Automatic matching method and system for delivery evidence and acceptance criteria

By converting acceptance criteria and delivered material attributes into associative data, forming candidate correspondences and clause matching facts, the accuracy problem in the acceptance of heterogeneous delivered materials is solved, enabling clear judgment of the validity of evidence and the coverage status of clauses, and supporting review and location-based supplementary evidence.

CN123114430APending Publication Date: 2026-11-17SHENZHEN ZHIDONG FUTURE TECHNOLOGY CO LTD +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611618901.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-10-20
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing manual inspection or simple keyword comparison methods are insufficient to accurately determine whether heterogeneous delivered materials meet acceptance requirements, leading to the adoption of invalid evidence or difficulty in identifying evidence gaps.

Method used

The acceptance calculation server converts the acceptance criteria clauses, clause logic, and delivered material attributes into associative data, forming candidate correspondences and clause matching facts, and outputs acceptance matching conclusions, including the validity of evidence, the status of clause satisfaction, and the reasons for exclusion.

Benefits of technology

It improves the accuracy of automatic matching between heterogeneous delivered materials and acceptance standards, ensures clear judgment of the validity of evidence and the coverage status of clauses, and supports review and location of supplementary evidence matters.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123114430A_ABST
    Figure CN123114430A_ABST
Patent Text Reader

Abstract

This application discloses a method and system for automatically matching delivery evidence with acceptance criteria. The method includes: acquiring acceptance processing data; determining a set of candidate evidence and clause pairs based on the acceptance processing data; determining a set of clause-matching facts based on the set of candidate evidence and clause pairs; and outputting an acceptance matching conclusion to the review terminal based on the set of clause-matching facts and the acceptance clause diagram. This application is beneficial for improving the accuracy of automatic matching and for locating supplementary evidence items and reviewing acceptance conclusions.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This application relates to the field of digital acceptance and data processing technology, and in particular to a method and system for automatically matching delivery evidence with acceptance criteria. Background Technology

[0002] After a company completes its R&D task, it typically needs to submit code versions, test results, technical documents, images, or videos as proof of delivery. Different acceptance criteria have different requirements regarding the type of materials, their source, applicable versions, generation time, and the content of the proof. The acceptance server needs to retrieve relevant materials from the code repository, testing platform, and file storage devices to form the basis for review.

[0003] Due to the heterogeneity of delivered materials, their diverse sources, and the potential mismatch between material versions and the acceptance target version, existing manual acceptance or simple keyword comparison methods are insufficient to accurately determine whether each material meets its corresponding acceptance requirements. This can easily lead to the adoption of invalid evidence or difficulties in locating evidence gaps. Summary of the Invention

[0004] This application proposes a method and system for automatically matching delivery evidence with acceptance criteria, aiming to improve the accuracy of automatic matching between heterogeneous delivery materials and acceptance criteria, and to provide verifiable and traceable acceptance evidence.

[0005] In a first aspect, embodiments of this application provide an automatic matching method for delivery evidence and acceptance standards, applied to an acceptance calculation server. The method includes: acquiring acceptance processing data, the acceptance processing data including a set of acceptance clause records representing acceptance standard clauses, an acceptance clause diagram representing clause logic, and a set of normative evidence objects representing the attributes of delivery materials; determining a set of candidate evidence and clause pairs based on the acceptance processing data, the set of candidate evidence and clause pairs representing a candidate correspondence between the set of normative evidence objects and the set of acceptance clause records; determining a set of clause matching facts based on the set of candidate evidence and clause pairs, the set of clause matching facts representing the validity of evidence, the status of clause satisfaction, and the reasons for exclusion; and outputting an acceptance matching conclusion to an audit terminal based on the set of clause matching facts and the acceptance clause diagram, the acceptance matching conclusion representing the clause coverage status, evidence identification, and the location of the basis.

[0006] As can be seen, this embodiment first converts the acceptance standard clauses, clause logic, and delivered material attributes into associative data, and then sequentially forms candidate correspondences and clause matching facts, so that the output acceptance matching conclusion can simultaneously reflect the validity of evidence, the coverage status of clauses, and their positioning basis, thereby facilitating the identification of supplementary evidence matters and the review of matching results.

[0007] In one possible example, determining the set of facts matching clauses based on the candidate evidence and clause pair set includes: determining validity verification data based on the candidate evidence and clause pair set, the validity verification data including source integrity verification data and version compatibility verification data; determining source integrity verification results based on the source integrity verification data; determining version compatibility verification results based on the version compatibility verification data; determining an acceptable set of evidence and clause pairs and exclusion reasons based on the source integrity verification results and the version compatibility verification results; and determining the set of facts matching clauses based on the judgment conditions and content characteristics of the acceptable evidence and clause pairs set and the exclusion reasons.

[0008] As can be seen, in this example, the integrity of the source of the evidence and version compatibility are verified separately, and then condition matching is performed on the evidence that passes the verification. This enables the clause matching facts to distinguish between valid supporting relationships and excluded relationships, reducing the risk of incorrect versions or content changes being included in the acceptance process.

[0009] In one possible example, determining the version compatibility verification result based on the version compatibility verification data includes: determining a version verification node group based on the version compatibility verification data, the version verification node group including evidence version nodes, target version nodes, and version relationship records, the evidence version nodes and the target version nodes corresponding to evidence version information and clause version conditions, respectively, and the version relationship records used to connect the two nodes; determining a version association path based on the version verification node group and the version relationship records, the version association path representing the version inheritance relationship or construction association relationship from the evidence version node to the target version node; and determining the version compatibility verification result based on the existence status and relationship type of the version association path.

[0010] As can be seen, in this example, converting the evidence version and the target version into connectable nodes and representing their inheritance or construction relationship through version association paths is helpful in identifying evidence with similar names but not belonging to the target version chain.

[0011] In one possible example, determining the version association path based on the version verification node group and the version relationship record includes: determining a path search range based on the version verification node group and the version relationship record, the path search range being limited by a task identifier and a deliverable identifier; determining a candidate path set based on the path search range and the evidence version node, the candidate path set being formed by connecting version parent-child relationship edges and construction association edges in the version relationship record; determining a destination path set based on the target version node and the candidate path set, the destination path set consisting of candidate paths ending at the target version node; and determining the version association path based on the inheritance direction of the version parent-child relationship edges in the destination path set and the consistency of the construction object of the construction association edges.

[0012] As can be seen, in this example, limiting the path search range first, and then verifying the consistency of the path endpoint, inheritance direction, and constructed object can reduce the interference caused by cross-task version relationships on compatibility judgment.

