Information tracing management system and method for API service interaction
By constructing local and implicit correlation coefficients to assess the importance of isolated data, the problem of broken traceability chains caused by data silos in API service interactions is solved, ensuring the accuracy and completeness of traceability results.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- GUOJI DIGITAL (QINGDAO) TECHNOLOGY CO LTD
- Filing Date
- 2026-05-13
- Publication Date
- 2026-07-31
AI Technical Summary
During API service interactions, the numerous and complex data sources lead to data silos, resulting in broken traceability chains, loss of contextual information, and difficulties in correlation analysis, thus affecting the accuracy and effectiveness of traceability results.
By collecting isolated data and raw data from isolated systems, local correlation coefficients and implicit correlation coefficients are constructed to assess the importance of isolated data. If the normalized value exceeds a preset threshold, the isolated data is used for tracing the source, ensuring the integrity and accuracy of the information tracing process.
It effectively solves the problem of broken traceability chains caused by data silos, ensuring the accuracy and effectiveness of traceability results. By constructing local and implicit correlation coefficients to assess the importance of isolated data, it ensures the integrity of the information traceability process.
Smart Images
Figure CN122491272A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of data traceability technology, specifically to an information traceability management system and method for API service interaction. Background Technology
[0002] In the contemporary digital economy, APIs (Application Programming Interfaces) have become a core infrastructure connecting different software systems, services, and applications. From internal enterprise system integration to cross-organizational business collaboration, from cloud computing platforms to IoT device interconnection, APIs ubiquitously support the operation of modern information society. However, with the explosive growth in the scale of API interactions and the increasing complexity of data flow, issues such as the credibility of information sources, the traceability of the flow process, and the assurance of data integrity are becoming increasingly prominent. Information traceability management, as a key technical means in the fields of data governance and information security, aims to record and track the complete lifecycle of data generation, dissemination, transformation, and use.
[0003] API service interactions typically involve data exchange across multiple systems, platforms, and terminals. Consequently, information tracing for API service interactions also involves data from multiple sources. When API calls span multiple independent systems or services, if the logs and tracking data of each system are stored in isolated platforms without a unified correlation mechanism, end-to-end complete tracing cannot be achieved, resulting in data silos. This ultimately leads to problems such as broken tracing chains, loss of contextual information, and difficulties in correlation analysis, severely impacting the accuracy and effectiveness of tracing results. Summary of the Invention
[0004] To address the aforementioned technical problems, the purpose of this application is to provide an information traceability management system and method for API service interaction, and the specific technical solution adopted is as follows: In a first aspect, embodiments of this application provide an information traceability management method for API service interactions, the method comprising the following steps: Collect isolated data and raw data within the isolated system during the source tracing process; Obtain a predetermined number of keywords from the original data and the isolated data respectively to obtain keyword sets for the original data and the isolated data; construct numerical data sequences for each field within each predetermined time window in the original data and the isolated data respectively; construct the local correlation coefficient between the isolated data and the original data based on the similarity between the numerical data sequences of the original data and the isolated data, and the differences between the keyword sets. Based on the generation time of each data point, the data in the original data and the isolated data are paired. Based on the generation time difference between all the matching data pairs and the similarity of data changes in the original data and the isolated data, the implicit correlation coefficient between the isolated data and the original data is constructed. Based on the local correlation coefficient and the implicit correlation coefficient, a source tracing importance coefficient for isolated data is constructed. If its normalized value exceeds a preset threshold, the isolated data is used for source tracing.
[0005] In one embodiment, the process of obtaining the keyword set is as follows: Arrange all field identifiers in the original data and isolated data in descending order of frequency, and use the first preset number of terms as keywords in the original data and isolated data respectively; use the sets of keywords in the original data and isolated data as keyword sets in the original data and isolated data respectively.
[0006] In one embodiment, the process of obtaining the numerical data sequence is as follows: The sequences of numerical data under the same field within each time window in the original data and the isolated data are respectively used as the numerical data sequences under each field within each time window in the original data and the isolated data.
[0007] In one embodiment, the process of obtaining the local correlation coefficient is as follows: The correlation coefficient between the numerical data sequence of any field in the original data and the numerical data sequence of any field in the isolated data within each time window is calculated using a correlation algorithm. The distance between the keyword sets of the original data and the isolated data is calculated using a distance metric algorithm and denoted as the first distance. The local correlation coefficient between the isolated data and the original data is determined based on the correlation coefficient and the first distance. The local correlation coefficient is positively correlated with the correlation coefficient and negatively correlated with the first distance.
