A crowdsourcing test report fusion method and system based on knowledge graph
By building a knowledge graph of requirements and test reports and using social technology consistency measurements, the problem of inaccurate results of crowdsourcing test reports is solved, efficient and accurate test reports are achieved, and the efficiency and quality of software testing is improved.
Patent Information
- Application Number
- CN202410583837.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-05-11
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2044-05-11
AI Technical Summary
The fusion results of the crowdsourcing test report are inaccurate, and the clustering results are not highly correlated with the demand documents, resulting in limited efficiency and accuracy of the software testing process.
Using a knowledge graph-based method, by constructing a demand knowledge graph and a test report knowledge graph, combining social technology consistency measurement (STC), selecting test reports with high consistency for integration, and optimizing and integrating test cases, test results and test descriptions. Exclude duplicate or contradictory information and generate a comprehensive test report.
It improves the information utilization rate and quality assurance efficiency in the software testing process, ensures the high quality of the test report and its practical guiding significance for the development process.
Smart Images

Figure CN118331884B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of software testing, and in particular, relates to a crowdsourcing test report fusion method and system based on knowledge graph. Background Art
[0002] Crowdsourcing testing is widely used in the field of mobile application testing because it can provide diverse feedback from different devices, locations, network environments, and user behaviors. Although crowdsourcing testing is popular and effective on commercial testing platforms, processing a large number of test reports, checking and classifying them is a time-consuming and necessary task. In order to improve the efficiency of processing test reports, software engineering researchers have developed a variety of techniques to detect and summarize duplicate information. These techniques are designed to help developers identify duplicate submissions in repositories, mainly focusing on two types of information: text descriptions and execution traces. In addition, traditional issue tracking systems, such as Bugzilla, Jira, and Mantis, provide keyword-based search functions to find similar reports to reduce duplication. Although researchers have proposed several methods (such as those based on image-text similarity, deep learning, and text frequency), these strategies still have certain limitations in practical applications. In particular, test reports mainly provide feedback based on test requirements, and developers often find it difficult to notice defects outside the scope of these requirements. Therefore, when integrating test reports, they should be combined with requirement documents to ensure comprehensiveness and accuracy. Summary of the invention
[0003] In order to solve the above problems, the purpose of the present invention is to provide a crowdsourcing test report fusion technology based on knowledge graph to effectively solve the current problems of inaccurate crowdsourcing test report fusion results and weak correlation between clustering results and requirement documents, thereby helping software developers to improve the efficiency of test report analysis and ensure the accuracy of the software testing process.
[0004] In order to achieve the above technical objectives, the present application provides a crowdsourcing test report fusion method based on knowledge graph, comprising the following steps:
[0005] Based on the initial crowdsourcing test report and the set requirements document, a knowledge graph is constructed through knowledge abstraction and processing, as well as maintenance of a set of mapping units of professional terms.
[0006] Based on the knowledge graph, we obtain the measure of social technical consistency between each test report and the knowledge graph according to the social interaction data between the project team and crowdsourced testers;
[0007] Test reports with higher consistency are selected for fusion. Through the matching and fusion algorithm of the knowledge graph, the test cases, test results and test descriptions are optimized and integrated, and duplicate or contradictory information is eliminated to generate the final comprehensive test report.
[0008] Preferably, in the process of obtaining the initial crowdsourcing test report, test reports from multiple crowdsourcing test platforms are collected to generate an initial crowdsourcing test report, each report containing at least a test case, a test result and a test description, wherein each test report is derived from a diverse network environment, device configuration and user behavior pattern.
[0009] Preferably, in the process of constructing the knowledge graph, natural language processing and machine learning techniques are used to extract entity, relationship and attribute information from requirement documents and test reports, and a knowledge graph is generated through knowledge abstraction and processing, as well as maintenance of a set of mapping units of professional terms.
[0010] Preferably, in the process of generating the knowledge graph, the requirement knowledge graph and the test report knowledge graph are constructed respectively by extracting entity, relationship and attribute information from the requirement document and the test report;
[0011] Based on the matching degree of text content, a deep learning model is used to evaluate the similarity of entities and relationships between the test report knowledge graph and the requirement knowledge graph;
[0012] By analyzing the developer social network, including but not limited to communication discussions, collaborative modification records, and feedback loops between testers and development teams, the consistency and relevance of the test reports are determined, and the structure and content of the knowledge graph are improved.