[0013] In one possible example, determining the candidate path set based on the path search range and the evidence version node includes: determining a path expansion state based on the path search range and the evidence version node, the path expansion state including a set of nodes to be expanded, a set of associated edges to be expanded, access markers, and the current path, the set of associated edges to be expanded including the version parent-child relationship edges and the construction relationship edges; determining a set of expandable associated edges based on unvisited nodes in the set of nodes to be expanded and the set of associated edges to be expanded; determining an updated path expansion state based on the set of expandable associated edges and the path expansion state, the updated path expansion state representing the updated data state of the access markers of the expandable associated edges connected to the current path and subsequent nodes; and determining the candidate path set based on the updated path expansion state iteratively updated to the point where there are no unvisited nodes in the set of nodes to be expanded.

[0014] As can be seen, in this example, by using access tags and path extension states to control the version relationship traversal process, nodes can be prevented from being extended repeatedly in an unordered manner, and candidate paths that can be used for compatibility verification can be fully preserved.

[0015] In one possible example, determining the expandable set of associated edges based on the unvisited nodes in the set of nodes to be expanded and the set of associated edges to be expanded includes: determining the current node based on the unvisited nodes in the set of nodes to be expanded and the set of associated edges to be expanded, wherein the current node is the unvisited node connected to the associated edges in the set of associated edges to be expanded; determining the adjacent set of associated edges based on the current node and the set of associated edges to be expanded; determining the deduplicated set of associated edges based on the associated object identifier, relationship type, and the adjacent set of associated edges; and determining the expandable set of associated edges based on the access marker in the path expansion state and the deduplicated set of associated edges.

[0016] As can be seen, in this example, by filtering adjacency relationships, deduplicating associated edges, and filtering by access tags to compress and expand the range layer by layer, redundant paths caused by duplicate relation edges can be reduced.

[0017] In one possible example, the step of outputting the acceptance matching conclusion to the audit terminal based on the clause matching fact set and the acceptance clause diagram includes: determining an initial coverage state based on the clause matching fact set and the acceptance clause diagram, wherein the initial coverage state characterizes the clause satisfaction status of each record in the acceptance clause record set; determining an evidence anomalous state based on the evidence validity, the exclusion reasons, and the initial coverage state in the clause matching fact set; determining an acceptance clause coverage matrix based on the logical relationships between clauses in the acceptance clause diagram, the initial coverage state, and the evidence anomalous state; and outputting the acceptance matching conclusion to the audit terminal based on the acceptance clause coverage matrix and the clause matching fact set.

[0018] As can be seen in this example, the clause satisfaction status and evidence exception status are summarized into the coverage matrix, enabling reviewers to distinguish between different situations such as no evidence submitted, invalid evidence, and unmet logical dependencies.

[0019] In one possible example, determining the acceptance clause coverage matrix based on the logical relationships between clauses in the acceptance clause diagram, the initial coverage state, and the evidence anomaly state includes: determining clause relationship evaluation units based on the logical relationships between clauses in the acceptance clause diagram, the initial coverage state, and the evidence anomaly state, wherein the clause relationship evaluation unit includes clause nodes with related relationships, logical relationship types, node coverage states, and node anomaly states; determining a set of node states based on the clause relationship evaluation units, wherein the set of node states includes the node coverage states and node anomaly states of the clause nodes; determining a relationship coverage state based on the set of node states and the logical relationship types, wherein the logical relationship types include all-satisfied relationships, at least one-satisfied relationships, or prerequisite dependency relationships; and determining the acceptance clause coverage matrix based on the relationship coverage state and the clause relationship evaluation units.

[0020] As can be seen, in this example, the status of the clause nodes is summarized according to different logical relationship types, so that the combined acceptance criteria can form a clear coverage judgment in a unified matrix.

[0021] In one possible example, after outputting the acceptance matching conclusion, the method further includes: determining acceptance evidence package data based on the acceptance matching conclusion, the acceptance processing data, and the clause matching fact set; determining a packaged acceptance evidence package based on the acceptance evidence package data, the packaged acceptance evidence package including a unique acceptance batch identifier and a content hash; determining an immutable storage record based on the packaged acceptance evidence package and the content hash, the immutable storage record representing the storage location of the packaged acceptance evidence package after it is written to an immutable object storage device; and determining retrieval information for the audit terminal based on the immutable storage record and the unique acceptance batch identifier.

[0022] As can be seen, in this example, the matching conclusion, processing data, and matching facts are encapsulated and saved, so that the audit terminal can use the acceptance batch identifier to retrieve the original evidence and check whether the content of the evidence package has changed based on the content hash.

[0023] Secondly, embodiments of this application provide an automatic matching system for delivery evidence and acceptance standards. The system is configured on an acceptance calculation server and includes: a data acquisition unit for acquiring acceptance processing data, which includes a set of acceptance clause records representing acceptance standard clauses, an acceptance clause diagram representing clause logic, and a set of normative evidence objects representing the attributes of delivery materials; a candidate determination unit for determining a set of candidate evidence and clause pairs based on the acceptance processing data, wherein the set of candidate evidence and clause pairs represents a candidate correspondence between the set of normative evidence objects and the set of acceptance clause records; a fact determination unit for determining a set of clause-matching facts based on the set of candidate evidence and clause pairs, wherein the set of clause-matching facts represents the validity of evidence, the status of clause fulfillment, and reasons for exclusion; and a conclusion output unit for outputting an acceptance matching conclusion to an audit terminal based on the set of clause-matching facts and the acceptance clause diagram, wherein the acceptance matching conclusion represents the clause coverage status, evidence identification, and basis location.

[0024] The system achieves the same or corresponding technical effects as the first aspect method by sequentially transmitting acceptance processing data, candidate evidence and clause pair sets, and clause matching fact sets between each unit.

[0025] Thirdly, embodiments of this application provide an electronic device including a processor, a memory, a communication interface, and one or more programs, the one or more programs being stored in the memory and configured to be executed by the processor, the programs including instructions for performing the steps in the first aspect of embodiments of this application.

[0026] Fourthly, embodiments of this application provide a computer storage medium storing a computer program for electronic data interchange, wherein the computer program causes a computer to perform some or all of the steps described in the first aspect of this embodiment.

