activity degree annotation; Building a tree-shaped credit primitive structure by a depth-first algorithm, and determining an authoritative data source by combining a Gaussian integral minimization model; resolving conflicts by dynamic weight distribution to output a consistent credit archive; Based on the consistent credit archive, sequentially performing structured verification, semantic consistency checking and business logic comparison to identify deviations in key fields and generate a difference report; Based on the difference report, completing consistency verification through version number checking and blockchain storage; If not resolved within a certain time, it is transferred to manual work order processing.

2. The traceability technology-based unified social credit code data quality control method of claim 1, characterized in that: The calculation of the activity degree comprises: Extracting the query frequency, update time difference and new data weight factor from each record in the raw data set, and calculating the activity degree according to the following formula: ； ； wherein, is the activity; is the challenge frequency factor, is the challenge frequency normalized to a continuous value in [0, 1]; , 2, 3 is a weight coefficient; is the time decay rate, is the half-life; is the time decay coefficient; is the difference between the current time and the last update time of the data; is the maximum valid time window; If the activity degree is greater than the 80th percentile of the historical activity degree distribution, it is marked as high activity; otherwise, it is marked as low activity.

3. The traceability technology-based unified social credit code data quality control method of claim 2, characterized in that: The tree-shaped credit primitive structure built by the depth-first algorithm comprises: Receiving the data stream and taking the unified social credit code field as the root node of the tree structure, while binding the activity degree value; Taking the key fields in the data stream as input, generating a semantic dependency matrix according to the inherent logical relationship between the key fields, wherein the matrix elements are 1 indicating the existence of parent-child relationship or logical dependency, otherwise 0; Starting from the root node, recursively expanding the key fields marked as dependent relationship in the semantic dependency matrix one by one, and hanging each key field and the corresponding activity degree together to form a child node and child node activity degree, and obtaining the tree-shaped credit primitive structure; The semantic dependency matrix is generated based on a predefined business rule library, which defines the parent-child dependency relationship between key fields.

4. The traceability technology-based unified social credit code data quality control method of claim 3, characterized in that: The determination of the authoritative data source by combining the Gaussian integral minimization model comprises: Extracting the path field and activity degree from each branch path in the tree-shaped credit primitive structure to form a path field set; Calculating the path field deviation for each path field: If the field type is a numerical field, use the arithmetic difference to calculate the absolute value of the difference between the current value of the field and the weighted mean of all candidate values of the corresponding field in the same data source set; If the field type is a date field, calculate the absolute value of the day difference between the current value of the field and the weighted mean of all candidate values of the corresponding field in the same data source set; If the field type is a text field, calculate the field deviation according to the edit distance or string similarity; For all path fields of each path, input each path field deviation into a normal distribution function to obtain a field Gaussian integral value; the mean of the normal distribution function is the mean of the historical value, and the standard deviation is the standard deviation of the historical value; Calculate the path matching score by weighted sum of the field Gaussian integral value and the activity degree in the path; Calculate the path matching score and the weighted sum of the activity degree to obtain the path residual; The data source corresponding to the path with a path residual less than a preset residual threshold is included in the authoritative data source set.

5. The traceability technology-based unified social credit code data quality control method of claim 4, characterized in that: The conflict resolution by dynamic weight distribution to output the consistent credit archive comprises: For the path field with conflict in the tree-shaped credit primitive structure, the path field deviation is obtained; According to the path field deviation and the path residual, the path field weight is proportionally adjusted, and the path weight of the same path field is normalized after the adjustment; The initial value of the path field weight is the activity value; The path field with conflict is weighted and fused using the adjusted path weight, the path field value with the highest weighted score is selected as the authoritative field value, and the corresponding data source is recorded; All path fields after dynamic weight allocation and conflict resolution are combined to generate a consistent credit file, including the authoritative field value, the corresponding authoritative data source and the activity.

6. The traceability technology-based unified social credit code data quality control method of claim 5, characterized in that: The calculation formula for proportionally adjusting the path field weight is: ； wherein, is a path bias ratio, calculated by dividing the path residual by the path field bias, is an initial value for the path field weight, is an adjusted path field weight.

7. The traceability technology-based unified social credit code data quality control method of claim 6, characterized in that: The deviation of the key field includes: Extract each authoritative field value, corresponding path residual and corresponding activity in the consistent credit file to construct a deviation detection vector, and the elements of the deviation detection vector are triplets <authoritative field name, path residual, activity>; The deviation detection vector is divided into deviation levels, the fields with path residuals lower than the first threshold value are marked as low deviation, the fields with path residuals between the first threshold value and the second threshold value are marked as medium deviation, and the fields with path residuals higher than the second threshold value are marked as high deviation; According to the activity and the deviation level, a key field deviation distribution table is generated, the fields with high activity and high deviation are marked as sensitive fields, and the fields with low activity and low deviation are marked as stable fields.

8. The traceability technology-based unified social credit code data quality control method of claim 7, characterized in that: The difference report generated according to the hierarchical decision tree model includes: Based on the key field deviation distribution table, a two-dimensional deviation projection matrix is constructed by deviation level and activity, wherein the rows represent deviation levels and the columns represent activity intervals; The two-dimensional deviation projection matrix is subjected to hierarchical clustering operation, and the authoritative fields with the same deviation level and activity level are divided into the same difference group; Each difference group is assigned a difference label, and a difference report is generated in combination with the historical version record of the authoritative field in the difference group.

9. The traceability technology-based unified social credit code data quality control method of claim 8, characterized in that: The consistency verification completed by version number verification and blockchain storage includes: For the sensitive fields in the difference report, the version number in the authoritative data source set is verified, if the version number sequence of the same unified social credit code field in different data source paths is discontinuous or the same version number corresponds to different authoritative field values, it is considered that the version number conflict occurs; If the version number conflict occurs, the consistency of the authoritative field value is preferentially compared, if the authoritative field values are consistent, only the version number record is updated; if the authoritative field values are different, the authoritative field value and the corresponding authoritative data source are written into the blockchain to generate an unalterable difference certificate.

10. A computer readable medium storing software, characterized in that: The software includes instructions executable by one or more computers, which, through such execution, cause the one or more computers to perform operations, the operations including the flow of the unified social credit code data quality control method based on the traceability technology according to any one of claims 1-9.

Images