[0013] Preferably, in the process of improving the structure and content of the knowledge graph, based on the demand knowledge graph and the test report knowledge graph, through equivalent relationship reasoning, different entities describing the same entity in different knowledge graphs are identified and entity alignment is performed to merge and complete the knowledge graphs.
[0014] Preferably, in the process of obtaining the measure of social technical consistency, based on the coverage and accuracy of the test report content, by combining the collaboration model in the developer social network, the degree to which each test report truly reflects the requirements is evaluated to generate the measure of social technical consistency.
[0015] Preferably, in the process of measuring the social technology consistency, the process of measuring the social technology consistency between the test report and the knowledge graph includes the following steps:
[0016] Collaboration requirements analysis: Based on the requirements knowledge graph, analyze the collaboration requirements of each requirement for testers and developers;
[0017] Collaboration activity evaluation: evaluate the actual collaboration activities through developer social networks, where the evaluation criteria include communication frequency and collaboration depth, and how these activities affect the quality and accuracy of test reports;
[0018] Consistency calculation: Compare the collaboration requirements with the collaboration activities to evaluate their consistency. A highly consistent test report means that its content is more likely to accurately reflect software requirements and potential problems.
[0019] Finally, according to the level of the metric STC, the test reports are selected for fusion. The knowledge graph fusion algorithm is used to merge similar test cases and results, exclude duplicate or contradictory information, and generate the final comprehensive test report.
[0020] Preferably, in the process of generating the comprehensive test report, a quality assessment is performed on the generated comprehensive test report, including quantitative analysis of test coverage, accuracy and completeness, so as to ensure the high quality of the test report and its practical guiding significance for the development process.
[0021] Preferably, during the consistency calculation process, a quantitative method is used to calculate the consistency metric between the test report and the requirements, so as to evaluate the coverage, accuracy and contribution of the test report to the software development process.
[0022] The present invention discloses a crowdsourcing test report fusion system based on knowledge graph, comprising:
[0023] The data processing module is used to construct a knowledge graph based on the initial crowdsourcing test report and the set requirements document through knowledge abstraction and processing, as well as the maintenance of the mapping unit set of professional terms;
[0024] The data measurement module is used to obtain the measurement value of the social technical consistency between each test report and the knowledge graph based on the social interaction data between the project team and the crowdsourcing testers;
[0025] The report generation module is used to select test reports with high consistency for fusion. Through the matching and fusion algorithm of the knowledge graph, it optimizes and integrates test cases, test results and test descriptions, eliminates duplicate or contradictory information, and generates the final comprehensive test report.
[0026] The present invention discloses the following technical effects:
[0027] The present invention constructs a requirement knowledge graph and a test report knowledge graph, and combines the social technology consistency metric (STC) to efficiently and accurately integrate crowdsourced test reports, so as to improve the information utilization and quality assurance efficiency in the software testing process. BRIEF DESCRIPTION OF THE DRAWINGS
[0028] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0029] Figure 1 is a schematic diagram of the method flow of the bottom-up method of the present invention;
[0030] Figure 2 It is a schematic diagram of the process of the method described in the present invention. DETAILED DESCRIPTION
[0031] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present application.
[0032] like Figure 1-2 As shown, the present invention designs a crowdsourcing test report fusion method based on the consistency of demand and report knowledge graph, and the method is characterized by comprising the following steps:
[0033] a) Crowdsourcing test report acquisition step: collect test reports from multiple crowdsourcing test platforms. Each report contains at least test cases, test results and test descriptions. The test reports are derived from diverse network environments, device configurations and user behavior patterns.
[0034] b) Knowledge graph construction steps: Use natural language processing and machine learning technology to automatically extract entities, relationships, attributes and other information from requirement documents and test reports to build requirement knowledge graphs and test report knowledge graphs. Further improve the structure and content of the knowledge graph through knowledge abstraction and processing, as well as the maintenance of the mapping unit set of professional terms;
[0035] c) Social Technical Consistency (STC) analysis step: Based on the constructed requirements knowledge graph and test report knowledge graph, as well as the social interaction data between the project team and crowdsourced testers, the social technical consistency metric between each test report and the requirements knowledge graph is calculated to determine the test reports with higher consistency;
[0036] d) Test report fusion step: According to the STC analysis results, select test reports with high consistency for fusion. Use the knowledge graph fusion algorithm to optimize and integrate test cases, test results, and test descriptions, eliminate duplicate or contradictory information, and finally generate a comprehensive and highly accurate test report.