[0027] As can be seen, in this embodiment of the application, compared with the existing manual item-by-item verification or delivery material comparison based solely on keywords, this application can perform unified matching processing for heterogeneous, scattered delivery materials that may have version mismatches. The acceptance calculation server of this application first acquires acceptance processing data, which includes a set of acceptance clause records representing acceptance standard clauses, an acceptance clause diagram representing clause logic, and a set of normative evidence objects representing the attributes of delivery materials. Secondly, based on the acceptance processing data, a set of candidate evidence and clause pairs is determined, representing the candidate correspondence between the set of normative evidence objects and the set of acceptance clause records. Then, based on the set of candidate evidence and clause pairs, a set of clause matching facts is determined, representing the validity of evidence, the status of clause fulfillment, and the reasons for exclusion. Finally, based on the set of clause matching facts and the acceptance clause diagram, an acceptance matching conclusion is output to the review terminal, representing the clause coverage status, evidence identification, and the location of the basis. By first unifying heterogeneous delivery materials and acceptance standards into associatable data objects, then limiting the scope of verification by candidate correspondences, and solidifying the validity of evidence, the status of fulfillment of clauses, and the reasons for exclusion into summable matching facts, the status of clause coverage can be traced back to the specific evidence and basis, thereby improving the accuracy of automatic matching and facilitating the location of supplementary evidence and the review of acceptance conclusions. Attached Figure Description

[0028] Figure 1 A schematic diagram of the hardware architecture for automatically matching delivery evidence and acceptance criteria provided in the embodiments of this application; Figure 2 A schematic diagram of the software architecture provided for an embodiment of this application; Figure 3 This is a schematic diagram of the robot R&D results acceptance scenario provided in an embodiment of this application; Figure 4 A flowchart illustrating the automatic matching method for delivery evidence and acceptance criteria provided in the embodiments of this application; Figure 5 This is a schematic diagram illustrating the data association relationship for acceptance processing provided in the embodiments of this application; Figure 6 A flowchart illustrating the method for determining the set of facts for clause matching provided in embodiments of this application; Figure 7 A flowchart illustrating the version compatibility verification method provided in this application embodiment; Figure 8 A flowchart illustrating the version association path determination method provided in this application embodiment; Figure 9 A flowchart illustrating the method for determining the candidate path set provided in this application embodiment; Figure 10 A flowchart illustrating the method for determining the set of scalable associated edges provided in an embodiment of this application; Figure 11 A flowchart illustrating the method for outputting acceptance matching conclusions provided in this application embodiment; Figure 12 A flowchart illustrating the method for determining the acceptance clause coverage matrix provided in this application embodiment; Figure 13 A flowchart illustrating the method for generating acceptance evidence packages provided in this application embodiment; Figure 14 A schematic diagram of the functional modules of the automatic matching system for delivery evidence and acceptance criteria provided in the embodiments of this application. Detailed Implementation

[0029] The present application will be further described below with reference to the accompanying drawings and embodiments. The described embodiments are used to explain the present application and are not intended to unduly limit the scope of protection of the present application. Where there is no conflict, the technical features of the various embodiments can be combined with each other.

[0030] See Figure 1 The hardware system 100 for automatically matching delivery evidence with acceptance criteria includes a deliverer terminal 110, a code repository server 120, a continuous integration or testing server 130, a document and multimedia storage device 140, an acceptance calculation server 150, an evidence metadata database 160, an immutable object storage device 170, an audit terminal 180, a time service device 190, and an identity authentication device 195. The acceptance calculation server 150 obtains code submissions, build results, test metrics, and original files through a read-only interface, writes the evidence metadata to the evidence metadata database 160, and writes the evidence package to the immutable object storage device 170. The audit terminal 180 receives the matching conclusions and retrieval information, while the time service device 190 and the identity authentication device 195 provide unified time and identity data, respectively.

[0031] The acceptance computing server 150 can be an enterprise intranet server, a private cloud computing node, or an authorized cloud server. A read-only interface means that the server can read evidence and records from the source system, but does not modify code branches, test results, or original files through the matching process of this application. This setting isolates the matching process from write operations on the development system.

[0032] See Figure 2 The software architecture 200 is deployed on the acceptance calculation server 150, including an acceptance criterion parsing pipeline 210, an evidence access and content extraction pipeline 220, a source and version verification runtime 230, a candidate matching runtime 240, a clause coverage calculation runtime 250, and a result encapsulation layer 260. The acceptance criterion parsing pipeline 210 generates clause data; the evidence access and content extraction pipeline 220 calls code, test, document, image, or video parsers; the source and version verification runtime 230 verifies the source, content, and version; the candidate matching runtime 240 establishes candidate correspondences between evidence and clauses; the clause coverage calculation runtime 250 summarizes the coverage status; and the result encapsulation layer 260 generates matching conclusions and acceptance evidence packages.

[0033] Each pipeline or runtime can be executed sequentially within the same process, or connected as multiple services by a message queue. Regardless of the deployment method, the data generated in the previous processing stage serves as the input for the next processing stage, and the data association within the same acceptance task is maintained using the acceptance batch identifier.

[0034] See Figure 3 In the robot R&D achievement acceptance scenario 300, the robot motion model R&D task 310 is configured with acceptance criteria 320, code evidence 331, test report 332, and demonstration video 333. The acceptance calculation server 150 reads code evidence 331 from the code repository server 120, test report 332 from the continuous integration or testing server 130, and demonstration video 333 from the document and multimedia storage device 140. If the build number of test report 332 does not correspond to the release version of code evidence 331, the acceptance calculation server 150 retains the coverage status of the code clauses in the acceptance matching conclusion 340, marks the test clauses as version incompatible, and displays a supplementary certification prompt 350 through the audit terminal 180.

[0035] For example, acceptance criterion 320 could require submitting model code from a specified release branch, completing motion generation tests on a specified test suite without joint overruns, and providing a demonstration video showcasing synchronized voice and motion in a bar scene. The location of the demonstration video 333 could indicate the start and end times of the hit segment, the location of the test report 332 could indicate the test item name and report page number, and the location of the code evidence 331 could indicate the repository address, branch, and commit number.

[0036] In one possible example, see Figure 4 The automatic matching method between delivery evidence and acceptance criteria can be achieved through the following steps S101 to S104.

[0037] S101. Obtain acceptance processing data, which includes a set of acceptance clause records representing acceptance standard clauses, an acceptance clause diagram representing clause logic, and a set of normative evidence objects representing the attributes of delivered materials.

[0038] The acceptance clause record is a machine-processable record converted from a natural language acceptance requirement. It can record the applicable task, scope of deliverables, type of evidence, version or time conditions, and judgment criteria. The acceptance clause diagram represents the fulfillment of all, at least one, or prerequisite dependencies among multiple acceptance clauses. The normative evidence object is a data object that uniformly encapsulates the type, source, version, generation time, content summary, content hash, and extractable content of a deliverable. Acceptance processing data is a higher-level summary of the above three types of data.

[0039] S102. Determine a set of candidate evidence and clause pairs based on the acceptance processing data. The set of candidate evidence and clause pairs represents the candidate correspondence between the set of normative evidence objects and the set of acceptance clause records.

