A test-oriented software quality evaluation method and device

Abstract

The application discloses a test-oriented software quality evaluation method and device, which is used for rapid, automatic and multiple quantitative evaluation of a quality state of a software or system under test in a software test process. The method and device are used in the software test process, and only test data such as software problem scale can drive the evaluation model to automatically operate, so that the cost of carrying out a software quality evaluation activity in the software test process can be reduced, and the efficiency and feasibility of the software quality evaluation can be improved. The extended comprehensive evaluation set vector used can be customized and adjusted by a user, so that flexibility of the evaluation model is ensured.

Description

Technical Field

[0001] This invention relates to the field of software quality assurance technology, and in particular to a test-oriented method and apparatus for assessing the state of software quality. Background Technology

[0002] With the rapid development of technologies such as computers, networks, communications, and artificial intelligence, and the continuous improvement of informatization levels, the functionality and performance of various systems or devices increasingly rely on related software. The business involved in software continues to expand in breadth and depth, and the scale and complexity of software are constantly increasing. Software development costs and economic losses caused by software problems are also increasing significantly. Stakeholders such as managers, testers, and users must grasp the software quality level from all aspects and at all stages in order to rationally plan and adjust corresponding measures and achieve effective control over the software status. The status of software quality is gradually improving.

[0003] Software testing is a crucial engineering method for software quality assurance and a vital part of the software lifecycle. It involves using manual or automated methods to run or measure a software system, aiming to verify whether the software meets specified requirements or to clarify the difference between expected and actual results. Software quality assurance generally requires that relevant software versions be released only after passing testing. Therefore, conducting timely software quality assessments throughout the software lifecycle not only provides direct quantitative evaluation data for subsequent formal technical reviews and program correctness verification activities, but more importantly, it allows software quality assurance personnel to analyze and track the overall software quality status evolution based on the comprehensive quantitative results of software quality assessments and specific issues discovered during software testing. This enables them to promptly grasp the quality level of the current version of the software, identify and eliminate potential software quality vulnerabilities as early as possible, reduce software quality risks, and enhance their control over large-scale, highly complex software systems. Therefore, conducting targeted, test-oriented software quality assessments in conjunction with the software testing process is becoming increasingly important.

[0004] Currently, there are no clear methods or devices for software quality assessment oriented towards testing in the market. Existing software quality evaluation systems mainly focus on the entire software lifecycle, covering various quality characteristics in multiple models such as software product quality and usage quality. There is no dedicated software quality assessment model for testing activities and engineering practices. Basically, it is necessary to refer to recommended quality models to determine quality requirements, establish quality measures, and perform quality evaluation. The implementation process involves systematic and cumbersome activities such as entity attribute quantification analysis, measure element selection, measurement method design, and measurement function construction, which require a lot of human and time resources. Because software testing in engineering practice is highly time-sensitive and generally tight in terms of schedule, time, and manpower, it cannot afford the operating costs of a systematic and comprehensive general software quality evaluation system. Therefore, general software quality evaluation systems cannot be directly applied to software testing engineering practices. There is an urgent need for a dedicated software quality assessment method and device for software testing activities that can achieve rapid and quantitative assessment of the quality of the software under test while minimizing resource consumption. Summary of the Invention

[0005] To address the aforementioned issues, the present invention provides a test-oriented software quality assessment method. This method aims to achieve rapid, automatic, and multiple quantitative assessments of the quality status of the software or system under test based on measured data such as the scale and severity distribution of software problems generated during testing, without affecting software testing engineering activities. This reduces the cost of software quality assessment and provides a feasible engineering solution for conducting software quality assessments during software testing, demonstrating strong relevance.