[0037] The consistency analysis step further includes the following process:
[0038] a) Similarity calculation sub-step: Use a deep learning model to evaluate the similarity of entities and relationships between the test report knowledge graph and the requirement knowledge graph, as well as the matching degree based on text content;
[0039] b) Social interaction analysis sub-step: Determine the consistency and relevance of the report by analyzing the developer social network, including but not limited to communication discussions, collaborative modification records, and feedback loops between testers and the development team.
[0040] The test report fusion step further includes the following process:
[0041] a) Comprehensive optimization of test cases and results: Advanced algorithms are used to integrate test cases and test results in test reports with high consistency to ensure the integrity and consistency of information;
[0042] b) Intelligent merging of test descriptions: Use text fusion technology to merge test descriptions in test reports with high consistency to generate a comprehensive test report with high information aggregation and easy to understand;
[0043] c) Quality assessment of comprehensive test reports: Perform quality assessment on the generated comprehensive test reports, including quantitative analysis of test coverage, accuracy, and completeness.
[0044] The steps for building the knowledge graph are further covered:
[0045] a) Requirements document analysis sub-step: Conduct in-depth analysis of the requirements document, automatically extract information such as software functional requirements and performance requirements, and convert this information into entities and relationships in the knowledge graph to form the basic structure of the requirements knowledge graph;
[0046] b) Test report analysis sub-step: Extract key information from the crowdsourced test report, including the execution path of the test case, the status description of the test result, and the defect information found during the test process, and build a test report knowledge graph to ensure the correct correspondence and links between various entities and relationships;
[0047] c) Mapping unit set construction and maintenance sub-step: Establish and maintain the mapping unit set between the requirement knowledge graph and the test report knowledge graph, use equivalence relationship reasoning and embedding-based methods to perform entity alignment, and ensure the correct matching of the same or similar entities in the two knowledge graphs.
[0048] The STC measurement steps include the following processes:
[0049] a) Collaboration requirements analysis sub-step: Analyze the collaboration requirements of each requirement based on the requirements knowledge graph, and identify the requirements that require close collaboration between different roles such as front-end and back-end developers and UI designers;
[0050] b) Collaboration activity evaluation sub-step: Using developer social network data, evaluate the actual collaboration activities for specific needs, such as communication frequency, collaboration depth (degree of participation in the collaboration process), etc., and how these activities affect the quality and accuracy of test reports;
[0051] c) Consistency calculation sub-step: Combined with the results of collaborative requirements analysis and collaborative activity evaluation, a quantitative method is used to calculate the consistency metric between the test report and the requirements, thereby evaluating the test report's coverage, accuracy, and contribution to the software development process.
[0052] The further implementation of the test report fusion step is as follows:
[0053] a) Fusion algorithm selection and application sub-step: According to the complexity and diversity of the test report content, select a suitable fusion algorithm, such as a rule-based method, a statistical method, or a deep learning model, to achieve efficient and accurate test report fusion;
[0054] b) Re-optimization sub-step of the test report after fusion: re-optimize the comprehensive test report generated by fusion, including semantic optimization, information redundancy removal and structural optimization, to ensure the readability and practicality of the comprehensive test report;
[0055] c) Test report fusion effect evaluation sub-step: A professional review team and automated tools are used to evaluate the effectiveness of the integrated test report to ensure its effectiveness and guiding value in the actual development environment.
[0056] Embodiment 1: The present invention designs a crowdsourcing test report fusion method based on the consistency of demand and report knowledge graphs. The method first generates a demand knowledge graph for structured text in the demand report, and generates a report knowledge graph for unstructured text in the test report. In this process, the structural features, language features, and metadata features of the crowdsourcing test field report text are used to complete the ontology matching and entity alignment subtasks. Then, based on the mapping unit set, the equivalent relationship reasoning is used to obtain the part of the report knowledge graph that is similar to the demand knowledge graph entity as the basis for the fusion of the crowdsourcing test report. In addition, the social interaction between the project development team and the crowdsourcing testers, such as communication and discussion, collaborative modification, etc., is analyzed to build a developer social network. The network reflects the collaboration mode and communication efficiency between developers and testers. And based on the demand knowledge graph, the test report knowledge graph and the developer social network, the consistency measure (STC) between the test report and the demand is calculated. The level of STC reflects the accuracy and completeness of the test report on demand coverage and feedback. Finally, according to the STC measurement result, the test report with high consistency is selected for fusion to generate a comprehensive test report. During the fusion process, the knowledge graph matching and fusion algorithms are used to integrate test cases and test results to ensure the integrity and accuracy of the information. In this way, the fusion of test reports with variable heat levels can be achieved, and a specific defect report set pointing to the software requirements can be generated.