[0040] In this context, candidate correspondences indicate that the evidence objects have the potential to meet the corresponding acceptance criteria, but the source and version validity have not yet been confirmed. The acceptance calculation server can perform pre-screening based on task identifier, deliverable scope, and evidence type. For example, a test report may form a candidate correspondence with a test result clause, but it usually does not form a candidate correspondence with a code file submission clause, thereby narrowing the data scope for subsequent verification.

[0041] S103. Determine the set of facts that match the clauses based on the candidate evidence and the set of clause pairs. The set of facts that match the clauses represents the validity of the evidence, the state of clause satisfaction, and the reasons for exclusion.

[0042] Among them, the validity of evidence indicates whether the source, content, and version of the evidence meet the conditions for participation in the matching; the clause satisfaction status indicates whether the valid evidence supports the corresponding acceptance clause; and the reasons for exclusion record situations such as inaccessible source, inconsistent content, unverifiable identity, or incompatible versions. The clause matching facts can be further linked to clause identifiers, evidence identifiers, supporting conditions, degree of support, and location of basis, so that each judgment can be located to the original material.

[0043] S104. Based on the set of facts matched with the clauses and the acceptance clause diagram, output the acceptance matching conclusion to the audit terminal. The acceptance matching conclusion represents the clause coverage status, evidence identification, and basis location.

[0044] The acceptance matching conclusion can include a recommendation to pass, request supplementary documentation, or return. The clause coverage status can include covered, uncovered, evidence abnormal, dependency unmet, or conflicting. In addition to displaying the conclusion, the audit terminal can also open code commits, test items, document paragraphs, or multimedia clips based on evidence identifiers and locations, allowing auditors to review them.

[0045] As can be seen in this example, the acceptance calculation server first forms a unified processing object, then limits the candidate corresponding range and generates interpretable matching facts, and finally outputs a conclusion in combination with the clause logic, so that the validity judgment of heterogeneous evidence and the clause coverage judgment are related, thereby improving the matching accuracy and review efficiency.

[0046] See Figure 5 The acceptance processing data 510 includes an acceptance clause record set 511, an acceptance clause diagram 512, and a set of normative evidence objects 513. The acceptance clause record set 511 and the set of normative evidence objects 513 form a candidate evidence and clause pair set 520. After validity and condition matching, the candidate evidence and clause pair set 520 forms a clause matching fact set 530. The clause matching fact set 530 and the acceptance clause diagram 512 are used together to form an acceptance matching conclusion 540, which is then sent to the review terminal 180.

[0047] The acceptance clause record set 511 describes "what needs to be proven", the normative evidence object set 513 describes "what the existing materials can prove", and the acceptance clause diagram 512 describes "how multiple clauses together constitute the acceptance conditions". The candidate evidence and clause pair set 520 is used to connect the first two types of data, the clause matching fact set 530 is used to store the validity and satisfaction status after connection, and the acceptance matching conclusion 540 is the output after logically summarizing the status of each clause.

[0048] In one possible example, see Figure 6 The process of determining the set of facts matching the clauses based on the candidate evidence and the set of clauses in step S103 can be achieved through the following steps S201 to S205.

[0049] S201. Determine validity verification data for the set based on the candidate evidence and terms. The validity verification data includes source integrity verification data and version compatibility verification data.

[0050] The source integrity verification data may include the source system address, access records, submitter's identity signature, original content hash, and the content hash calculated after rereading the content. Version compatibility verification data may include evidence version identifiers, clause version conditions, version parent-child relationship records, and construction association records. Both types of data are associated with the identifiers of candidate evidence and acceptance clauses to prevent the misuse of verification records between different candidate pairs.

[0051] In one possible implementation of this application, the acceptance calculation server can simultaneously record the reading time, interface return status, and source system object identifier when reading external links, and utilize a time service device to unify the time base of different source systems. This implementation can solve the problem of inconsistent access record times making it difficult to determine whether evidence belongs to the same submission batch, thus improving the stability of source verification.

[0052] S202. Determine the source integrity verification result based on the source integrity verification data.

[0053] The source integrity verification result is used to characterize whether the source system is accessible, whether the submitter's identity is verifiable, and whether the currently read content is consistent with the content submitted. For file evidence, the saved content hash and the reread content hash can be compared; for repository links, the repository object identifier, submitter signature, and access records can be verified; for test results, the build task, execution entity, and report source can be verified.

[0054] S203. Determine the version compatibility verification result based on the version compatibility verification data.

[0055] The version compatibility verification result is used to characterize whether there is an allowed inheritance or build relationship between the evidence version and the target version of the clause. For example, the build number of the test report should be able to be associated with the version node of the submitted code, and the model version marked in the demonstration video should belong to the accepted release version or its allowed successor version.

[0056] S204. Determine the set of acceptable evidence and clauses and the reasons for exclusion based on the source integrity verification result and the version compatibility verification result.

[0057] Specifically, when both the source integrity verification result and the version compatibility verification result meet the requirements of the corresponding acceptance clauses, the corresponding candidate pair can be written into the set of acceptable evidence and clause pairs; otherwise, the reason for exclusion is recorded. The reason for exclusion can be recorded as a combination of reason type and location information, such as "Version incompatibility - Build number B17 not associated with target submission C42".

[0058] S205. Determine the set of facts that match the terms based on the judgment conditions and content characteristics of the set of acceptable evidence and terms, as well as the reasons for exclusion.

[0059] Among them, structured field clauses can compare task identifiers, deliverable identifiers, or version fields; test-type clauses can compare test sets, build numbers, and test results; and document, image, or video clauses can locate corresponding paragraphs, regions, or time segments within the specified materials. The reasons for excluding candidate pairs are also written into the matching facts of the corresponding clauses to distinguish between the two states of no evidence and invalid evidence.

[0060] As can be seen in this example, through the above steps, source integrity and version compatibility respectively form interpretable verification results, and together limit the scope of acceptable evidence before condition matching, so that the fact of clause matching reflects the evidence support situation, while retaining the basis for excluding invalid evidence.

[0061] In one possible example, see Figure 7 The process of determining the version compatibility verification result based on the version compatibility verification data in step S203 can be implemented through the following steps S301 to S303.

[0062] S301. Determine a version verification node group based on the version compatibility verification data. The version verification node group includes evidence version nodes, target version nodes, and version relationship records. The evidence version nodes and the target version nodes correspond to evidence version information and clause version conditions, respectively. The version relationship records are used to connect the two nodes.

