A comprehensive information archiving service system based on supply chain data

Abstract

The application discloses a kind of integrated information archiving service systems based on supply chain data, it is related to data archiving field, and multiple supply chain main body information is acquired in advance by data interface, and unified division is carried out according to business category, multiple role attribute and data type, and structured database is built with supply chain main body ID-business category ID-role attribute-data type as composite index;Real-time acquisition enterprise supply chain docking demand, build demand parameter mapping table, generate high demand docking main list;The main body in list is carried out multi-dimensional data quality score, and the main body below threshold triggers review verification task to form quality closed loop, and the main body above threshold carries out fine-grained fragmentation to multiple categories of data in it, and business and security label are calibrated for each fragmentation, and multi-dimensional data label retrieval library is built, and structured review report is automatically generated for the main body reaching optimal threshold, to realize the optimal authorization of combination of automation screening and artificial professional judgment.

Description

Technical Field

[0001] This invention relates to the field of data archiving, specifically a comprehensive information archiving service system based on supply chain data. Background Technology

[0002] With the increasing specialization of global supply chain division of labor and the deepening of inter-enterprise collaboration, supply chain management has evolved from the traditional linear procurement-production-sales model into a complex network structure that includes multi-level suppliers, manufacturers, distributors and end customers. The efficiency of information exchange and data quality among supply chain enterprises directly affect the accuracy of procurement decisions, the stability of production plans and the overall supply chain's ability to withstand risks. Currently, most systems simply store business data within a single enterprise, without pre-classifying it according to supply chain categories, entity roles, and data types at multiple levels and dimensions. This results in a situation where, when an enterprise needs to quickly find high-quality partners under a specific role, it cannot accurately retrieve data through a unified index architecture. Often, it requires manually reviewing multiple reports or cross-system queries, leading to inefficiency. Furthermore, it cannot obtain real-time information on the enterprise's dynamic connection needs. For candidate supply chain entity data, existing technologies lack a multi-dimensional quantitative scoring mechanism and have not set up a linkage process between scoring thresholds and review verification task sheets. Consequently, entities with substandard data quality cannot be automatically identified and guided to rectify the issues, allowing low-quality data to directly enter the decision-making process, increasing cooperation risks.

[0003] This application aims to pre-acquire information on multiple supply chain entities through data interfaces, and uniformly categorize them according to business categories, multiple role attributes, and data types. It constructs a structured database with a composite index of supply chain entity ID, business category ID, role attribute, and data type. Real-time acquisition of enterprise supply chain connection needs is performed, a demand parameter mapping table is built, and a list of high-demand connection entities is generated. Entities on the list are evaluated for multi-dimensional data quality. Entities below a certain threshold trigger review and verification task sheets to form a quality closure loop. Entities above the threshold perform fine-grained segmentation of multi-category data within their entities, and each segment is labeled with business and security tags. A multi-dimensional data tag retrieval library is constructed. Structured audit reports are automatically generated for entities that meet the selection threshold. This achieves a combination of automated screening and professional human judgment for optimal authorization, significantly improving the standardization of supply chain information archiving, demand response speed, data quality controllability, and the intelligence level of partner selection. Summary of the Invention

[0004] The purpose of this invention is to provide a comprehensive information archiving service system based on supply chain data to solve the problems in the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: A comprehensive information archiving service system based on supply chain data includes a supply chain entity information acquisition and pre-division module, a real-time supply chain entity demand matching module, a comprehensive data quality scoring module, a multi-category data segmentation and labeling module, a supply chain entity selection review and authorization module, and an enterprise data center. Several modules are deployed within the enterprise data center. The supply chain entity information acquisition and pre-segmentation module obtains information data of multiple supply chain entities in advance through the data interface, and divides each supply chain entity by business category, multiple role attributes, and data type, and constructs a supply chain information database indexed by supply chain entity ID-business category ID-role attribute-data type; The real-time supply chain entity demand matching module obtains the current enterprise's supply chain matching needs in real time, constructs a demand parameter mapping table, dynamically calculates the demand heat value of eligible supply chain entities, and generates a list of supply chain entities with high demand for matching for the current enterprise. The comprehensive data quality scoring module performs multi-dimensional data quality scoring on the extracted supply chain entities with high demand. For supply chain entities whose scores are below the set scoring threshold, a review and verification task is triggered to conduct data verification. The multi-category data segmentation and labeling module extracts supply chain entities with quality scores greater than a set threshold, performs multi-category data segmentation on their data, and labels each data segment to build a multi-dimensional data label retrieval library. The supply chain entity selection and authorization module extracts all the labeled data fragment indexes from the supply chain entities whose comprehensive quality scores are greater than the set selection threshold, generates a supply chain entity data review report, and conducts manual review and decision-making.

[0006] Further configuration: The pre-division module for acquiring supply chain entity information includes a data platform channel access sub-module, a supply chain entity layer pre-division sub-module, and a supply chain information database construction sub-module. The data platform channel access sub-module connects to the business systems of different supply chain entities that are connected to this enterprise, establishes a data channel between this enterprise and the business systems of multiple supply chain entities, and collects business data of multiple supply chain entities in real time according to a pre-set collection cycle. The business data of multiple supply chain entities includes basic information of the entity enterprise and business flow data. The collected business data is then connected to the pre-division sub-module of the supply chain entity layer. The supply chain entity layer pre-segmentation submodule acquires business data from multiple supply chain entities. Based on the enterprise information corresponding to each supply chain entity, it extracts the industry classification code and product category corresponding to that enterprise information. Supply chain entities with the same industry classification code and product category are initially classified into major categories. For supply chain entities in the same production category within the initial major category classification, corresponding business category tags are predefined. ; Based on the multi-role attributes pre-defined in the enterprise data center, the roles of the supply chain entities in each initial category are defined according to different multi-role attributes. The multi-role attributes include suppliers, manufacturers, distributors, and customers. The defined role attributes of different supply chain entities are summarized, and secondary classification is carried out in the categories of different initial categories according to different defined attributes. The supply chain information database construction submodule acquires the business data of each supply chain entity with defined role attributes within each initial major category. This business data is then categorized according to different data types, including procurement data, production data, warehousing data, logistics data, financial data, and pre-sales and after-sales data. The business data of different supply chain entities is then grouped according to these data types, and the data of each supply chain entity is summarized. Business category tags Define role attributes and data types, and build a structured supply chain information database.