[0008] In one embodiment, the expression for the local correlation coefficient is: In the formula, A is the local correlation coefficient between the isolated data and the original data; The distance is the first distance; T is the number of time windows involving both the original data and the isolated data. Let I be the numerical correlation between the isolated data and the original data within the t-th time window; I is the total number of numerical data sequences in the isolated data within the t-th time window; and J is the total number of numerical data sequences in the original data within the t-th time window. Let be the correlation coefficient between the i-th numerical data sequence in the isolated data and the j-th numerical data sequence in the original data within the t-th time window; It is a preset minimum positive number.
[0009] In one embodiment, the pairing of the original data with the isolated data based on the generation time of each data item specifically involves: Within each preset time window, for any data in the original data to be paired, calculate the generation time difference between it and each unpaired data in the isolated data, obtain the data corresponding to the minimum generation time difference in the isolated data, and use it and the data to be paired as two data that are matched with each other.
[0010] In one embodiment, the process of obtaining the implicit correlation coefficient is as follows: Calculate the variance of the time difference between the isolated data and all matching pairs of data in the original data. The similarity algorithm is used to calculate the similarity between the data sequences under each field in the original data and the data sequences under each field in the isolated data. Two fields with a similarity greater than a preset similarity threshold are regarded as a group of synonym fields. For any group of synonym fields, the difference between each data value belonging to the original data field and the sum of all data values belonging to the isolated data field is calculated and recorded as the first difference. Based on the variance and the first difference, a latent correlation coefficient is constructed between the isolated data and the original data. The latent correlation coefficient is negatively correlated with the variance and the first difference, respectively.
[0011] In one embodiment, the expression for the implicit association coefficient is: In the formula, D is the implicit correlation coefficient between the isolated data and the original data. Here, P is the normalized value of the variance, and P is the number of sets of synonymous fields between the isolated data and the original data. This represents the total number of data values belonging to the original data in the p-th group of synonym fields. This represents the q-th data value among the fields belonging to the original data in the p-th group of synonym fields. This represents the sum of data values belonging to the isolated data fields within the p-th group of synonym fields. It is a preset minimum positive number.
[0012] In one embodiment, the source importance coefficient is a weighted sum of the local correlation coefficient and the latent correlation coefficient.
[0013] Secondly, embodiments of this application also provide an information traceability management system for API service interaction, including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the steps of any of the methods described above.
[0014] The embodiments of this application have at least the following beneficial effects: This application constructs a local correlation coefficient by comparing the keyword differences between the original data and the isolated data, as well as the correlation of numerical data changes between the original data and the isolated data. This reflects the degree of correlation between the data in the isolated system and the original data, and preliminarily assesses the importance of the isolated data to the tracing process. It analyzes the correlation between the generation time of each data point in the original data and each data point in the isolated data, and, combined with the similarity of data changes in the original data and the isolated data, constructs a latent correlation coefficient. This reflects the degree of potential latent correlation between the isolated data and the original data, further assessing the importance of the isolated data to the tracing process. By combining the above two correlation coefficients to calculate the latent correlation coefficient between the isolated data and the original data, it comprehensively judges the importance of the isolated data in the isolated system to the tracing process. If its normalized value exceeds a preset threshold, it indicates that the isolated data is relatively important to the tracing process, and the isolated data is then used for tracing, thereby ensuring the integrity and accuracy of the information tracing process. This solves the problem of data silos arising from the multiple and complex sources of API service interaction data in the current traditional tracing process, leading to a broken tracing chain, and ensures the accuracy and effectiveness of the tracing results. Attached Figure Description
[0015] To more clearly illustrate the technical solutions and advantages in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0016] Figure 1 A flowchart illustrating the steps of an information traceability management method for API service interaction provided in one embodiment of this application; Figure 2 This is a schematic diagram illustrating the process of obtaining local correlation coefficients. Detailed Implementation
[0017] To further illustrate the technical means and effects adopted by this application to achieve the intended purpose of the invention, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of the information traceability management system and method for API service interaction proposed in this application. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.
[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.
[0019] The following description, in conjunction with the accompanying drawings, details the specific scheme of the information traceability management system and method for API service interaction provided in this application.