[0063] In this context, a version node is a unified representation of a commit number, branch version, release version, build number, or model version. An evidence version node represents the version that generated the current evidence, and a target version node represents the version allowed or required by the acceptance criteria. Version relationship records can originate from code repository commit history, build pipeline records, artifact lists, or model release records.

[0064] S302. Determine the version association path based on the version verification node group and the version relationship record. The version association path represents the version inheritance relationship or construction relationship from the evidence version node to the target version node.

[0065] Version inheritance represents the parent-child commit, branch merge, or release inheritance relationship between two version nodes; build association represents the generation relationship between code versions and build artifacts, test reports, or model artifacts. For example, code commit C42 generates build number B17 through a build task, and test report T08 is generated from build number B17. Therefore, test report T08 can be associated with code commit C42 through the build relationship.

[0066] In one possible implementation of this application, the source and confirmation status of each version relationship record can be saved. For direct records from code repositories or continuous integration platforms, the confirmation status can be set to verified; for relationships inferred from file metadata, the confirmation status can be set to pending review. This implementation can solve the problem of inconsistent reliability of version relationships from different sources, enabling version compatibility results to carry relationship basis.

[0067] S303. Determine the version compatibility verification result based on the existence status and relationship type of the version association path.

[0068] Among them, when there is a version association path that conforms to the allowed relationship type of the terms, it can be determined as version compatible; when there is no path to the target version node, or the path contains an unallowed reverse inheritance relationship or cross-task construction relationship, it can be determined as version incompatible, and the unreached node or abnormal relationship edge is recorded.

[0069] As can be seen in this example, through the above steps, the version information that was originally scattered in the code repository, build platform and file metadata is converted into nodes and relationship paths, which can be used to judge version compatibility from the evidence generation link and reduce misjudgments caused by comparing only version names.

[0070] In one possible example, see Figure 8 The process of determining the version association path based on the version verification node group and version relationship record in step S302 can be implemented through the following steps S401 to S404.

[0071] S401. Determine the path search range based on the version verification node group and the version relationship record. The path search range is limited by the task identifier and the deliverable identifier.

[0072] Task identifiers are used to distinguish different R&D or delivery tasks, while deliverable identifiers are used to distinguish code, models, test reports, or multimedia materials. The path search scope can include only version nodes and version relationship records directly related to the current task and deliverable, avoiding incorrect connections between different tasks with the same branch name or build number.

[0073] In one possible implementation of this application, when a build task generates multiple deliverables simultaneously, the build relationships can be further broken down using the deliverable identifiers in the artifact list. This implementation can solve the problem of mixed artifact version relationships caused by a shared build pipeline, and helps to keep the path search range consistent with the current acceptance criteria.

[0074] S402. Determine a candidate path set based on the path search range and the evidence version node. The candidate path set is formed by connecting the parent-child relationship edges and the association relationship edges in the version relationship record.

[0075] In this context, parent-child relationship edges can represent commit inheritance, branch merging, or release inheritance, while relationship edges can represent code generating artifacts, artifacts generating test results, or models generating demo materials. Candidate paths can contain only one relationship edge or consist of multiple consecutive relationship edges.

[0076] S403. Determine a set of arrival paths based on the target version node and the set of candidate paths, wherein the set of arrival paths consists of candidate paths that terminate at the target version node.

[0077] The set of arrival paths excludes intermediate paths that do not reach the target version node. If the terms and conditions allow subsequent versions of the target version, the target version node may also represent the set of allowed version nodes determined by the terms and conditions, and candidate paths to any of these allowed nodes will be retained as arrival paths.

[0078] S404. Determine the version association path based on the inheritance direction of the version parent-child relationship edge in the arrival path set and the consistency of the construction object of the construction association relationship edge.

[0079] Specifically, inheritance direction is used to exclude paths that treat errors from older versions as successors to newer versions, and build object consistency is used to exclude paths where test reports and code artifacts belong to different build objects. The paths retained after filtering both reach the target version node and meet the requirements of direction and object consistency.

[0080] As can be seen in this example, through the above steps, the path search is restricted to the corresponding task and deliverable, and gradually goes through the verification of the destination, direction and build object, so that the version-related path has clear business boundaries and generation link basis.

[0081] In one possible example, see Figure 9 The process of determining the candidate path set based on the path search range and evidence version nodes in step S402 can be implemented through the following steps S501 to S504.

[0082] S501. Determine the path expansion status based on the path search range and the evidence version node. The path expansion status includes a set of nodes to be expanded, a set of related edges to be expanded, an access marker, and the current path. The set of related edges to be expanded includes the version parent-child relationship edges and the construction relationship edges.

[0083] The system consists of several parts: a set of nodes to be expanded to store nodes whose adjacent relationships still need to be read; a set of edges to be expanded to store usable edge relationships within the path search range; an access marker to record whether a node or edge has been processed; and a current path to record the sequence of relationships from the evidence version node to the current node. Initially, the evidence version node enters the set of nodes to be expanded, forming a current path that contains only that node.

[0084] S502. Determine the expandable associated edge set based on the unvisited nodes in the node set to be expanded and the associated edge set to be expanded.

[0085] The scalable set of associated edges consists of edges that are connected to currently unvisited nodes, have available relationship directions, and have not yet resulted in duplicate visits. For the same current node, version parent-child relationship edges and build relationship edges can coexist, thus supporting continuous searches among code versions, build artifacts, and test results.

[0086] S503. Determine the updated path extension state based on the set of scalable associated edges and the path extension state. The updated path extension state characterizes the updated data state of the access markers of the scalable associated edges connected to the current path and subsequent nodes.

[0087] In this process, each expandable associated edge is connected to the corresponding current path to form a new current path, and the other end of the associated edge becomes a successor node. Unvisited successor nodes are added to the set of nodes to be expanded, and the access markers of processed current nodes and associated edges are updated for the next round of expansion reading.

[0088] In one possible implementation of this application, "node identifier, relationship type sequence, and path endpoint" can be used together as the criteria for deduplication of candidate paths. If different relationship records form the same object link, the path record with higher source credibility can be retained. This implementation can solve the problem of the number of candidate paths being inflated due to duplicate synchronization records, and can reduce the subsequent screening burden while retaining valid generated links.

[0089] S504. Determine the candidate path set based on the updated path expansion state where no unvisited node exists in the set of nodes to be expanded, according to the iterative update.

[0090] In this process, if there are no unvisited nodes in the set of nodes to be expanded, it indicates that the reachable relationships within the path search range have been expanded. The acceptance calculation server aggregates all current paths to form a candidate path set. For version data with circular relationships, the access flag can prevent the same node from being repeatedly added to the set of nodes to be expanded along the same relationship sequence.