[0057] The method proposed by the present invention specifically includes the following process:
[0058] 1. Construction of demand and report knowledge graph:
[0059] In the initial stage, it is necessary to build the demand knowledge graph G1 = {A1, B1, C1} and the report knowledge graph G2 = {A2, B2, C2}. In the formal definition of the knowledge graph, A, B and C represent the entities (Entities), relations (Relations), and facts (Facts) in each knowledge graph respectively.
[0060] This patent uses a bottom-up construction method to extract entities, attributes and relationships from the requirements report and test report, and add them to the data layer of the knowledge graph; then these knowledge elements are summarized and organized, gradually abstracted into concepts, and finally form a model layer. The process of the bottom-up method is as follows Figure 1 shown.
[0061] 2. Knowledge extraction:
[0062] Knowledge extraction, similar to ontology learning, uses machine learning technology to automatically or semi-automatically extract entities, relationships, attributes and other elements of the knowledge graph from requirements and report documents. Knowledge extraction includes entity extraction, relationship extraction and attribute extraction. Entity extraction automatically discovers specific entities such as defect names, defect generation methods, and defect generation locations. It is also called named entity learning or named entity recognition. Relationship extraction refers to the use of methods and techniques from disciplines such as linguistics, statistics, and information science to discover semantic relationships between entities from text. Attribute extraction is for entities, and the problem of extracting entity attributes can be transformed into a relationship extraction problem.
[0063] 3. Knowledge Fusion:
[0064] Knowledge fusion is similar to ontology integration. Since the data sources used by knowledge graphs for knowledge extraction are diverse, problems such as knowledge duplication and unclear relationships between knowledge may occur. Knowledge fusion can eliminate the ambiguity between entities, relationships, attributes and other referents and factual objects, so that knowledge from different sources can be standardized and integrated. Knowledge fusion is divided into:
[0065] (1) Entity alignment: It can be used to determine whether multiple entities in the same or different data sets refer to the same entity in the objective world, solving the problem of one entity corresponding to multiple names.
[0066] (2) Attribute value filling: When different values appear for the same attribute, a decision is made based on the number and reliability of data sources to provide a more accurate attribute value.
[0067] 4. Knowledge processing:
[0068] Abstract the concepts of the constructed data layer, that is, construct the model layer of the knowledge graph. Knowledge processing includes ontology construction and quality assessment. The knowledge base formed based on ontology not only has a strong hierarchical structure, but also has a low degree of redundancy. Due to technical limitations, the obtained knowledge elements may be wrong, so an evaluation process is required before adding knowledge to the knowledge base. The accuracy of the knowledge base is ensured by quantifying the credibility of existing knowledge and retaining knowledge with high confidence.
[0069] 5. Correspondence between professional terms in the knowledge base and text similarity:
[0070] After the knowledge graph is built, since the requirements document is a guiding and standard document, its content is relatively abstract and unclear. The task of this section is to design a knowledge base, parse the entities in the requirements knowledge graph, and find the entities in the knowledge base that may appear in the report knowledge graph in the future. When such a correspondence is found, the relationship and facts of the entity are used to expand the requirements knowledge graph.
[0071] Pre-analyze the keywords contained in future defect reports from the requirements, and add entities and relationships to the requirements knowledge graph as a supplement.
[0072] 6. Construction and maintenance of mapping unit collection:
[0073] Initialize a triple set W = <G1, G2, M>, including a source ontology G1, a target ontology G2, and a mapping unit set M{m1, m2, ..., m i}. Among them, m i Represents a basic mapping unit, which can be written as a four-tuple m i =<a,b,c,d> , where a is the identifier of the mapping unit, which is used to uniquely identify the quadruple, b and c are elements in G1 and G2 respectively, and d represents the similarity between G1 and G2, satisfying d∈[0,1].
[0074] 7. Entity alignment under equivalence relation reasoning:
[0075] Entity alignment under equivalence reasoning refers to identifying entities that describe the same real-world object in two or more knowledge graphs and determining the equivalence relationship between them. Entity alignment is a key technology for knowledge graph fusion and completion, which can associate knowledge from different sources.
[0076] In the equivalence relation reasoning, the present invention focuses on how to accurately identify different entities describing the same entity in different knowledge graphs. Entity alignment methods are mainly divided into embedding-based methods and graph neural network-based methods.