[0007] Further settings: The real-time supply chain entity demand matching module includes a sub-module for interacting with current enterprise supply chain demand parameters and a sub-module for executing multi-condition joint searches. The current enterprise supply chain demand parameter interaction sub-module allows users to manually submit the specific business categories of the supply chain entities to be matched with for the current enterprise. Define role attributes, business time windows, and connection conditions data to construct a demand parameter mapping table for the current enterprise's supply chain entities to be connected, and send the demand parameter mapping table to the structured supply chain information database. The multi-condition joint retrieval execution submodule obtains the current enterprise's supply chain entity demand parameter mapping table, retrieves the structured supply chain information database based on the conditions in the parameter mapping table, pre-locates the supply chain business categories that meet the conditions, extracts the supply chain entities that meet the entity role attribute requirements from the supply chain business categories that meet the conditions, and at the same time screens the extracted supply chain entities based on the docking condition data, extracts the supply chain entities that all meet the above conditions, and marks them as preliminary candidate entities. The system aggregates supply chain entities with preliminary candidate tags, dynamically calculates the demand heat value for each entity, and pre-sets quantitative indicators for demand heat values ​​in the enterprise data center. These indicators include the matching degree between each supply chain entity and the enterprise's connection conditions data, the update frequency of each entity's business data, the number of historical collaborations between each entity and the enterprise, and the historical collaboration amount. Each quantitative indicator is assigned a different weight. The system calculates and extracts the demand heat value for each supply chain entity, and pre-sets a high demand threshold (demand heat threshold) in the enterprise data center. When the demand heat value of an extracted supply chain entity exceeds the set threshold, it is determined to be a high-demand supply chain entity, and the current list of high-demand supply chain entities for the enterprise is output. If the number of selected high-demand supply chain entities in the current enterprise's connection conditions list is greater than the preset maximum limit or less than the preset minimum limit, the enterprise data center will manually modify the current enterprise's connection conditions data.

[0008] Further configuration: The comprehensive data quality scoring module includes a multi-dimensional data review sub-module and an abnormal supply chain entity handling sub-module. The multi-dimensional data review sub-module obtains the list of supply chain entities with high demand for the current enterprise, and obtains the procurement data, production data, warehousing data, logistics data, financial data, and pre-sales and after-sales data of each supply chain entity in the list. It then conducts a completeness, timeliness, and volatility analysis and review for each type of data.

[0009] Further settings: The multi-dimensional data review submodule retrieves the required data fields manually set by the backend for each data type, summarizes them into a set of required data fields for each data type, divides the daily data records for this data type according to timestamps, screens each data record for each data type for missing required data fields, counts the number of missing required data fields in a specific daily data record for this data type, and sets the number of missing required data fields in a specific data record for this data type as [value missing]. The amount of required data fields in the current data record of this type of data. Calculate the missing rate of required data fields in the current data record. , Analyze the missing rate of required data fields in all records of this type of data. Calculate the mean missing rate of required data fields in this type. , ,in, For the number of data records of this type, if If the supply chain entity does not possess this type of data, If the value is less than the set threshold, a warning flag for this data type will be set for the supply chain entity. The average overall missing rate of required data fields for each data type is aggregated for each supply chain entity, and the average overall missing rate of required data fields for each data type is converted into a completeness score for the current data type. ,Right now The integrity score for the current data type ranges from [value missing]. .

[0010] Further settings: The multi-dimensional data review submodule pre-sets a data timeliness assessment window through the enterprise data center, obtains the collection of all records for each type of data within the timeliness assessment window for different supply chains within the list, and sets the data records for a certain type of data within a specific supply chain entity within the timeliness assessment window as follows: , ,in, To record the total, if If the score is less than the set threshold, the timeliness score for this type of data is recorded as 0. The data is then sorted according to its collection time, and a specific data record is extracted. Post-update timestamp Set the real-time evaluation timestamp as Calculate data records Time difference between the current real-time evaluation timestamp ,Right now Calculate the time difference between different types of data for each supply chain entity and the current real-time assessment timestamp, screen the timeliness score of individual records within different data types in the data timeliness assessment window, and set data records. The timeliness score is : in For data recording The corresponding data type has a manually set time-sensitivity decay factor. The time-sensitivity decay factor for each data type is preset by humans. Summarize the timeliness score of each data record within each data type in the data timeliness assessment window, and calculate the average timeliness score of each data record within the data timeliness assessment window for the current data type. This refers to the timeliness score of the current data type. The timeliness scores of different data types within each supply chain entity are normalized within the timeliness assessment window, and the timeliness score range is mapped to... .