[0091] As can be seen in this example, through the above steps, the candidate path search process is encapsulated as an iteratively updatable path expansion state, keeping nodes, associated edges, current paths, and access markers synchronized, which is beneficial for complete search of version relationships and controlling repeated expansion.

[0092] In one possible example, see Figure 10 The process of determining the expandable set of associated edges in step S502 based on the unvisited nodes in the set of nodes to be expanded and the set of associated edges to be expanded can be implemented through the following steps S601 to S604.

[0093] S601. Determine the current node based on the unvisited nodes in the set of nodes to be expanded and the set of associated edges to be expanded, wherein the current node is the unvisited node connected to the associated edges in the set of associated edges to be expanded.

[0094] The current node can be selected according to the order in which it entered the set of nodes to be expanded, or it can prioritize nodes that are closer to the target version node. Regardless of the selection method, the current node is a node that has not yet been visited and has a connected edge. Isolated nodes that cannot be expanded do not need to be included in the subsequent edge filtering.

[0095] S602. Determine the set of adjacent associated edges based on the current node and the set of associated edges to be expanded.

[0096] The set of adjacent edges includes edges that start or end at the current node. The acceptance calculation server can simultaneously read the direction, relationship type, associated object identifier, and relationship source of the relationship edges, providing data for subsequent deduplication and access status filtering.

[0097] S603. Determine the set of deduplicated associated edges based on the associated object identifier, relationship type, and the set of adjacent associated edges.

[0098] When multiple adjacent edges point to the same associated object and have the same relationship type, they can be considered as duplicate relationships, and the edge with the most complete source record or the highest confirmation status should be retained. Edges pointing to the same object but with different relationship types should be retained separately to avoid losing the differences between version inheritance relationships and the construction of relationships.

[0099] S604. Determine the expandable association edge set based on the access marker in the path expansion state and the deduplicated association edge set.

[0100] The access marker is used to exclude associated edges that have already been processed in the same direction. If an associated edge causes the current path to return to a node that has already appeared, the acceptance calculation server can mark the associated edge as non-extendable; the remaining associated edges are written into the extensible associated edge set and provided to step S503.

[0101] As can be seen in this example, by taking the above steps, adjacent related edges are first deduplicated according to the related object and relationship type, and then duplicate extended relationships are excluded by combining access tags. This can reduce redundant paths and avoid invalid traversal caused by circular version relationships.

[0102] In one possible example, see Figure 11 The process of outputting the acceptance matching conclusion to the audit terminal based on the fact set of matching clauses and the acceptance clause diagram can be achieved through the following steps S701 to S704.

[0103] S701. Determine the initial coverage state based on the set of matching facts and the acceptance clause diagram, wherein the initial coverage state characterizes the condition of each record in the set of acceptance clause records satisfying the clause.

[0104] The acceptance calculation server maps the clause identifiers in the clause matching facts to the corresponding clause nodes in the acceptance clause diagram. Clause nodes with valid supporting facts can be marked as covered, clause nodes without supporting facts can be marked as uncovered, and clause nodes with partial supporting content can be left in a pending review state.

[0105] S702. Determine the evidence anomalous status based on the validity of the evidence in the fact set matching the terms, the reason for exclusion, and the initial coverage status.

[0106] The "Abnormal Evidence Status" is used to distinguish the reasons why a clause is not covered. For example, failure to submit relevant materials can indicate missing evidence; submitted materials but whose source cannot be verified can indicate an abnormal source; a test report associated with an incorrect build number can indicate an abnormal version; and different pieces of evidence giving opposite results for the same judgment condition can indicate a conflict of evidence.

[0107] S703. Determine the acceptance clause coverage matrix based on the logical relationships between the clauses in the acceptance clause diagram, the initial coverage state, and the evidence anomaly state.

[0108] The acceptance clause coverage matrix can organize rows according to clause identifiers and record evidence identifiers, initial coverage status, evidence anomaly status, relationship coverage status, and the reason for the status. In the robot R&D achievement acceptance example, code clauses can be recorded as covered, test clauses can be recorded as version anomalies, and demonstration video clauses can be recorded as covered with an attached hit time segment.

[0109] S704. Output the acceptance matching conclusion to the audit terminal based on the acceptance clause coverage matrix and the clause matching fact set.

[0110] Specifically, when all mandatory clauses are covered by valid evidence, a recommendation to pass can be issued; when there are evidentiary gaps that can be eliminated by submitting supplementary materials, a recommendation to supplement evidence can be issued; and when key evidence is invalid, conflicting, or key clauses are not met, a recommendation to return can be issued. Each recommendation is associated with the clause identifier, evidence identifier, and the location of the basis.

[0111] As can be seen in this example, through the above steps, the clause satisfaction status, evidence abnormality status, and clause logic are centralized into the acceptance clause coverage matrix, so that the conclusion output by the audit terminal not only presents the overall recommendations, but also explains the coverage status and abnormality reasons of each clause.

[0112] In one possible example, see Figure 12 The process of determining the acceptance clause coverage matrix in step S703 based on the logical relationship between the clauses, the initial coverage status, and the evidence anomaly status of the acceptance clause diagram can be achieved through the following steps S801 to S804.

[0113] S801. Determine the clause relationship evaluation unit based on the logical relationship between clauses in the acceptance clause diagram, the initial coverage status, and the evidence anomaly status. The clause relationship evaluation unit includes clause nodes with related relationships, logical relationship types, node coverage status, and node anomaly status.

[0114] The clause relationship evaluation unit is an encapsulation of a group of clause nodes and their states that need to be jointly evaluated. A clause node can belong to one evaluation unit, or it can participate in different evaluation units as a preceding node and a subsequent node in a preceding dependency relationship.

[0115] S802. Determine a node status set based on the clause relationship evaluation unit. The node status set includes the node coverage status and the node abnormal status of the clause node.

[0116] The node coverage status indicates whether the corresponding clause is supported by valid evidence, while the node abnormal status indicates that evidence is missing, the source is abnormal, the version is abnormal, or there is an evidence conflict. Writing both types of statuses into the node status set avoids ignoring the validity of evidence based solely on "the existence of evidence."

[0117] S803. Determine the relationship coverage state based on the node state set and the logical relationship type, wherein the logical relationship type includes all satisfied relationships, at least one satisfied relationship, or prerequisite dependency relationship.

[0118] Specifically, for a fully satisfied relation, the relation coverage status is considered satisfied when all necessary nodes in the evaluation unit are satisfied and there are no blocking anomalies; for a relation with at least one satisfied relation, the relation coverage status can be considered satisfied when any allowed node is satisfied and the evidence is valid; for a prerequisite dependency relation, the coverage status of subsequent nodes is evaluated only after the prerequisite node is satisfied. For example, "submitting code" can be used as a prerequisite for "valid test results" to prevent the use of test reports independently of the target code version.