[0006] The present invention achieves the above objectives through the following technical solutions:

[0007] A test-oriented software quality assessment method, comprising:

[0008] Process 001: Based on the software problem evaluation set input by the testers, add the labeling of the software's problem-free state to construct a comprehensive evaluation set;

[0009] Process 002, based on the evaluation values, failure level evaluation conclusions, evaluation levels and other extended attributes set by the testers for each component in the comprehensive evaluation set, further constructs an extended comprehensive evaluation vector set on the basis of the comprehensive evaluation set;

[0010] Process 003 constructs a corresponding test evaluation factor hierarchical structure storage structure based on the input of the test evaluation factor hierarchical structure. In the test evaluation factor hierarchical structure storage structure, nodes are classified into different types according to leaf nodes and evaluation objects. Information of different types of nodes is collected and stored. In addition, the test evaluation factor hierarchical structure storage structure also stores relevant information of each layer and the link and subordinate relationship between nodes.

[0011] Process 004 receives the test question size distribution input by the testers and stores it in the corresponding leaf node in the test evaluation factor hierarchical storage structure;

[0012] Process 005: Traverse all leaf nodes and perform single-factor experimental evaluation on all leaf nodes;

[0013] Process 006 traverses each evaluation object node in each layer in a bottom-up order, performs single evaluation atom set failure evaluation on each node, and traverses to the top root node to complete the quality evaluation of the tested system or software.

[0014] As a further optimization, the software problem evaluation set is a domain of discourse for judging software problems, based on the severity level classification of software problems used in the software testing process. Let V represent the software problem evaluation set, then V = {v1, v2, ..., v...} j , ..., v m The software problem evaluation set is also simply referred to as the evaluation set.

[0015] As a further optimization, the comprehensive evaluation set, based on the software problem evaluation set, introduces the labeling of the "no problem (0 problems)" state, forming a complete set of evaluation components required for system failure assessment. This set can still be used to label the severity level of software problems, and is also suitable for judging the degree of system failure and quality status. Let V... s This is a comprehensive evaluation set constructed based on the general software problem severity level evaluation set V. ∞ To define the evaluation components (called 0-item increments) for the "no problem (0 problems)" state, V is... s Represented as V s =V∪{v ∞} = {v1, v2, ..., v j , ..., v m ,|v ∞}

[0016] As a further optimization, the extended comprehensive evaluation vector set is a vector set constructed by expanding any evaluation component in the comprehensive evaluation set from a single semantic expression to multiple dimensions such as evaluation level and quantitative assessment. This enables quantitative evaluation and multi-perspective description of the failure degree and quality status of the tested system. Let the extended V... s The resulting extended comprehensive evaluation set vector is It can be represented as Will any component Defined as

[0017] As a further optimization, the hierarchical structure of test evaluation factors is established by testers based on their understanding and analysis of the system or software under test. From the perspective of software testing, the testers decompose the various indicators or characteristics of the system or software under test, such as functions, performance, and interfaces, as well as software testing requirements such as security, resilience, boundary conditions, and strength, layer by layer, extract evaluation factors, and establish a corresponding tree structure model.

[0018] As a further optimization, the test evaluation factor hierarchical structure input is the original test evaluation factor hierarchical structure information input by the tester to the device, but it has not yet been structured and stored in the device.

[0019] As a further optimization, the test evaluation factor hierarchical structure storage structure is the corresponding form of the test evaluation factor hierarchical structure stored in the computer system in any way, such as database, table, array, or file, after the device performs structuring and other processing on the input of the test evaluation factor hierarchical structure.

[0020] As a further optimization, the leaf node is the lowest level node of a certain path in the hierarchical structure of test evaluation factors, and this node has no child nodes.

[0021] As a further optimization, the evaluation object is the specific object on which failure evaluation is performed (it can be an entity or an abstract concept such as an attribute or characteristic). The concept of the evaluation object is relative. In the hierarchical structure of test evaluation factors, each node in each layer except the leaf node (the lowest node of a certain path, which has no child nodes) can be regarded as the evaluation object of its lower-level child nodes. The final evaluation object of the system failure evaluation is the root node, that is, the software or system under test.