[0020] Please see Figure 1 The diagram illustrates a flowchart of an information traceability management method for API service interaction provided in an embodiment of this application. The method includes the following steps: Step S1: Collect isolated data and raw data within the isolated system during the tracing process.
[0021] Due to the numerous interaction objects and processes involved in API service interactions, this solution uses the processes of the supplier system, sales and purchase system, and inventory management system as examples. The information traceability system audits the current API call chain using chain tracing tools such as Jaeger and SkyWalking to identify specific nodes where data cannot flow. When a node fails to flow, metadata management tools are used to compare the interface definitions of different systems to confirm whether there are data structure mismatches or missing data. When deterministic traceability identifiers (such as TraceID and SpanID) are lost or fail to resolve, silo correlation analysis based on data characteristics is initiated to obtain siloed data from the siloed systems. A data adapter is added before the siloed systems to extract the raw data and convert it into a unified format.
[0022] Thus, the original data of the information tracing starting point and the isolated data of the isolated system were obtained.
[0023] Step S2: Obtain a preset number of keywords from the original data and the isolated data respectively to obtain the keyword sets of the original data and the isolated data; construct numerical data sequences for each field within each preset time window in the original data and the isolated data respectively; construct the local correlation coefficient between the isolated data and the original data based on the similarity between the numerical data sequences of the original data and the isolated data, and the differences between the keyword sets.
[0024] In the operation of an information traceability system that interacts with API services, raw data flows through multiple levels of nodes, including supplier systems, sales and procurement systems, and inventory management systems. These systems seamlessly connect through unified interface specifications and data standards. Data entity identifiers, timestamps, attribute fields, and relational references maintain semantic consistency and numerical continuity during transmission, resulting in a high degree of correlation between the data flowing through each system throughout the process. However, when data silos occur, differences in interface definitions or format heterogeneity between isolated systems lead to a break in the semantic mapping between the raw data and the isolated data, causing a shift in numerical references. This significantly reduces the correlation between the isolated data and the raw data. However, isolated data may actually be critical for information traceability; therefore, it is necessary to assess the traceability importance of isolated data.
[0025] Specifically, under normal circumstances, data may undergo some additions, deletions, and modifications during its circulation. However, the additions, deletions, and modifications of local keywords within the data are minimal, such as product names, supplier names, and inherent attribute data. Therefore, data maintains a certain degree of correlation when circulating across different systems. For isolated data within isolated systems, if this data is one of the key data involved in the circulation process, its correlation with the keywords in the original data is higher, and the changes in local data show a higher correlation.
[0026] Based on the above analysis, in order to characterize the correlation between isolated data and original data, the local correlation coefficient can be represented by the keyword consistency characteristics and the similarity of the numerical data change trends between the isolated data and the original data. That is, the higher the degree of keyword consistency between the isolated data and the original data, and the greater the similarity of the numerical data change trends, the higher the importance of the isolated data to the original data, and the more important this part of the data should not be ignored in the source tracing process.
[0027] In a preferred embodiment, to characterize the consistency of keywords between the isolated data and the original data, key-value pairs and structured fields are extracted from both the original and isolated data. The TF-IDF algorithm is used to calculate the weight of each field identifier, and the top N field identifiers with the highest weights in both the original and isolated data are selected as keywords, thereby constructing keyword sets for both data. In this embodiment, N=5. In other embodiments of this application, the implementer can set the value of N according to the actual situation. The TF-IDF algorithm is a well-known technology, and its specific process will not be described in detail.
[0028] It should be noted that this application provides only one field weighting algorithm for field identifier weighting. There are many existing field weighting algorithms, and implementers may also use other field weighting algorithms to obtain the weight of field identifiers. This application does not impose any specific restrictions.
[0029] To characterize the similarity in the changing trends of numerical data between isolated data and original data, firstly, time windows are set; in this embodiment, the length of each time window is 1 hour. In other embodiments of this application, the implementer can set the length of the time window according to the actual situation. Then, for the numerical data in both isolated data and original data, the time series sequence composed of all numerical data under the same field within each time window is taken as the numerical data sequence under that field within that time window, thereby obtaining the numerical data sequences within each time window in both isolated data and original data, and calculating the Pearson correlation coefficient between them. If the lengths of the data sequences are not uniform, the shorter of the two lengths is used as the baseline length; data exceeding the baseline length are not included in the correlation coefficient calculation. The Pearson correlation coefficient is a well-known technique, and the specific process will not be elaborated further.