[0119] S804. Determine the acceptance clause coverage matrix based on the relationship coverage status and the clause relationship evaluation unit.

[0120] The acceptance calculation server writes the relationship coverage status of each evaluation unit into the corresponding clause node, and fills the matrix with the clause identifier, evidence identifier, coverage status, abnormal status, and status reason. For clause nodes belonging to multiple evaluation units, the evaluation results under the corresponding relationship type can be saved separately.

[0121] As can be seen in this example, through the above steps, acceptance requirements with different combinations can be converted into a unified relational evaluation unit, and the status can be summarized according to the logical relation type, thereby improving the consistency of complex acceptance criteria coverage judgment.

[0122] In one possible example, see Figure 13 After the acceptance matching conclusion is output, the process of generating and storing the acceptance evidence package can be implemented through the following steps S901 to S904.

[0123] S901. Determine the acceptance evidence package data based on the acceptance matching conclusion, the acceptance processing data, and the clause matching fact set.

[0124] The acceptance evidence package data can include the clause records corresponding to the acceptance batch, clause logic, evidence object summary, candidate correspondence, evidence validity, clause satisfaction status, reasons for exclusion, evidence location information, and final conclusion. For cases with large original file volumes, the evidence package data can save the storage reference and content hash of the original file instead of repeatedly writing the complete file.

[0125] S902. Determine the packaged acceptance evidence package based on the acceptance evidence package data. The packaged acceptance evidence package includes a unique acceptance batch identifier and a content hash.

[0126] The unique acceptance batch identifier is used to distinguish evidence packages formed by different tasks, different submission rounds, or different acceptance times. The content hash is an identifier used to verify the integrity of the packaged content, and can be used to determine whether the content of the evidence package has changed during subsequent readings.

[0127] S903. Determine an immutable storage record based on the encapsulated acceptance evidence package and the content hash. The immutable storage record represents the storage location after the encapsulated acceptance evidence package is written into the immutable object storage device.

[0128] The immutable object storage device retains the corresponding object version and storage location after the evidence package is written, and does not change the existing evidence package through ordinary overwrite. The immutable storage record can include object address, object version, write time, content hash, and storage status, which are used to prove that the currently retrieved object is consistent with the content at the time of encapsulation.

[0129] S904. Determine the retrieval information for the audit terminal based on the immutable storage record and the unique acceptance batch identifier.

[0130] The retrieval information may include the acceptance batch identifier, controlled access address, and object version. After reading the evidence package using the retrieval information, the audit terminal can re-display the relevant clauses, evidence citations, and judgment criteria, and check the integrity of the read content through content hashing.

[0131] As can be seen in this example, through the above steps, the data used for acceptance matching, the facts formed, and the conclusions output are associated and encapsulated, and retained using a unique acceptance batch identifier and immutable storage records, so that the matching results can be verified and reproduced.

[0132] See Figure 14 The delivery evidence and acceptance criteria automatic matching system 1400 is set on the acceptance calculation server 150, including a data acquisition unit 1410, a candidate determination unit 1420, a fact determination unit 1430 and a conclusion output unit 1440.

[0133] The data acquisition unit 1410 is used to acquire acceptance processing data, which includes a set of acceptance clause records representing acceptance standard clauses, an acceptance clause diagram representing clause logic, and a set of normative evidence objects representing the attributes of delivered materials.

[0134] The candidate determination unit 1420 is used to determine a set of candidate evidence and clause pairs based on the acceptance processing data. The set of candidate evidence and clause pairs represents the candidate correspondence between the set of normative evidence objects and the set of acceptance clause records.

[0135] The fact-determining unit 1430 is used to determine a set of facts that match clauses based on the candidate evidence and the set of clauses. The set of facts that match clauses represents the validity of the evidence, the state of clause satisfaction, and the reasons for exclusion.

[0136] The conclusion output unit 1440 is used to output an acceptance matching conclusion to the audit terminal 180 based on the set of facts matching the clauses and the acceptance clause diagram. The acceptance matching conclusion represents the clause coverage status, evidence identification, and basis location.

[0137] The data acquisition unit 1410, candidate determination unit 1420, fact determination unit 1430, and conclusion output unit 1440 can be implemented by software modules in the acceptance calculation server 150, or by multiple services working together. Each unit sequentially transmits acceptance processing data, a set of candidate evidence and clause pairs, and a set of clause matching facts, thereby forming an interpretable acceptance matching conclusion and achieving the same or corresponding technical effects as the aforementioned method embodiments.

[0138] This application also provides a computer storage medium storing a computer program for electronic data interchange, which causes a computer to perform some or all of the steps of any of the methods described in the above method embodiments, wherein the computer includes an electronic device.

[0139] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0140] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0141] In the several embodiments provided in this application, it should be understood that the disclosed apparatus can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of the units described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical or other forms.

[0142] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0143] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0144] If the aforementioned integrated units are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage device (CMD). Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several instructions to cause a computer device (which may be a personal computer, electronic device, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned memory includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0145] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, which may include: a flash drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, etc.

[0146] The embodiments of this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this application. The description of the above embodiments is only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for automatically matching delivery evidence with acceptance criteria, characterized in that, The method, applied to acceptance computing servers, includes: Acquire acceptance processing data, which includes a set of acceptance clause records representing acceptance standard clauses, an acceptance clause diagram representing clause logic, and a set of normative evidence objects representing the attributes of delivered materials. The acceptance clause records in the set of acceptance clause records are machine-processable records converted from a natural language acceptance requirement. The acceptance clause diagram is used to represent the full satisfaction, at least one satisfaction, or prerequisite dependency relationships among multiple acceptance clauses. The normative evidence objects in the set of normative evidence objects are data objects that uniformly encapsulate the type, source, version, generation time, content summary, content hash, and extractable content of a delivered material. Based on the acceptance processing data, a set of candidate evidence and clause pairs is determined. The set of candidate evidence and clause pairs represents the candidate correspondence between the set of normative evidence objects and the set of acceptance clause records. The candidate correspondence indicates that the normative evidence object has the possibility of satisfying the corresponding acceptance clause. Based on the candidate evidence and clause pair set, a clause matching fact set is determined, which represents the validity of the evidence, the clause satisfaction status, and the reasons for exclusion. Based on the set of facts matched with the clauses and the acceptance clause diagram, the acceptance matching conclusion is output to the audit terminal. The acceptance matching conclusion represents the clause coverage status, evidence identification, and basis location.