[0022] As a further optimization, the failure assessment is based on the distribution characteristics of the severity level and scale of various defects or errors that occur during the operation of the assessment object. It quantitatively analyzes and evaluates the operational failure of the assessment object and achieves an indirect assessment of the quality status of the assessment object.

[0023] As a further optimization, the single-factor measured evaluation is a fuzzy evaluation implemented on a single node in the hierarchical structure of test evaluation factors based on the problem scale discovered during the measured process, assuming a given comprehensive evaluation set V. s =V∪{v ∞ Let U be any set of evaluation factors, and let u i For any evaluation factor in U, based on the data collected in U i A fuzzy subset constructed based on the scale of various problems discovered in actual testing. That is, for u i Single-factor experimental evaluation.

[0024] As a further optimization, the evaluation factor set is the complete set of all evaluation factors that influence and constrain a certain evaluation object. The number of evaluation factors contained in the evaluation factor set U is called the degree of the evaluation factor set, denoted as deU(U). U is called the deU(U) degree evaluation factor set. Let U represent an evaluation factor set containing n evaluation factors, then U = {u1, u2, ..., u...} i , ..., u n}

[0025] As a further optimization, the single-evaluation atom set failure evaluation means that for any evaluation atom set ES =<et,U> deU(U) = n, given an m+1 degree comprehensive evaluation set V s =V∪{v ∞ The failure assessment of the evaluation atom set ES is based on the first-level fuzzy comprehensive evaluation, and adopts methods such as weighted comprehensive analysis of the single-factor fuzzy evaluation results of all evaluation factors under the evaluation factor set U of ES. The universe of discourse for the evaluation object et is then determined as the comprehensive evaluation set V. s =V∪{v ∞ fuzzy subset of} To achieve multi-factor fuzzy comprehensive evaluation of et, In the form of The component function form is in It is the fuzzy vector obtained by performing fuzzy evaluation on the evaluation set V. Component functions.

[0026] As a further optimization, the evaluation atom set defines the minimum analytical scope (analysis object) for implementing failure assessment, including the evaluation object and its corresponding set of evaluation factors. Let ES be the evaluation atom set, then ES can be represented in the ordered pair form ES =<et,U> , where et is the evaluation object, and set U is the set of evaluation factors corresponding to et. In the hierarchical structure of test evaluation factors, any set of evaluation atoms represents an ordered pair consisting of a non-leaf node et (evaluation object) and all its child nodes U (evaluation factor set) at the next level.

[0027] As a further optimization, the fuzzy subset... This means that, assuming the size of the software problem discovered by test et is S... et ,but Where n = deU(U), 1 ≤ i ≤ n, S U For a given comprehensive evaluation set V, the actual problem size is U. s =V∪{v ∞}, Defined as The fuzzy subset For each evaluation factor in U, a dynamic weight set is constructed. The operator o represents the composition operator used for comprehensive evaluation. The fuzzy matrix... This is a fuzzy evaluation matrix constructed for the evaluation object et.

[0028] As a further optimization, the dynamic weight set is defined as follows: when the software problem size S... et When ≠0, by comprehensively considering factors such as the actual problem scale and the severity (harm level) of various problems, and by quantifying the failure degree of each evaluation factor, a dynamic weight set is constructed based on the actual measurement weight distribution.

[0029] As a further optimization, the fuzzy evaluation matrix means that the single-factor fuzzy evaluation results of all evaluation factors in the evaluation factor set U are n fuzzy subsets.

[0030]

[0031] The fuzzy evaluation matrix is ​​constructed from all fuzzy sets.

[0032]

[0033] Another aspect of the present invention provides a test-oriented software quality assessment apparatus, comprising:

[0034] The evaluation element construction module constructs a hierarchical structure of test evaluation factors based on the hierarchical structure of the system or software under test input by the user.