[0030] It should be noted that this application provides only one correlation algorithm for calculating the correlation coefficient between numerical data sequences. There are many existing correlation algorithms, and implementers may also use other correlation algorithms to calculate the correlation coefficient between numerical data sequences. This application does not impose any specific restrictions.
[0031] Furthermore, in this embodiment, the expression for the local correlation coefficient between the isolated data and the original data is as follows: In the formula, A is the local correlation coefficient between the isolated data and the original data; The edit distance between the keyword sets of the isolated data and the original data is denoted as the first distance; T is the number of time windows involving both the original data and the isolated data. Let I be the numerical correlation between the isolated data and the original data within the t-th time window; I is the total number of numerical data sequences in the isolated data within the t-th time window; and J is the total number of numerical data sequences in the original data within the t-th time window. Let be the Pearson correlation coefficient between the i-th numerical data sequence in the isolated data and the j-th numerical data sequence in the original data within the t-th time window; The value is a preset minimum positive number, used to prevent the denominator from being 0. In this embodiment... The value is 0.1. In other embodiments of this application, the implementer may set the value according to the actual situation. The value of . Among them, the edit distance is a well-known technique, and the specific process will not be described in detail.
[0032] It should be noted that this application provides only one distance measurement algorithm for the distance between keyword sets. There are many existing distance measurement algorithms, and implementers may also use other distance measurement algorithms to calculate the distance between keyword sets. This application does not impose any specific restrictions.
[0033] A represents the local data correlation characteristics between isolated data and the original data. The first part of the relation is... This represents the keyword association between the original data and the isolated data. It is achieved by first identifying keywords in the Chinese text data within both the original and isolated data sets, forming keyword sets, and then analyzing the edit distance between the two sets. Since there may be cases where keywords are completely identical, a very small positive number is added. This is to prevent the denominator from being 0.
[0034] The second part of the relation is This represents the correlation characteristics of data changes between the isolated data and the numerical data values in the original data. It is characterized by calculating the Pearson correlation coefficient between the numerical data sequences in the isolated data and the original data. Similar to the existing principle of calculating the correlation between data changes in data sequences, it reflects the correlation by analyzing the trend of change between data sequences.
[0035] The meaning of this relationship is: the greater the correlation between the keywords of the isolated data and the Chinese text data in the original data, and the higher the correlation between the changes in the numerical data, the higher the correlation between the data in the current isolated system and the original data, and the stronger its importance to the information traceability system.
[0036] Step S3: Pair the data in the original data with the data in the isolated data based on the generation time of each data point. Based on the generation time difference between all the matching data pairs and the similarity of data changes in the original data and the isolated data, construct the implicit correlation coefficient between the isolated data and the original data.
[0037] In actual business processing, the same data in different processes may be abbreviated or have different meanings. This may result in fields with the same meaning being represented differently in different process systems. For example, the supplier name in the supplier system may be "XXX Technology Company", while in the sales and purchase system it may be simplified to "XXX Company". Consequently, when analyzing the keyword association between the original data and isolated data, the correlation between the isolated data and the original data may be misjudged, leading to the conclusion that the isolated data is less important to the traceability process, ultimately affecting the accuracy of the traceability process. Therefore, further analysis is needed.
[0038] Specifically, in a normal business process, raw data undergoes a series of system processing steps after creation before being stored or transmitted. Therefore, the overall time interval for new data generation is similar across different systems. For example, each new business data entry generated in the supplier system flows through the sales and purchase system and the inventory management system after a series of processing steps. At this point, the time intervals for each business data entry generated in both systems are quite similar. However, for isolated systems, if the raw data does not flow through the isolated system during its circulation, the generation time of each data entry in the isolated system differs significantly from the generation time of the raw data. Furthermore, raw data may undergo different processing steps during transmission, resulting in variations in the representation of identical data across different systems. That is, a certain data entry in an isolated system might be a summation of the raw data, leading to a decrease in the correlation between actually related data. For instance, the raw data in the supplier system might record product sales, with the data under this field representing daily sales that change over time. In the inventory management system, however, product sales are the sum of all daily sales in the raw data.