[0011] Further settings: The multi-dimensional data review submodule acquires different types of data for each supply chain entity within the list, and extracts a specific data type by pre-setting key numerical fields for each data type through the enterprise data center. The backend manually determines the time window for analyzing data fluctuations of the current type, and extracts the corresponding data types within the supply chain theme within the set time window. The valid values ​​within different predefined key numeric fields are used to construct the data type. The sequence values ​​of different key numeric fields within the code, with data types set. A key numerical field inside sequence values , To record the total number of values, if If the value is less than a set threshold, then the key numerical fields within this data type will be... The volatility score is set to 0. Null values ​​and obvious outliers are filtered from the sequence, and this key numerical field is calculated. The mean of all valid values ​​within the sequence and standard deviation Calculate the current key numerical field Dispersion within a set time window ,Right now: If this key numerical field The mean of all valid values ​​inside If the value is 0, then volatility analysis will not be performed on the key numerical fields of that side. The data will be sent to the enterprise data center for manual review, and the degree of dispersion will be set through the enterprise data center backend. The upper limit value of this key numerical field. The mean of all valid values ​​inside Approaching 0, degree of dispersion If it exceeds the upper limit, the dispersion will be... The maximum value is directly calculated and the current data type is summarized. The degree of dispersion of different key numerical fields within a set time window; Using a linear mapping model, the dispersion of different key numerical fields within a set time window is mapped to a volatility score, and the data type is defined. Internal key numerical fields Volatility score According to the formula: Summarize the current data types The volatility scores of different key numerical fields within a set time window are calculated, and the mean volatility score of different key numerical fields within the set time window is also calculated. That is, the current data type The comprehensive volatility score is obtained by summarizing the comprehensive volatility scores of different data types extracted from each supply chain entity, normalizing them, and mapping the comprehensive volatility score range to... ; The system retrieves the completeness score, timeliness score, and volatility score of different data types for each supply chain entity within the current list of high-demand supply chain partners. Based on the role attributes of each supply chain entity, the system pre-sets completeness, timeliness, and volatility weight thresholds for each data type in the enterprise's data center backend, and defines the data types accordingly. Internal integrity weight threshold is The timeliness weight threshold is The volatility weight threshold is ,in, Get data type Integrity score Timeliness score Volatility score Calculate the overall quality score for this type. : Calculate the data type The comprehensive quality score is obtained by summarizing the comprehensive quality scores of different data types for each supply chain entity within the list. Through the enterprise data center backend, based on the role attributes of different supply chain entities, different comprehensive quality score influence weights are configured for the internal procurement data, production data, warehousing data, logistics data, financial data, and pre-sales and after-sales data of the supply chain entities with different role attributes. The sum of the influence weights of different comprehensive quality scores is 1. A role attribute-data type weight matrix is ​​constructed. The comprehensive quality scores of the current supply chain entity's internal procurement data, production data, warehousing data, logistics data, financial data, and pre-sales and after-sales data, as well as the corresponding configured influence weights, are statistically analyzed to calculate the comprehensive data quality score of the current supply chain entity.

[0012] Further settings: The abnormal supply chain entity handling submodule obtains the comprehensive data quality score of different supply chain entities within the current list of high-demand supply chain entities of the enterprise. The minimum threshold for the comprehensive data quality score is preset through the enterprise data center backend. If the comprehensive data quality score of a certain supply chain entity within the list is less than the minimum threshold, the supply chain entity review process is triggered, and a review task form is generated and sent to the enterprise data center backend. The enterprise data center backend will then manually determine whether the entity meets the supply chain entity standards for demand matching.

[0013] Further configuration: The multi-category data fragmentation and labeling module includes a multi-dimensional data type independent fragmentation sub-module and a data fragmentation semantic tagging sub-module. The multi-dimensional data type independent fragmentation sub-module, through the enterprise data center backend, pre-sets multi-dimensional data fragmentation preset rules based on procurement data, production data, warehousing data, logistics data, financial data, and pre-sales and after-sales data. It performs preliminary data fragmentation on procurement data, production data, warehousing data, logistics data, financial data, and pre-sales and after-sales data according to different fragmentation identifiers. It retrieves the data volume within each preliminary fragment. If the data volume exceeds a set threshold, it further fragments the data according to a set time granularity, generating independent fragmented data packets. Each independent fragmented data packet includes a unique fragmentation identifier. Corresponding supply chain entities Includes the name, corresponding data type, time range, storage path, and data size of each shard; The data sharding semantic labeling submodule uses natural language processing to extract labels from different independent shards within each type of data, generating sharding label vectors. It then aggregates the label vectors of several independent shards under different types of data within each supply chain entity, constructing a multi-dimensional data label retrieval library. This multi-dimensional data label retrieval library partitions the time range of several independent shards under different types of data within each supply chain entity into time ranges, constructing a time index list.

[0014] Further settings: The supply chain entity selection and authorization module pre-sets a comprehensive quality selection scoring threshold. It screens supply chain entities whose comprehensive quality scores within the current list of high-demand supply chain entities exceed the set threshold. It then generates a secondary list of preferred supply chain entities, extracts all marked independent data fragments from this list, and generates a supply chain entity data review report. This report includes information about the supply chain entities. and name, business category labels The data review report is sent to the preset data review department role based on the assigned role type, overall quality score, and quality scores for each dimension of completeness, timeliness, and volatility. Based on human decision-making, a final list of recommended supply chain entities is generated and sorted by recommendation priority.