[0035] The actual measurement and evaluation module enables actual measurement and evaluation of all leaf nodes;

[0036] The evaluation set construction module constructs a comprehensive evaluation set and an extended comprehensive evaluation vector set based on user input.

[0037] The hierarchical evaluation module enables the evaluation of all elements within a given level.

[0038] The evaluation and scheduling module, the actual scheduling evaluation module, the evaluation set construction module, and the hierarchical evaluation module are used to implement the evaluation.

[0039] The evaluation results display module shows the evaluation results of the system and each node.

[0040] As a further optimization, the components of each vector in the extended comprehensive evaluation vector set can be adjusted, thereby enabling the adjustment of the evaluation model.

[0041] The beneficial effects of this invention are as follows:

[0042] This invention discloses a software quality assessment method and apparatus for testing, which is mainly used for rapid, automatic, and multiple quantitative assessments of the quality status of the software or system under test during software testing. Using this method and apparatus during software testing, the assessment model can be driven to operate automatically using only test data such as the scale of software problems. This can reduce the cost of carrying out software quality assessment activities during software testing and improve the efficiency and feasibility of software quality assessment. The extended comprehensive evaluation set vector used can be customized and adjusted by the user, ensuring the flexibility of the assessment model. Attached Figure Description

[0043] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this embodiment. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0044] Figure 1 This invention provides a schematic diagram of the overall process of a test-oriented software quality assessment method.

[0045] Figure 2 This is a schematic diagram of one implementation of the present invention. Detailed Implementation

[0046] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be described in detail below. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. Based on the embodiments of this invention, all other implementation methods obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0047] like Figure 2 As shown in the schematic diagram of one implementation of the present invention, it includes:

[0048] The device, based on the software problem evaluation set V input by the tester, adds a calibration of the software's problem-free state to form a comprehensive evaluation set V. s .

[0049] Based on the evaluation values, failure level evaluation conclusions, evaluation levels, and other extended attributes set by the testers for each component in the comprehensive evaluation set, the device further constructs an extended comprehensive evaluation vector set.

[0050] Based on their understanding and analysis of the system or software under test, testers decompose the system or software under test from a software testing perspective, including its functions, performance, interfaces, and other indicators or characteristics, as well as software testing requirements such as security, resilience, boundary conditions, and strength. They then extract evaluation factors and establish a corresponding tree structure model, forming a hierarchical structure of test evaluation factors.

[0051] The device constructs a hierarchical framework storage structure (Structure) based on the overall structure of the test evaluation factors hierarchy input by the testers. In addition to storing relevant information for each layer in the test evaluation factor hierarchy, the hierarchical framework storage structure (Structure) must also store relevant information for virtual layers. Virtual layers are abstract layers composed of all leaf nodes. The information stored in the hierarchical framework storage structure (Structure) for each layer includes layer number, layer type, input data link, output result link, number of nodes, layer description, and other information.

[0052] Based on the distribution of leaf nodes in the hierarchical structure of test evaluation factors input by the testers, the device constructs a leaf node original data storage structure Layer0_in_leafOriginal corresponding to the virtual layer. The leaf node original data storage structure Layer0_in_leafOriginal stores relevant information of each leaf node and the distribution of the scale of test problems found by that leaf node in the software problem evaluation set.

[0053] The device constructs an evaluation object node information storage structure, the_nodes_of_ETs, based on the connection relationships between nodes in the hierarchical structure of test evaluation factors input by the testers. The evaluation object node information storage structure, the_nodes_of_ETs, stores the information of all child nodes directly under each evaluation object node in each layer in a hierarchical manner.