[0039] Based on the above analysis, to characterize the implicit correlation between the original data and the isolated data, the implicit correlation coefficient can be represented by the approximate characteristics of the time difference between the generation of the original data and the isolated data, as well as the numerical characteristics between the data in the synonymous fields. That is, the higher the approximation of the time difference between the generation of each data point between the original data and the isolated data, and the more obvious the numerical characteristic relationship between the data in the synonymous fields, the more implicit the correlation between the isolated data and the original data is, and the stronger the importance of the isolated data to the tracing process.
[0040] In a preferred embodiment, to characterize the approximate time difference between the original data and the isolated data, the original data and the isolated data are first initially filtered using a business type identifier or interface protocol header. Then, in the initially filtered original data and isolated data, the generation time in the inherent attributes of each data item is obtained. Within each time window, each data item in the original data and the isolated data is paired. Specifically, for the first data item in the original data, the absolute value of the difference between its generation time and the generation times of each data item in the isolated data is calculated to obtain the generation time of the isolated data item. The data with the smallest absolute difference is matched with the first data in the original data. For the second data in the original data, the absolute difference between its generation time and the generation times of all other data in the isolated data (excluding the already matched data) is calculated. The data with the smallest absolute difference in the isolated data is then matched with the second data in the original data. This process is repeated, matching each data in the original data with each data in the isolated data one-to-one. Matched data will not appear again in subsequent matches, and redundant data that cannot be matched is discarded.
[0041] Furthermore, the absolute value of the difference between the generation times of each pair of matching data is calculated and denoted as the first time difference. The variance of all first time differences within all time windows in both the isolated data and the original data is calculated and used as the generation time variance between the isolated data and the original data. The ratio of this generation time variance to the time window length is calculated to normalize the generation time variance. It should be noted that implementers may also use other normalization methods to normalize the generation time variance, and this application does not impose specific restrictions.
[0042] To characterize the numerical features between synonymous field data, sequences of data under each field in both the original data and the isolated data are obtained and denoted as field data sequences. Further, the cosine similarity between the field data sequences is calculated. When the cosine similarity between a field data sequence in the original data and a field data sequence in the isolated data is greater than or equal to a preset similarity threshold H, these two fields are considered a set of synonymous fields between the original data and the isolated data; otherwise, they are considered non-synonymous fields. In this embodiment, based on historically known similarities between synonymous field data sequences in the original data and the isolated data, H is set to 0.9. In other embodiments of this application, the implementer can set the value of H according to the actual situation. All data values under each set of synonymous fields are normalized using the maximum-minimum normalization method, and subsequent calculations are performed using the normalized data values.
[0043] Based on the above analysis, the implicit correlation coefficient between isolated data and the original data is constructed, and its expression is as follows: In the formula, D is the implicit correlation coefficient between the isolated data and the original data. This represents the normalized value of the time variance between the isolated data and the original data, where P is the number of groups of synonymous fields between the isolated data and the original data. This represents the total number of data values belonging to the original data in the p-th group of synonym fields. This represents the q-th data value among the fields belonging to the original data in the p-th group of synonym fields. This represents the sum of data values belonging to the isolated data fields within the p-th group of synonym fields. It is a preset minimum positive number.
[0044] It should be noted that if the total number of synonymous fields between the isolated data and the original data is 0, then... Take the smallest positive number, 0.01.
[0045] D represents the implicit correlation between the isolated data and the original data. This represents the approximate time interval between the generation of isolated data and the original data. It is characterized by calculating the generation time interval between each data point in the original data and its corresponding isolated data point, and then determining the variance of this time interval. Furthermore, if there is an actual correlation between the isolated data and the original data, and the isolated system also generates corresponding isolated data points after a fixed actual interval whenever the original data is generated, then the variance of the time interval between the generation of the original data and the isolated data is 0. Therefore, a very small positive number needs to be added. This ensures that the denominator is not zero.
[0046] The second part of the relation is This represents the numerical characteristics between isolated data and the original data under synonymous fields, which is characterized by calculating the differences in data values under synonymous fields between isolated data and the original data.
[0047] The meaning of this relationship is: the higher the approximation of the time interval between each data point between the isolated data and the original data, and the greater the possibility that the data values under the synonymous fields of the isolated data and the original data have a summation relationship, the more likely there is an implicit correlation between the isolated data and the original data, and the higher the importance of the isolated data to the tracing process.
[0048] Step S4: Construct the source importance coefficient of the isolated data based on the local correlation coefficient and the implicit correlation coefficient. If its normalized value exceeds a preset threshold, the isolated data will be used for source tracing.