[0015] Compared with the prior art, the beneficial effects of the present invention are: it aims to obtain information of multiple supply chain entities in advance through data interfaces, and uniformly classify them according to business categories, multiple role attributes and data types, and construct a structured database with supply chain entity ID-business category ID-role attribute-data type as a composite index, thereby laying a data foundation for efficient retrieval; Real-time acquisition of enterprise supply chain connection needs, construction of demand parameter mapping table, dynamic calculation of demand popularity value of eligible entities, and generation of a list of high-demand connection entities; The entities on the list are assessed for data quality across multiple dimensions, including data completeness, timeliness, and volatility. Entities with data below a certain threshold are subject to review and verification tasks to create a quality loop, guiding them to conduct data verification and correction. Fine-grained segmentation of multi-category data within entities exceeding the threshold is performed, and business and security tags are assigned to each segment. A multi-dimensional data tag retrieval library is constructed to achieve refined data management and rapid location. Structured audit reports are automatically generated for entities that reach the selection threshold, realizing the best authorization through a combination of automated screening and professional human judgment. This significantly improves the standardization of supply chain information archiving, the speed of demand response, the controllability of data quality, and the level of intelligence in partner selection. Attached Figure Description

[0016] To make the content of this invention easier to understand, the invention will be further described in detail below with reference to specific embodiments and accompanying drawings.

[0017] Figure 1 This is a schematic diagram of the module structure of a comprehensive information archiving service system based on supply chain data according to the present invention; Figure 2 This is a flowchart illustrating the implementation of a comprehensive information archiving service system based on supply chain data according to the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] Please see Figures 1-2 In this embodiment of the invention, a comprehensive information archiving service system based on supply chain data is provided. The system includes a supply chain entity information acquisition and pre-division module, a real-time supply chain entity demand docking module, a comprehensive data quality scoring module, a multi-category data segmentation and calibration module, a supply chain entity preferential review and authorization module, and an enterprise data center. Several modules are deployed inside the enterprise data center. The supply chain entity information acquisition and pre-segmentation module obtains information data of multiple supply chain entities in advance through the data interface, and divides each supply chain entity by business category, multiple role attributes, and data type, and constructs a supply chain information database indexed by supply chain entity ID-business category ID-role attribute-data type; Further explanation is needed: the pre-division module for obtaining supply chain entity information includes a data platform channel access sub-module, a supply chain entity layer pre-division sub-module, and a supply chain information database construction sub-module. The data middle platform channel access submodule connects to the business systems of different supply chain entities that are connected to this enterprise, establishes a data channel between this enterprise and the business systems of multiple supply chain entities, and collects business data of multiple supply chain entities in real time according to the preset collection cycle. The business data of multiple supply chain entities includes basic information of the entity enterprise and business flow data. The collected business data is connected to the pre-divided submodules of the supply chain entity layer. The supply chain entity layer pre-segmentation submodule acquires business data from multiple supply chain entities. Based on the enterprise information corresponding to each supply chain entity, it extracts the industry classification code and product category corresponding to that enterprise information. Supply chain entities with the same industry classification code and product category are initially classified into major categories. For supply chain entities in the same production category within the initial major category classification, corresponding business category tags are predefined. ; Based on the multi-role attributes pre-defined in the enterprise data center, the roles of the supply chain entities in each initial category are defined according to different multi-role attributes. The multi-role attributes include suppliers, manufacturers, distributors, and customers. The defined role attributes of different supply chain entities are summarized, and secondary classification is carried out in the categories of different initial categories according to different defined attributes. The supply chain information database construction submodule acquires the business data of each supply chain entity with defined role attributes within each initial major category. This business data is then categorized according to different data types, including procurement data, production data, warehousing data, logistics data, financial data, and pre-sales and after-sales data. The business data of different supply chain entities is then grouped according to these data types, and the data of each supply chain entity is summarized. Business category tags Define role attributes and data types, and build a structured supply chain information database.

[0020] The real-time supply chain entity demand matching module obtains the current enterprise's supply chain matching needs in real time, constructs a demand parameter mapping table, dynamically calculates the demand heat value of eligible supply chain entities, and generates a list of supply chain entities with high demand for matching for the current enterprise. The real-time supply chain entity demand matching module includes a sub-module for interacting with current enterprise supply chain demand parameters and a sub-module for executing multi-condition joint search. Further explanation is needed: the current enterprise supply chain demand parameter interaction submodule allows users to manually submit the specific business categories of the supply chain entities that the enterprise needs to connect with. Define role attributes, business time windows, and connection conditions data to construct a demand parameter mapping table for the current enterprise's supply chain entities to be connected, and send the demand parameter mapping table to the structured supply chain information database. Further explanation is needed. The multi-condition joint retrieval execution submodule obtains the current enterprise's supply chain entity demand parameter mapping table. Based on the conditions in the parameter mapping table, it retrieves the structured supply chain information database, pre-locates the supply chain business categories that meet the conditions, extracts the supply chain entities that meet the entity role attribute requirements from the supply chain business categories that meet the conditions, and at the same time, it screens the extracted supply chain entities based on the docking condition data. It then extracts the supply chain entities that all meet the above conditions and marks them as preliminary candidate entities. The system aggregates supply chain entities with preliminary candidate tags, dynamically calculates the demand heat value for each entity, and pre-sets quantitative indicators for demand heat values ​​in the enterprise data center. These indicators include the matching degree between each supply chain entity and the enterprise's connection conditions data, the update frequency of each entity's business data, the number of historical collaborations between each entity and the enterprise, and the historical collaboration amount. Each quantitative indicator is assigned a different weight. The system calculates and extracts the demand heat value for each supply chain entity, and pre-sets a high demand threshold (demand heat threshold) in the enterprise data center. When the demand heat value of an extracted supply chain entity exceeds the set threshold, it is determined to be a high-demand supply chain entity, and the current list of high-demand supply chain entities for the enterprise is output. If the number of selected high-demand supply chain entities in the current enterprise's connection conditions list is greater than the preset maximum limit or less than the preset minimum limit, the enterprise data center will manually modify the current enterprise's connection conditions data.