[0054] The device constructs a node information storage structure Layerx_in_nodesInfo layer by layer based on the hierarchical structure of test evaluation factors input by the testers. Each layer of the node information storage structure Layerx_in_nodesInfo stores the node information of each node contained in that layer. The node information storage includes the node type nodeType, the list of child nodes childrens, the layer sequence number layerIndex, etc. The device traverses and tests each node layer by layer in sequence. If the node type is a leaf node, the node list childrens of that node is set to empty. If the node type is an evaluation object, the device reads the child node information of that node from the evaluation object node information storage structure the_nodes_of_ETs and sets it as the child node list childrens of that node.

[0055] The device iterates through each leaf node and performs experimental evaluation: it obtains the input data link information of the virtual layer from the hierarchical framework storage structure Structure, loads the experimental raw data of all leaf nodes from the original data storage structure Laycr0_in_leafOriginal, expands the vector to ∞ for all leaf nodes, and constructs the universe of discourse V for each leaf node as the comprehensive evaluation set. s The device uses a fuzzy subset to implement single-factor fuzzy evaluation based on the measured problem size. The device stores the measured evaluation results of the leaf nodes into the measured evaluation storage structure Layer0_RealTestValuation.

[0056] Device Loading Extended Comprehensive Evaluation Vector Set

[0057] Based on the hierarchical information of test evaluation factors stored in the hierarchical framework storage structure Structure, the device constructs the evaluation result storage structure Layerx_Valuation for each layer except the virtual layer. Each layer's evaluation result storage structure Layerx_Valuation stores the node evaluation result output architecture of all nodes under that layer, and the node evaluation result output architecture includes the problem size S related to the evaluation of that node. et Dynamic weight set Fuzzy evaluation matrix Evaluation results, Valuation, and other information.

[0058] Following a bottom-up order, the device traverses and analyzes each node in each layer of the hierarchical framework storage structure (excluding the virtual layer) to evaluate each node in each layer and even the entire system: The device reads the hierarchical framework storage structure (excluding the virtual layer), reads the current layer information and the previous layer information layer by layer from bottom to top, and evaluates each node in each layer.

[0059] If the device detects that the node type being analyzed is a leaf node, it takes the measured evaluation result of the leaf node stored in the measured evaluation storage structure Layer0_RealTestValuation as its final evaluation result and saves it to the evaluation result Valuation of the node evaluation result output architecture.

[0060] If the device detects that the currently analyzed node type is an evaluation object, the device accumulates the problem sizes of all child nodes in the previous layer of that node to obtain the current node's problem size, and saves it to the node evaluation result output architecture's problem size S. etIn the process, the device reads the evaluation results (Valuation) of all child nodes in the previous layer of the current node to construct the fuzzy evaluation matrix of the current node, and stores it in the fuzzy evaluation matrix of the node evaluation result output architecture of the current node. In the middle, the device detects the current node problem size S. et If the current node has a problem size S et If the value is 0, then the "0-item increment" of the current node's evaluation result is set to 1, and the remaining components are set to 0. If the problem size S et If it is not 0, then use the extended comprehensive evaluation vector set. The failure degree of the current node is calculated by summing the products of the evaluation values ​​of each vector and the corresponding problem sizes. According to the degree of failure Find the dynamic weight set Dynamic weight set With fuzzy evaluation matrix The evaluation result of the current node is synthesized and saved to the evaluation result Valuation of the current node's node evaluation result output architecture.

[0061] Another aspect of the present invention provides a test-oriented software quality assessment apparatus, comprising:

[0062] The evaluation element construction module constructs a hierarchical structure of test evaluation factors based on the hierarchical structure of the system or software under test input by the user.

[0063] The actual measurement and evaluation module enables actual measurement and evaluation of all leaf nodes;

[0064] The evaluation set construction module constructs a comprehensive evaluation set and an extended comprehensive evaluation vector set based on user input.

[0065] The hierarchical evaluation module enables the evaluation of all elements within a given level.

[0066] The evaluation and scheduling module, the actual scheduling evaluation module, the evaluation set construction module, and the hierarchical evaluation module are used to implement the evaluation.

[0067] The evaluation results display module shows the evaluation results of the system and each node.