2. The method according to claim 1, characterized in that, The step of determining the set of facts that match the terms based on the candidate evidence and the set of terms includes: Based on the candidate evidence and terms, validity verification data is determined for the set, including source integrity verification data and version compatibility verification data. The source integrity verification result is determined based on the source integrity verification data. The version compatibility verification result is determined based on the version compatibility verification data. Based on the source integrity verification results and the version compatibility verification results, determine the set of acceptable evidence and terms and the reasons for exclusion; The set of facts matching the terms is determined based on the judgment criteria and content characteristics of the set of acceptable evidence and terms, as well as the reasons for exclusion.

3. The method according to claim 2, characterized in that, Determining the version compatibility verification result based on the version compatibility verification data includes: Based on the version compatibility verification data, a version verification node group is determined. The version verification node group includes evidence version nodes, target version nodes, and version relationship records. The evidence version nodes and the target version nodes correspond to evidence version information and clause version conditions, respectively. The version relationship records are used to connect the two nodes. The version association path is determined based on the version verification node group and the version relationship record. The version association path represents the version inheritance relationship or construction relationship from the evidence version node to the target version node. The version compatibility verification result is determined based on the existence status and relationship type of the version association path.

4. The method according to claim 3, characterized in that, The step of determining the version association path based on the version verification node group and the version relationship record includes: The path search range is determined based on the version verification node group and the version relationship record, and the path search range is limited by the task identifier and the deliverable identifier. A candidate path set is determined based on the path search range and the evidence version node. The candidate path set is formed by connecting the parent-child relationship edges and the association relationship edges in the version relationship record. A set of arrival paths is determined based on the target version node and the set of candidate paths, wherein the set of arrival paths consists of candidate paths that terminate at the target version node; The version association path is determined based on the inheritance direction of the version parent-child relationship edge in the arrival path set and the consistency of the construction object of the construction association edge.

5. The method according to claim 4, characterized in that, The step of determining the candidate path set based on the path search range and the evidence version node includes: The path expansion state is determined based on the path search range and the evidence version node. The path expansion state includes a set of nodes to be expanded, a set of associated edges to be expanded, an access marker, and the current path. The set of associated edges to be expanded includes the version parent-child relationship edges and the construction relationship edges. The set of expandable associated edges is determined based on the unvisited nodes in the set of nodes to be expanded and the set of associated edges to be expanded. The updated path extension state is determined based on the set of scalable associated edges and the path extension state. The updated path extension state characterizes the updated data state of the access markers of the scalable associated edges connected to the current path and subsequent nodes. The candidate path set is determined based on the updated path expansion state, where no unvisited node exists in the set of nodes to be expanded, as determined by iterative updates.

6. The method according to claim 5, characterized in that, The step of determining the expandable set of related edges based on the unvisited nodes in the set of nodes to be expanded and the set of related edges to be expanded includes: The current node is determined based on the unvisited nodes in the set of nodes to be expanded and the set of associated edges to be expanded, wherein the current node is the unvisited node connected to the associated edges in the set of associated edges to be expanded; The set of adjacent associated edges is determined based on the current node and the set of associated edges to be expanded; The set of deduplicated associated edges is determined based on the associated object identifier, relationship type, and the set of adjacent associated edges; The expandable associated edge set is determined based on the access marker in the path expansion state and the deduplicated associated edge set.

7. The method according to any one of claims 1 to 6, characterized in that, The step of outputting the acceptance matching conclusion to the audit terminal based on the fact set of matching terms and the acceptance terms diagram includes: The initial coverage state is determined based on the set of facts matching the terms and the acceptance terms diagram, wherein the initial coverage state characterizes the state in which the terms are satisfied for each record in the set of acceptance terms records; The evidence anomalous status is determined based on the validity of the evidence in the fact set matched with the terms, the reasons for exclusion, and the initial coverage status. The acceptance clause coverage matrix is ​​determined based on the logical relationships between the clauses in the acceptance clause diagram, the initial coverage state, and the evidence anomaly state. The acceptance matching conclusion is output to the audit terminal based on the acceptance clause coverage matrix and the clause matching fact set.

8. The method according to claim 7, characterized in that, The step of determining the acceptance clause coverage matrix based on the logical relationships between the clauses in the acceptance clause diagram, the initial coverage state, and the evidence anomaly state includes: The clause relationship evaluation unit is determined based on the logical relationship between the clauses in the acceptance clause diagram, the initial coverage status, and the evidence anomaly status. The clause relationship evaluation unit includes clause nodes with related relationships, logical relationship type, node coverage status, and node anomaly status. The node status set is determined based on the clause relationship evaluation unit, and the node status set includes the node coverage status and the node abnormal status of the clause node; The relationship coverage state is determined based on the node state set and the logical relationship type, wherein the logical relationship type includes all satisfied relationships, at least one satisfied relationship, or a prerequisite dependency relationship. The acceptance clause coverage matrix is ​​determined based on the relationship coverage status and the clause relationship evaluation unit.

9. The method according to any one of claims 1 to 6, characterized in that, After outputting the acceptance matching conclusion, the method further includes: The acceptance evidence package data is determined based on the acceptance matching conclusion, the acceptance processing data, and the clause matching fact set. Based on the data in the acceptance evidence package, an acceptance evidence package is determined to be packaged, and the packaged acceptance evidence package includes a unique acceptance batch identifier and a content hash. An immutable storage record is determined based on the encapsulated acceptance evidence package and the content hash. The immutable storage record represents the storage location after the encapsulated acceptance evidence package is written into the immutable object storage device. The retrieval information for the audit terminal is determined based on the immutable storage record and the unique acceptance batch identifier.

10. An automatic matching system for delivery evidence and acceptance criteria, characterized in that, For executing the automatic matching method of delivery evidence and acceptance criteria as described in any one of claims 1-9, the system is configured on an acceptance calculation server, comprising: The data acquisition unit is used to acquire acceptance processing data, which includes a set of acceptance clause records representing acceptance standard clauses, an acceptance clause diagram representing clause logic, and a set of normative evidence objects representing the attributes of delivered materials. A candidate determination unit is used to determine a set of candidate evidence and clause pairs based on the acceptance processing data. The set of candidate evidence and clause pairs represents the candidate correspondence between the set of normative evidence objects and the set of acceptance clause records. The fact-determining unit is used to determine a set of facts that match clauses based on the candidate evidence and the set of clauses, wherein the set of facts that match clauses represents the validity of the evidence, the state of clause satisfaction, and the reasons for exclusion. The conclusion output unit is used to output an acceptance matching conclusion to the audit terminal based on the fact set of matching clauses and the acceptance clause diagram. The acceptance matching conclusion represents the clause coverage status, evidence identification, and basis location.