[0021] The comprehensive data quality scoring module performs multi-dimensional data quality scoring on the extracted supply chain entities with high demand. For supply chain entities whose scores are below the set scoring threshold, a review and verification task is triggered to conduct data verification. The comprehensive data quality scoring module includes a multi-dimensional data review sub-module and an abnormal supply chain entity processing sub-module. The multi-dimensional data review submodule obtains the list of supply chain entities with high demand for the current enterprise, and obtains the procurement data, production data, warehousing data, logistics data, financial data, and pre-sales and after-sales data of each supply chain entity in the list. It then conducts a completeness, timeliness, and volatility analysis and review of each type of data.

[0022] Specifically, the integrity analysis and multi-dimensional data review submodules retrieve the required data fields manually set by the backend for each data type, summarize them into a set of required data fields for each data type, divide the daily data records for that data type according to timestamps, screen each data record for each data type for missing required data fields, count the number of missing required data fields in a specific daily data record for that data type, and set the number of missing required data fields in a specific data record for that data type as a threshold. The amount of required data fields in the current data record of this type of data. Calculate the missing rate of required data fields in the current data record. , Analyze the missing rate of required data fields in all records of this type of data. Calculate the mean missing rate of required data fields in this type. , ,in, For the number of data records of this type, if If the supply chain entity does not possess this type of data, If the value is less than the set threshold, a warning flag for this data type will be set for the supply chain entity. The average overall missing rate of required data fields for each data type is aggregated for each supply chain entity, and the average overall missing rate of required data fields for each data type is converted into a completeness score for the current data type. ,Right now The integrity score for the current data type ranges from [value missing]. .

[0023] Specifically, the timeliness analysis and multi-dimensional data review submodule pre-set a data timeliness assessment window through the enterprise data center. It then obtains the collection of all records for each type of data within the timeliness assessment window for different supply chains within the list. Finally, it defines the data records for a specific type of data within a particular supply chain entity that are present in the timeliness assessment window. , ,in, To record the total, if If the score is less than the set threshold, the timeliness score for this type of data is recorded as 0. The data is then sorted according to its collection time, and a specific data record is extracted. Post-update timestamp Set the real-time evaluation timestamp as Calculate data records Time difference between the current real-time evaluation timestamp ,Right now Calculate the time difference between different types of data for each supply chain entity and the current real-time assessment timestamp, screen the timeliness score of individual records within different data types in the data timeliness assessment window, and set data records. The timeliness score is : in For data recording The corresponding data type has a manually set time-sensitivity decay factor. The time-sensitivity decay factor for each data type is preset by humans. Summarize the timeliness score of each data record within each data type in the data timeliness assessment window, and calculate the average timeliness score of each data record within the data timeliness assessment window for the current data type. This refers to the timeliness score of the current data type. The timeliness scores of different data types within each supply chain entity are normalized within the timeliness assessment window, and the timeliness score range is mapped to... .

[0024] Specifically, the volatility analysis and multi-dimensional data review submodule acquire different types of data for each supply chain entity within the list. It then extracts a specific data type by pre-setting key numerical fields for each data type in the enterprise data center. The backend manually determines the time window for analyzing data fluctuations of the current type, and extracts the corresponding data types within the supply chain theme within the set time window. The valid values ​​within different predefined key numeric fields are used to construct the data type. The sequence values ​​of different key numeric fields within the code, with data types set. A key numerical field inside sequence values , To record the total number of values, if If the value is less than a set threshold, then the key numerical fields within this data type will be... The volatility score is set to 0. Null values ​​and obvious outliers are filtered from the sequence, and this key numerical field is calculated. The mean of all valid values ​​within the sequence and standard deviation Calculate the current key numerical field Dispersion within a set time window ,Right now: If this key numerical field The mean of all valid values ​​inside If the value is 0, then volatility analysis will not be performed on the key numerical fields of that side. The data will be sent to the enterprise data center for manual review, and the degree of dispersion will be set through the enterprise data center backend. The upper limit value of this key numerical field. The mean of all valid values ​​inside Approaching 0, degree of dispersion If it exceeds the upper limit, the dispersion will be... The maximum value is directly calculated and the current data type is summarized. The degree of dispersion of different key numerical fields within a set time window; Using a linear mapping model, the dispersion of different key numerical fields within a set time window is mapped to a volatility score, and the data type is defined. Internal key numerical fields Volatility score According to the formula: Summarize the current data types The volatility scores of different key numerical fields within a set time window are calculated, and the mean volatility score of different key numerical fields within the set time window is also calculated. That is, the current data type The comprehensive volatility score is obtained by summarizing the comprehensive volatility scores of different data types extracted from each supply chain entity, normalizing them, and mapping the comprehensive volatility score range to... ; The system retrieves the completeness score, timeliness score, and volatility score of different data types for each supply chain entity within the current list of high-demand supply chain partners. Based on the role attributes of each supply chain entity, the system pre-sets completeness, timeliness, and volatility weight thresholds for each data type in the enterprise's data center backend, and defines the data types accordingly. Internal integrity weight threshold is The timeliness weight threshold is The volatility weight threshold is ,in, Get data type Integrity score Timeliness score Volatility score Calculate the overall quality score for this type. : Calculate the data type The comprehensive quality score is obtained by summarizing the comprehensive quality scores of different data types for each supply chain entity within the list. Through the enterprise data center backend, based on the role attributes of different supply chain entities, different comprehensive quality score influence weights are configured for the internal procurement data, production data, warehousing data, logistics data, financial data, and pre-sales and after-sales data of the supply chain entities with different role attributes. The sum of the influence weights of different comprehensive quality scores is 1. A role attribute-data type weight matrix is ​​constructed. The comprehensive quality scores of the current supply chain entity's internal procurement data, production data, warehousing data, logistics data, financial data, and pre-sales and after-sales data, as well as the corresponding configured influence weights, are statistically analyzed to calculate the comprehensive data quality score of the current supply chain entity.