[0049] The analyses in steps S2 and S3 are combined using the following relationship: In the formula, Z is the source importance coefficient of the isolated data, A is the local correlation coefficient between the isolated data and the original data, and D is the implicit correlation coefficient between the isolated data and the original data. , These are the first weight and the second weight, respectively. In this embodiment , Both are set to 0.5 to add A and D with equal weight. In other embodiments of this application, the implementer can set the weight according to the actual situation. and The value of .
[0050] The stronger the local correlation between isolated data and the original data, and the greater the degree of implicit correlation, the higher the correlation between the isolated data in the isolated system and the original data, the stronger its importance to the tracing process, and the more important the isolated data should not be ignored.
[0051] Furthermore, the source importance coefficient of the isolated data is normalized using the maximum value normalization method. In this embodiment, the maximum value used for normalization is the maximum value of the source importance coefficients corresponding to the isolated data of all isolated systems that appeared in the historical source tracing process. If the currently calculated Z value is greater than the historical maximum value, the normalization result is directly set to 1.
[0052] At this point, the traceability importance coefficient of isolated data in the isolated system can be calculated. When the traceability importance coefficient exceeds a preset threshold K, it indicates that the isolated data plays a crucial role in the traceability process. The isolated data is then applied to the traceability process, and the following steps are performed: The information traceability management system collects data from both the original data source and isolated systems, integrating them into a unified data asset view based on established relationships. Original data serves as a reliable benchmark, while isolated data provides a supplementary perspective, together forming a complete traceability dataset. The lineage graph extends to include isolated data nodes, clearly marking the data flow path from the original system to the isolated system via API transmission or batch synchronization. The graph distinguishes between original and isolated data nodes, marking their synchronization timeliness and consistency status. When initiating a traceability request, the system first determines whether the target data resides in the original system or an isolated system. If it's in an isolated system, the system traces upstream along the lineage relationship to the original data source to obtain the authoritative data version; simultaneously, it searches downstream to determine which downstream applications rely on the isolated data, assessing its impact. The traceability results comprehensively display a comparison view of the original and isolated data, including numerical differences, time delays, and synchronization status. For paths involving isolated data, the system automatically marks the data's timeliness level, indicating the user's assessment of the data's freshness relative to the original data, assisting in judging the reliability of decisions. In this embodiment, the value of K is set to 0.9 based on the range of values for the traceability importance coefficient of important isolated data identified during the historical traceability process. In other embodiments of this application, the implementer may set the value of K according to the actual situation.
[0053] A schematic diagram illustrating the process of obtaining local correlation coefficients is shown below. Figure 2 As shown.
[0054] Based on the same inventive concept as the above methods, this application also provides an information traceability management system for API service interaction, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any one of the above-described methods for information traceability management for API service interaction.
[0055] In summary, this application provides an information traceability management method for API service interactions. By comparing the keyword differences between the original data and isolated data, and the correlation of numerical data changes between the original data and isolated data, a local correlation coefficient is constructed to reflect the degree of correlation between data in the isolated system and the original data, thus preliminarily assessing the importance of isolated data to the traceability process. Furthermore, by analyzing the correlation between the generation time of each data point in the original data and each data point in the isolated data, and combining this with the similarity of data changes in the original data and isolated data, a latent correlation coefficient is constructed to reflect the degree of potential latent correlation between the isolated data and the original data. This study further assessed the importance of isolated data to the tracing process. By combining the two correlation coefficients mentioned above to calculate the implicit correlation coefficient between isolated data and original data, the importance of isolated data in the isolated system to the tracing process was comprehensively judged. If its normalized value exceeds a preset threshold, it indicates that the isolated data is relatively important to the tracing process, and thus the isolated data is used for tracing, thereby ensuring the integrity and accuracy of the information tracing process. This solves the problem of data silos appearing due to the multiple and complex sources of API service interaction data in the current traditional tracing process, which leads to the breakage of the tracing chain, and ensures the accuracy and effectiveness of the tracing results.
[0056] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this application. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.
[0057] The various embodiments in this application are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.
[0058] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the principles of this application should be included within the protection scope of this application.