[0077] Embedding-based entity alignment methods measure the similarity of entities by generating embedding vectors of entities and calculating the similarity between vectors. The general framework of this method includes an embedding module, an alignment module, and a reasoning module. The embedding module is responsible for generating embedding vectors of entities, relations, and attributes; the alignment module determines the equivalence relationship between entities by calculating the similarity between vectors; and the reasoning module may use some rules or models to further optimize the alignment results.
[0078] Entity alignment methods based on graph neural networks use graph neural networks to capture the complex interactions between entities and relations, so as to better identify equivalent entity pairs. Such methods usually represent entities and relations as nodes and edges in the graph, and optimize the results of entity alignment by learning neural networks on the graph.
[0079] 8. STC measurement:
[0080] Next, the present invention calculates the social technical consistency (STC) between each test report and the demand knowledge graph. The calculation of STC not only considers the coverage and accuracy of the test report content, but also combines the collaboration mode in the developer social network to evaluate the degree to which each test report truly reflects the demand.
[0081] Specifically, STC measurement is achieved through the following sub-steps:
[0082] Collaboration requirements analysis: Based on the requirements knowledge graph, analyze the collaboration requirements of each requirement for testers and developers. For example, a requirement may require close collaboration between front-end and back-end developers, as well as communication with UI designers.
[0083] Collaboration activity assessment: Evaluate actual collaboration activities through developer social networks. For example, analyze whether there are effective communication and collaboration records between developers for specific needs.
[0084] Consistency calculation: Compare the collaboration requirements with the collaboration activities to assess their consistency. A highly consistent test report means that its content is more likely to accurately reflect software requirements and potential issues.
[0085] Finally, test reports are selected for fusion based on the STC value. Selecting test reports with high STC values means that these reports are not only accurate in content, but also reflect effective developer collaboration. These reports will be given priority, and similar test cases and results will be merged through the knowledge graph fusion algorithm to exclude duplicate or contradictory information and generate the final comprehensive test report.
[0086] This comprehensive test report not only includes possible defects and problems in the software, but also provides opinions on the collaboration model during the development process, providing a basis for improving the software development process and improving team collaboration efficiency.
[0087] The method and implementation of the present invention provide a new crowdsourcing test report fusion strategy, which can effectively improve the quality and efficiency of software testing, especially showing its unique advantages when processing large-scale crowdsourcing test data. By combining the demand knowledge graph, the test report knowledge graph and the social technology consistency metric, the present invention can not only improve the accuracy of test report fusion, but also deeply analyze the collaboration mode in the software development process, providing new perspectives and tools for software project management.
[0088] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0089] In the description of the present invention, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.
[0090] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.
Claims
1. A crowdsourcing test report fusion method based on knowledge graph, characterized in that: The following steps are involved: Based on the initial crowdsourcing test report and the set requirements document, a knowledge graph is constructed through knowledge abstraction and processing, as well as maintenance of a set of mapping units of professional terms. Based on the knowledge graph, obtaining a measure of social technical consistency between each test report and the knowledge graph according to social interaction data between the project team and crowdsourced testers; Select consistent test reports for fusion, optimize and integrate test cases, test results, and test descriptions through knowledge graph matching and fusion algorithms, eliminate duplicate or contradictory information, and generate the final comprehensive test report; In the process of generating the knowledge graph, the entity, relationship and attribute information are extracted from the requirement document and the test report to construct the requirement knowledge graph and the test report knowledge graph respectively; Based on the matching degree of text content, a deep learning model is used to evaluate the similarity of entities and relationships between the test report knowledge graph and the requirement knowledge graph; By analyzing the developer social network, including but not limited to communication discussions, collaborative modification records, and feedback loops between testers and development teams, we can determine the consistency and relevance of test reports and improve the structure and content of the knowledge graph; In the process of improving the structure and content of the knowledge graph, based on the demand knowledge graph and the test report knowledge graph, through equivalent relationship reasoning, different entities describing the same entity in different knowledge graphs are identified and entity alignment is performed to fuse and complete the knowledge graphs; In the process of obtaining the social technology consistency metric, based on the coverage and accuracy of the test report content, by combining the collaboration mode in the developer social network, the degree to which each test report truly reflects the requirements is evaluated, and the social technology consistency metric is generated; In the process of measuring social technology consistency, the process of measuring the social technology consistency between the test report and the knowledge graph includes the following steps: Collaboration requirements analysis: Based on the requirements knowledge graph, analyze the collaboration requirements of each requirement for the project team and crowdsourcing testers; Collaboration activity evaluation: evaluate the actual collaboration activities through developer social networks, where the evaluation criteria include communication frequency and collaboration depth, and how these activities affect the quality and accuracy of test reports; Consistency calculation: Compare the collaboration requirements with the collaboration activities to evaluate their consistency. The consistency of the test report indicates that its content accurately reflects the software requirements and potential problems. Finally, according to the level of the metric STC, the test reports are selected for fusion. The knowledge graph fusion algorithm is used to merge similar test cases and results, exclude duplicate or contradictory information, and generate the final comprehensive test report.