[0025] Further explanation is needed. The abnormal supply chain entity processing submodule obtains the comprehensive data quality scores of different supply chain entities within the current list of high-demand supply chain entities for the enterprise. The enterprise data center backend pre-sets a minimum threshold for the comprehensive data quality score. If the comprehensive data quality score of a certain supply chain entity within the list is less than the minimum threshold, the supply chain entity review process is triggered, and a review task sheet is generated and sent to the enterprise data center backend. The enterprise data center backend then manually judges whether the entity meets the supply chain entity standards for demand matching.

[0026] The multi-category data segmentation and labeling module extracts supply chain entities with quality scores greater than a set threshold, performs multi-category data segmentation on their data, and labels each data segment to build a multi-dimensional data label retrieval library. The multi-category data fragmentation and labeling module includes a multi-dimensional data type independent fragmentation sub-module and a data fragmentation semantic labeling sub-module. Further explanation is needed. The multi-dimensional data type independent sharding submodule, through the enterprise data center backend, pre-sets multi-dimensional data sharding rules based on procurement data, production data, warehousing data, logistics data, financial data, and pre-sales and after-sales data. It performs preliminary data sharding on procurement data, production data, warehousing data, logistics data, financial data, and pre-sales and after-sales data according to different sharding identifiers. It then retrieves the data volume within each preliminary shard. If the data volume exceeds a set threshold, it further shards the data according to a set time granularity, generating independent sharded data packets. Each independent sharded data packet includes a unique sharding identifier. Corresponding supply chain entities Includes the name, corresponding data type, time range, storage path, and data size of each shard; The data sharding semantic labeling submodule uses natural language processing to extract labels from different independent shards within each type of data, generating sharding label vectors. It then aggregates the label vectors of several independent shards under different types of data within each supply chain entity, constructing a multi-dimensional data label retrieval library. This multi-dimensional data label retrieval library partitions the time range of several independent shards under different types of data within each supply chain entity into time ranges, constructing a time index list.

[0027] The supply chain entity selection and authorization module extracts all the labeled data fragment indexes from the supply chain entities whose comprehensive quality scores are greater than the set selection threshold, generates a supply chain entity data review report, and conducts manual review and decision-making.

[0028] The supply chain entity selection and authorization module pre-sets a comprehensive quality selection score threshold. It then screens supply chain entities whose comprehensive quality scores within the current list of high-demand supply chain entities exceed the set threshold. This process generates a secondary list of preferred supply chain entities, extracts all labeled independent data fragments from this list, and generates a supply chain entity data review report. The data review report includes information about the supply chain entities. and name, business category labels The data review report is sent to the preset data review department role based on the assigned role type, overall quality score, and quality scores for each dimension of completeness, timeliness, and volatility. Based on human decision-making, a final list of recommended supply chain entities is generated and sorted by recommendation priority.

[0029] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A comprehensive information archiving service system based on supply chain data, characterized in that: The system includes a supply chain entity information acquisition and pre-division module, a real-time supply chain entity demand matching module, a comprehensive data quality scoring module, a multi-category data segmentation and calibration module, a supply chain entity selection review and authorization module, and an enterprise data center. Several modules are deployed within the enterprise data center. The supply chain entity information acquisition and pre-segmentation module obtains information data of multiple supply chain entities in advance through the data interface, and divides each supply chain entity by business category, multiple role attributes, and data type, and constructs a supply chain information database indexed by supply chain entity ID-business category ID-role attribute-data type; The real-time supply chain entity demand matching module obtains the current enterprise's supply chain matching needs in real time, constructs a demand parameter mapping table, dynamically calculates the demand heat value of eligible supply chain entities, and generates a list of supply chain entities with high demand for matching for the current enterprise. The comprehensive data quality scoring module performs multi-dimensional data quality scoring on the extracted supply chain entities with high demand. For supply chain entities whose scores are below the set scoring threshold, a review and verification task is triggered to conduct data verification. The multi-category data segmentation and labeling module extracts supply chain entities with quality scores greater than a set threshold, performs multi-category data segmentation on their data, and labels each data segment to build a multi-dimensional data label retrieval library. The supply chain entity selection and authorization module extracts all the labeled data fragment indexes from the supply chain entities whose comprehensive quality scores are greater than the set selection threshold, generates a supply chain entity data review report, and conducts manual review and decision-making.

2. The comprehensive information archiving service system based on supply chain data according to claim 1, characterized in that... The pre-division module for acquiring supply chain entity information includes a data platform channel access sub-module, a supply chain entity layer pre-division sub-module, and a supply chain information database construction sub-module. The data platform channel access sub-module connects to the business systems of different supply chain entities that are connected to the enterprise, establishes a data channel between the enterprise and the business systems of multiple supply chain entities, and collects business data of multiple supply chain entities in real time according to a pre-set collection cycle. The business data of multiple supply chain entities includes basic information of the entity enterprise and business flow data. The collected business data is then connected to the pre-division sub-module of the supply chain entity layer. The supply chain entity layer pre-segmentation submodule acquires business data from multiple supply chain entities. Based on the enterprise information corresponding to each supply chain entity, it extracts the industry classification code and product category corresponding to that enterprise information. Supply chain entities with the same industry classification code and product category are initially classified into major categories. For supply chain entities in the same production category within the initial major category classification, corresponding business category tags are predefined. ; Based on the multi-role attributes pre-defined in the enterprise data center, the roles of the supply chain entities in each initial category are defined according to different multi-role attributes. The multi-role attributes include suppliers, manufacturers, distributors, and customers. The defined role attributes of different supply chain entities are summarized, and secondary classification is carried out in the categories of different initial categories according to different defined attributes. The supply chain information database construction submodule acquires the business data of each supply chain entity with defined role attributes within each initial major category. This business data is then categorized according to different data types, including procurement data, production data, warehousing data, logistics data, financial data, and pre-sales and after-sales data. The business data of different supply chain entities is then grouped according to these data types, and the data of each supply chain entity is summarized. Business category tags Define role attributes and data types, and build a structured supply chain information database.