[0068] The components of each vector in the extended comprehensive evaluation vector set can be adjusted, thereby enabling the adjustment of the evaluation model.

[0069] The method of this application uses test data such as the size of software problems as input during the software testing process to achieve rapid, automatic, and multiple quantitative assessments of the quality status of the software or system under test. It can reduce the cost of carrying out software quality assessment activities during the software testing process and improve the efficiency and feasibility of software quality assessment. The extended comprehensive evaluation set vector used can be customized and adjusted by the user, ensuring the flexibility of the assessment model.

[0070] In summary, the above are specific embodiments of the present invention. Those skilled in the art can easily conceive of variations or substitutions when implementing the present invention. Therefore, the various specific technical features described in the specific embodiments can be combined in various suitable ways. Furthermore, various different embodiments of the present invention can also be arbitrarily combined.

Claims

1. A test-oriented software quality assessment method, characterized in that, include: Step S1: Constructing the Comprehensive Evaluation Set: Based on the software problem evaluation set input by the testers, a "no problem" state is introduced. This "no problem" state is also called the "0 problem" state. This constitutes the complete set of evaluation components required for system failure assessment, which is the comprehensive evaluation set. Let V... s This is a comprehensive evaluation set constructed based on the software problem evaluation set V. ∞ To calibrate the evaluation components for the "no problem" state, v ∞ If it is called a zero-term increment, then V s Represented as V s =V∪{v ∞ } = {v1, v2, ..., v j , ..., v m ,|v ∞ }; The software problem evaluation set is a domain of evaluation levels for software problems, based on the severity levels of software problems used in the software testing process. Let V represent the software problem evaluation set, then V = {v1, v2, ..., v...} j , ..., v m The software problem evaluation set is also simply referred to as the evaluation set. Step S2: Construct an extended comprehensive evaluation vector set: Based on the extended attributes set by the testers for each component in the comprehensive evaluation set, further construct an extended comprehensive evaluation vector set on the basis of the comprehensive evaluation set. The extended attributes include evaluation values, failure level evaluation conclusions, and evaluation levels. Step S3: Construct a hierarchical structure of test evaluation factors: Based on the input of the hierarchical structure of test evaluation factors, construct a corresponding storage structure of the hierarchical structure of test evaluation factors. In the storage structure of the hierarchical structure of test evaluation factors, nodes are classified into different types according to leaf nodes and evaluation objects. Information of different types of nodes is collected and stored. In addition, the storage structure of the hierarchical structure of test evaluation factors also stores relevant information of each layer and the link and subordinate relationship between nodes. The evaluation object is the specific object on which failure evaluation is performed. The concept of evaluation object is relative. In the hierarchical structure of test evaluation factors, each node in each layer except the leaf nodes can be regarded as the evaluation object of its lower-level child nodes. The final evaluation object of system failure evaluation is the root node, i.e., the software or system under test. The hierarchical structure of test evaluation factors is a tree-like model in which testers, based on their understanding and analysis of the system or software under test, decompose the test requirements of the system or software under test layer by layer from the perspective of software testing, extract evaluation factors, and establish corresponding evaluation factors. Step S4: Input Test Problem Size Distribution: Receive the test problem size distribution input by the testers and store it in the corresponding leaf node of the test evaluation factor hierarchical storage structure; Step S5: Perform single-factor experimental evaluation: Traverse all leaf nodes and perform single-factor experimental evaluation on all leaf nodes; Step S6: Comprehensive quality assessment: Traverse each assessment object node in each layer in a bottom-up order, perform single assessment atom set failure assessment on each node, and traverse to the top root node to complete the quality assessment of the tested system or software. The evaluation atom set defines the minimum analytical scope for performing