Claims
1. A method for information provenance management for API service interaction, characterized in that, The method includes the following steps: Collect isolated data and raw data within the isolated system during the source tracing process; Obtain a predetermined number of keywords from the original data and the isolated data respectively to obtain keyword sets for the original data and the isolated data; construct numerical data sequences for each field within each predetermined time window in the original data and the isolated data respectively; construct the local correlation coefficient between the isolated data and the original data based on the similarity between the numerical data sequences of the original data and the isolated data, and the differences between the keyword sets. Based on the generation time of each data point, the data in the original data and the isolated data are paired. Based on the generation time difference between all the matching data pairs and the similarity of data changes in the original data and the isolated data, the implicit correlation coefficient between the isolated data and the original data is constructed. Based on the local correlation coefficient and the implicit correlation coefficient, a source tracing importance coefficient for isolated data is constructed. If its normalized value exceeds a preset threshold, the isolated data is used for source tracing. 2.The API service interaction oriented information provenance management method of claim 1, wherein, The process of obtaining the keyword set is as follows: Arrange all field identifiers in the original data and isolated data in descending order of frequency, and use the first preset number of nouns as keywords in the original data and isolated data respectively; use the sets of keywords in the original data and isolated data as keyword sets in the original data and isolated data respectively.
3. The information traceability management method for API service interaction as described in claim 1, characterized in that, The process of obtaining the numerical data sequence is as follows: The sequences of numerical data under the same field within each time window in the original data and the isolated data are respectively used as the numerical data sequences under each field within each time window in the original data and the isolated data.
4. The information traceability management method for API service interaction as described in claim 1, characterized in that, The process of obtaining the local correlation coefficient is as follows: The correlation coefficient between the numerical data sequence of any field in the original data and the numerical data sequence of any field in the isolated data within each time window is calculated using a correlation algorithm. The distance between the keyword sets of the original data and the isolated data is calculated using a distance metric algorithm and denoted as the first distance. The local correlation coefficient between the isolated data and the original data is determined based on the correlation coefficient and the first distance. The local correlation coefficient is positively correlated with the correlation coefficient and negatively correlated with the first distance.
5. The information traceability management method for API service interaction as described in claim 4, characterized in that, The expression for the local correlation coefficient is: In the formula, A is the local correlation coefficient between the isolated data and the original data; The distance is the first distance; T is the number of time windows involving both the original data and the isolated data. Let be the numerical correlation between the isolated data and the original data within the t-th time window; I is the total number of numerical data sequences in the isolated data within the t-th time window; J is the total number of numerical data sequences in the original data within the t-th time window. Let be the correlation coefficient between the i-th numerical data sequence in the isolated data and the j-th numerical data sequence in the original data within the t-th time window; It is a preset minimum positive number.
6. The information traceability management method for API service interaction as described in claim 1, characterized in that, The process of pairing data in the original data with data in the isolated data based on the generation time of each data item is as follows: Within each preset time window, for any data in the original data to be paired, calculate the generation time difference between it and each unpaired data in the isolated data, obtain the data corresponding to the minimum generation time difference in the isolated data, and use it and the data to be paired as two data that are matched with each other.
7. The information traceability management method for API service interaction as described in claim 1, characterized in that, The process of obtaining the implicit correlation coefficient is as follows: Calculate the variance of the time difference between the isolated data and all matching pairs of data in the original data. The similarity algorithm is used to calculate the similarity between the data sequences under each field in the original data and the data sequences under each field in the isolated data. Two fields with a similarity greater than a preset similarity threshold are regarded as a group of synonym fields. For any group of synonym fields, the difference between each data value belonging to the original data field and the sum of all data values belonging to the isolated data field is calculated and recorded as the first difference. Based on the variance and the first difference, a latent correlation coefficient is constructed between the isolated data and the original data. The latent correlation coefficient is negatively correlated with the variance and the first difference, respectively.
8. The information traceability management method for API service interaction as described in claim 7, characterized in that, The expression for the implicit association coefficient is as follows: In the formula, D is the implicit correlation coefficient between the isolated data and the original data. Here, P is the normalized value of the variance, and P is the number of sets of synonymous fields between the isolated data and the original data. This represents the total number of data values belonging to the original data in the p-th group of synonym fields. This represents the q-th data value among the fields belonging to the original data in the p-th group of synonym fields. This represents the sum of data values belonging to the isolated data fields within the p-th group of synonym fields. It is a preset minimum positive number.
9. The information traceability management method for API service interaction as described in claim 1, characterized in that, The source tracing importance coefficient is the weighted sum of the local correlation coefficient and the latent correlation coefficient.
10. An information traceability management system for API service interaction, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1-9.