3. The comprehensive information archiving service system based on supply chain data according to claim 1, characterized in that... The real-time supply chain entity demand matching module includes a current enterprise supply chain demand parameter interaction submodule and a multi-condition joint search and execution submodule. The current enterprise supply chain demand parameter interaction submodule allows users to manually submit the business categories of the supply chain entities to be matched with for the current enterprise. Define role attributes, business time windows, and connection conditions data to construct a demand parameter mapping table for the current enterprise's supply chain entities to be connected, and send the demand parameter mapping table to the structured supply chain information database. The multi-condition joint retrieval execution submodule obtains the current enterprise's supply chain entity demand parameter mapping table, retrieves the structured supply chain information database based on the conditions in the parameter mapping table, pre-locates the supply chain business categories that meet the conditions, extracts the supply chain entities that meet the entity role attribute requirements from the supply chain business categories that meet the conditions, and at the same time screens the extracted supply chain entities based on the docking condition data, extracts the supply chain entities that all meet the above conditions, and marks them as preliminary candidate entities. The system aggregates supply chain entities with preliminary candidate tags, dynamically calculates the demand heat value for each entity, and pre-sets quantitative indicators for demand heat values ​​in the enterprise data center. These indicators include the matching degree between each supply chain entity and the enterprise's connection conditions data, the update frequency of each entity's business data, the number of historical collaborations between each entity and the enterprise, and the historical collaboration amount. Each quantitative indicator is assigned a different weight. The system calculates and extracts the demand heat value for each supply chain entity, and pre-sets a high demand threshold (demand heat threshold) in the enterprise data center. When the demand heat value of an extracted supply chain entity exceeds the set threshold, it is determined to be a high-demand supply chain entity, and the current list of high-demand supply chain entities for the enterprise is output. If the number of selected high-demand supply chain entities in the current enterprise's connection conditions list is greater than the preset maximum limit or less than the preset minimum limit, the enterprise data center will manually modify the current enterprise's connection conditions data.

4. The comprehensive information archiving service system based on supply chain data according to claim 1, characterized in that... The comprehensive data quality scoring module includes a multi-dimensional data review submodule and an abnormal supply chain entity processing submodule. The multi-dimensional data review submodule obtains the current list of supply chain entities with high demand from the enterprise, and obtains the procurement data, production data, warehousing data, logistics data, financial data, and pre-sales and after-sales data of each supply chain entity in the list. It then performs a completeness, timeliness, and volatility analysis and review on each type of data.

5. A comprehensive information archiving service system based on supply chain data according to claim 4, characterized in that... The multi-dimensional data review submodule obtains the required data fields manually set by the backend for each data type, summarizes them into a set of required data fields for each data type, divides the daily data records of this data type according to timestamps, screens each data record of this data type for whether there are any missing required data fields, counts the number of missing required data fields in a certain daily data record of this data type, and sets the number of missing required data fields in a certain data record of this data type as . The amount of required data fields in the current data record of this type of data. Calculate the missing rate of required data fields in the current data record. , Analyze the missing rate of required data fields in all records of this type of data. Calculate the mean missing rate of required data fields in this type. , ,in, For the number of data records of this type, if If the supply chain entity does not possess this type of data, If the value is less than the set threshold, a warning flag for this data type will be set for the supply chain entity. The average overall missing rate of required data fields for each data type is aggregated for each supply chain entity, and the average overall missing rate of required data fields for each data type is converted into a completeness score for the current data type. ,Right now The integrity score for the current data type ranges from [value missing]. .

6. A comprehensive information archiving service system based on supply chain data according to claim 4, characterized in that... The multi-dimensional data review submodule pre-sets a data timeliness assessment window through the enterprise data center, obtains the collection of all records for each type of data within the timeliness assessment window for different supply chains within the list, and sets the data records for a certain type of data within a certain supply chain entity within the timeliness assessment window as follows: , ,in, To record the total, if If the score is less than the set threshold, the timeliness score for this type of data is recorded as 0. The data is then sorted according to its collection time, and a specific data record is extracted. Post-update timestamp Set the real-time evaluation timestamp as Calculate data records Time difference between the current real-time evaluation timestamp ,Right now Calculate the time difference between different types of data for each supply chain entity and the current real-time assessment timestamp, screen the timeliness score of individual records within different data types in the data timeliness assessment window, and set data records. The timeliness score is : in For data recording The corresponding data type has a manually set time-sensitivity decay factor. The time-sensitivity decay factor for each data type is preset by humans. Summarize the timeliness score of each data record within each data type in the data timeliness assessment window, and calculate the average timeliness score of each data record within the data timeliness assessment window for the current data type. This refers to the timeliness score of the current data type. The timeliness scores of different data types within each supply chain entity are normalized within the timeliness assessment window, and the timeliness score range is mapped to... .