failure assessment, including the assessment object and its corresponding set of evaluation factors. Let ES be the evaluation atom set, then ES can be represented in the ordered pair form ES =<et,U> , where et is the evaluation object, and set U is the set of evaluation factors corresponding to et. In the hierarchical structure of test evaluation factors, any set of evaluation atoms represents an ordered pair consisting of a non-leaf node et and all its child nodes U in the next layer. The evaluation factor set is the complete set of all evaluation factors that influence and constrain a certain evaluation object. The number of evaluation factors contained in the evaluation factor set U is called the degree of the evaluation factor set, denoted as deU(U). U is called the deU(U) degree evaluation factor set. Let U represent the evaluation factor set containing n evaluation factors, then U = {u1, u2, ..., u...} i ,…,u n }; The single-evaluation atom set failure evaluation is as follows: for any evaluation atom set ES =<et,U> deU(U) = n, given an m+1 degree comprehensive evaluation set V s =V∪{v ∞ The failure assessment of the evaluation atom set ES is a single-factor fuzzy evaluation result of all evaluation factors under the weighted comprehensive evaluation factor set U of ES. The universe of discourse for et is the comprehensive evaluation set V. s =V∪{v ∞ fuzzy subset of} The component function form is in It is the fuzzy vector obtained by performing fuzzy evaluation on the evaluation set V. The component functions; let the size of the software problem discovered by test et be S. et ,but Where n = deU(U), 1 ≤ i ≤ n, S U For a given comprehensive evaluation set V, the actual problem size is U. s =V∪{v ∞ }, Defined as The fuzzy subset For each evaluation factor in U, the operator is used to construct a dynamic weight set. The fuzzy matrix represents the composite operator used in the comprehensive evaluation. The fuzzy evaluation matrix constructed for the evaluation object et; The dynamic weight set is: when the software problem size S... et When ≠0, by comprehensively considering the measured problem scale and severity of various problems of each evaluation factor, and by quantifying the failure degree of each evaluation factor, a weight distribution based on actual measurements is constructed. The process of obtaining the dynamic weight set is as follows: If the node type being analyzed is an evaluation object, the problem size of the current node is obtained by summing the problem sizes of all child nodes in the previous layer of the current node, and this problem size S is saved to the node evaluation result output architecture of the current node. et In the middle, the evaluation results of all child nodes in the previous layer of the current node are read to construct the fuzzy evaluation matrix of the current node, and stored in the node evaluation result output architecture fuzzy evaluation matrix of the current node. middle; Detect the current node's problem size S et If the problem size S et If it is not 0, then use the extended comprehensive evaluation vector set. The failure degree of the current node is calculated by summing the products of the evaluation values ​​of each vector and the corresponding problem sizes. According to the degree of failure Find the dynamic weight set The fuzzy evaluation matrix is ​​defined as follows: the single-factor fuzzy evaluation results of all evaluation factors in the evaluation factor set U are n fuzzy subsets. The fuzzy evaluation matrix is ​​constructed from all fuzzy subsets. 。 2. The test-oriented software quality assessment method according to claim 1, characterized in that, The comprehensive evaluation set V s Any evaluation component is expanded from a single semantic expression to multiple dimensions such as evaluation level and quantitative assessment, constructing a corresponding extended comprehensive evaluation vector set; let the extended V The extended comprehensive evaluation vector set obtained after s is It can be represented as Will any component Defined as .

3. The test-oriented software quality assessment method according to claim 2, characterized in that, Single-factor measured evaluation is a fuzzy evaluation of a single node in the hierarchical structure of test evaluation factors, based on the problem scale discovered during the measured process. Given a comprehensive evaluation set V... s =V∪{v ∞ Let U be any set of evaluation factors, and let u i For any evaluation factor in U, based on the data collected in U i A fuzzy subset constructed from the scale of various severity levels of problems discovered during actual testing. That is, the evaluation factor u i Single-factor experimental evaluation.

4. A software quality assessment device for testing, characterized in that, The apparatus is used to perform the method according to any one of claims 1 to 3.

Images