2. According to the knowledge graph-based crowdsourcing test report fusion method of claim 1, it is characterized by: In the process of obtaining the initial crowdsourcing test report, test reports from multiple crowdsourcing test platforms are collected to generate an initial crowdsourcing test report, each report at least contains a test case, a test result, and a test description, wherein each test report is derived from a diverse network environment, device configuration, and user behavior pattern.
3. According to claim 2, a crowdsourcing test report fusion method based on knowledge graph is characterized in that: In the process of constructing the knowledge graph, natural language processing and machine learning techniques are used to extract entity, relationship and attribute information from requirement documents and test reports, and the knowledge graph is generated through knowledge abstraction and processing, as well as maintenance of a set of mapping units of professional terms.
4. According to the knowledge graph-based crowdsourcing test report fusion method of claim 1, it is characterized by: In the process of generating comprehensive test reports, quality assessment is performed on the generated comprehensive test reports, including quantitative analysis of test coverage, accuracy and completeness, to ensure the high quality of the test reports and their practical guiding significance for the development process.
5. A crowdsourcing test report fusion system based on knowledge graph, characterized in that: include: The data processing module is used to construct a knowledge graph based on the initial crowdsourcing test report and the set requirements document through knowledge abstraction and processing, as well as the maintenance of the mapping unit set of professional terms; A data measurement module, for obtaining a measurement value of social technical consistency between each test report and the knowledge graph based on the knowledge graph and according to social interaction data between the project team and crowdsourcing testers; The report generation module is used to select consistent test reports for fusion. Through the matching and fusion algorithm of the knowledge graph, it optimizes and integrates test cases, test results and test descriptions, eliminates duplicate or contradictory information, and generates the final comprehensive test report. In the process of generating the knowledge graph, the system extracts entity, relationship and attribute information from the requirement document and test report to construct the requirement knowledge graph and test report knowledge graph respectively; Based on the matching degree of text content, a deep learning model is used to evaluate the similarity of entities and relationships between the test report knowledge graph and the requirement knowledge graph; By analyzing the developer social network, including but not limited to communication discussions, collaborative modification records, and feedback loops between testers and development teams, we can determine the consistency and relevance of test reports and improve the structure and content of the knowledge graph; In the process of improving the structure and content of the knowledge graph, the system identifies different entities describing the same entity in different knowledge graphs through equivalence relationship reasoning based on the demand knowledge graph and the test report knowledge graph, performs entity alignment, and fuses and completes the knowledge graphs; In the process of obtaining the social technology consistency metric, the system generates the social technology consistency metric based on the coverage and accuracy of the test report content and by combining the collaboration mode in the developer social network to evaluate the degree to which each test report truly reflects the requirements; In the process of measuring the social technology consistency of the system, the process of measuring the social technology consistency between the test report and the knowledge graph includes the following steps: Collaboration requirements analysis: Based on the requirements knowledge graph, analyze the collaboration requirements of each requirement for the project team and crowdsourcing testers; Collaboration activity evaluation: evaluate the actual collaboration activities through developer social networks, where the evaluation criteria include communication frequency and collaboration depth, and how these activities affect the quality and accuracy of test reports; Consistency calculation: Compare the collaboration requirements with the collaboration activities to evaluate their consistency. The consistency of the test report indicates that its content accurately reflects the software requirements and potential problems. Finally, according to the level of the metric STC, the test reports are selected for fusion. The knowledge graph fusion algorithm is used to merge similar test cases and results, exclude duplicate or contradictory information, and generate the final comprehensive test report.
Citation Information
Patent Citations
Crowdsourcing test report recommendation method and device, and electronic equipment
CN114048148A
Crowdsourcing annotation supported knowledge graph updating method and system and readable storage medium
CN114528413A
Power industry information analysis method and equipment based on knowledge graph
CN115080694A
Software testing method and device based on knowledge graph, equipment and storage medium
CN117667702A
Crowdsourcing data acquisition method and system based on knowledge graph, and electronic equipment
CN117993772A