7. A comprehensive information archiving service system based on supply chain data according to claim 4, characterized in that... The multi-dimensional data review submodule acquires different types of data for each supply chain entity within the list, and extracts a specific data type by pre-setting key numerical fields for each type of data through the enterprise data center. The backend manually determines the time window for analyzing data fluctuations of the current type, and extracts the corresponding data types within the supply chain theme within the set time window. The valid values ​​within different predefined key numeric fields are used to construct the data type. The sequence values ​​of different key numeric fields within the code, with data types set. A key numerical field inside sequence values , To record the total number of values, if If the value is less than a set threshold, then the key numerical fields within this data type will be... The volatility score is set to 0. Null values ​​and obvious outliers are filtered from the sequence, and this key numerical field is calculated. The mean of all valid values ​​within the sequence and standard deviation Calculate the current key numerical field Dispersion within a set time window ,Right now: If this key numerical field The mean of all valid values ​​inside If the value is 0, then volatility analysis will not be performed on the key numerical fields of that side. The data will be sent to the enterprise data center for manual review, and the degree of dispersion will be set through the enterprise data center backend. The upper limit value of this key numerical field. The mean of all valid values ​​inside Approaching 0, degree of dispersion If it exceeds the upper limit, the dispersion will be... The maximum value is directly calculated and the current data type is summarized. The degree of dispersion of different key numerical fields within a set time window; Using a linear mapping model, the dispersion of different key numerical fields within a set time window is mapped to a volatility score, and the data type is defined. Internal key numerical fields Volatility score According to the formula: Summarize the current data types The volatility scores of different key numerical fields within a set time window are calculated, and the mean volatility score of different key numerical fields within the set time window is also calculated. That is, the current data type The comprehensive volatility score is obtained by summarizing the comprehensive volatility scores of different data types extracted from each supply chain entity, normalizing them, and mapping the comprehensive volatility score range to... ; The system retrieves the completeness score, timeliness score, and volatility score of different data types for each supply chain entity within the current list of high-demand supply chain partners. Based on the role attributes of each supply chain entity, the system pre-sets completeness, timeliness, and volatility weight thresholds for each data type in the enterprise's data center backend, and defines the data types accordingly. Internal integrity weight threshold is The timeliness weight threshold is The volatility weight threshold is ,in, Get data type Integrity score Timeliness score Volatility score Calculate the overall quality score for this type. : Calculate the data type The comprehensive quality score is obtained by summarizing the comprehensive quality scores of different data types for each supply chain entity within the list. Through the enterprise data center backend, based on the role attributes of different supply chain entities, different comprehensive quality score influence weights are configured for the internal procurement data, production data, warehousing data, logistics data, financial data, and pre-sales and after-sales data of the supply chain entities with different role attributes. The sum of the influence weights of different comprehensive quality scores is 1. A role attribute-data type weight matrix is ​​constructed. The comprehensive quality scores of the current supply chain entity's internal procurement data, production data, warehousing data, logistics data, financial data, and pre-sales and after-sales data, as well as the corresponding configured influence weights, are statistically analyzed to calculate the comprehensive data quality score of the current supply chain entity.

8. A comprehensive information archiving service system based on supply chain data according to claim 1, characterized in that... The abnormal supply chain entity processing submodule obtains the comprehensive data quality scores of different supply chain entities within the current list of high-demand supply chain entities for the enterprise. The minimum threshold for the comprehensive data quality score is preset in the enterprise data center backend. If the comprehensive data quality score of a certain supply chain entity within the list is less than the minimum threshold, the supply chain entity review process is triggered, and a review task order is generated and sent to the enterprise data center backend. The enterprise data center backend then manually judges whether the entity meets the supply chain entity standards for demand matching.

9. A comprehensive information archiving service system based on supply chain data according to claim 1, characterized in that... The multi-category data fragmentation and labeling module includes a multi-dimensional data type independent fragmentation submodule and a data fragmentation semantic tagging submodule. The multi-dimensional data type independent fragmentation submodule, through the enterprise data center backend, pre-sets multi-dimensional data fragmentation rules based on procurement data, production data, warehousing data, logistics data, financial data, and pre-sales and after-sales data. It performs preliminary data fragmentation on procurement data, production data, warehousing data, logistics data, financial data, and pre-sales and after-sales data according to different fragmentation identifiers. It retrieves the data volume within each preliminary fragment; if the data volume exceeds a set threshold, it further fragments the data according to a set time granularity, generating independent fragmented data packets. Each independent fragmented data packet includes a unique fragmentation identifier. Corresponding supply chain entities Includes the name, corresponding data type, time range, storage path, and data size of each shard; The data sharding semantic labeling submodule uses natural language processing to extract labels from different independent shards within each type of data, generating sharding label vectors. It then aggregates the label vectors of several independent shards under different types of data within each supply chain entity, constructing a multi-dimensional data label retrieval library. This multi-dimensional data label retrieval library partitions the time range of several independent shards under different types of data within each supply chain entity into time ranges, constructing a time index list.

10. A comprehensive information archiving service system based on supply chain data according to claim 1, characterized in that... The supply chain entity selection and authorization module pre-sets a comprehensive quality selection scoring threshold. It then screens supply chain entities whose comprehensive quality scores within the current list of high-demand supply chain entities exceed the set threshold. A secondary list of preferred supply chain entities is generated, and all marked independent data fragments within this list are extracted. A supply chain entity data review report is then generated, including information on the supply chain entities. and name, business category labels The data review report is sent to the preset data review department role based on the assigned role type, overall quality score, and quality scores for each dimension of completeness, timeliness, and volatility. Based on human decision-making, a final list of recommended supply chain entities is generated and sorted by recommendation priority.

Images