Software development quality evaluation system based on big data

Abstract

The invention discloses a software development quality evaluation system based on big data, and relates to the technical field of big data, the software development quality evaluation system comprises a management center, the management center is connected with a software acquisition module, a development processing module, a quality analysis module and an intelligent estimation module; constructing a simulation virtual space to perform simulation operation on the collected software operation codes, performing interactive execution on the software operation codes, and collecting software interactive data and software comprehensive information; the software interaction data and the software comprehensive information are extracted and classified to obtain comprehensive influence parameters, and the comprehensive influence parameters are converted and extracted to obtain multi-dimensional influence coefficients; carrying out digital-graph conversion on the multi-dimensional influence coefficient to obtain an influence coefficient dynamic graph, carrying out region conversion on the influence coefficient dynamic graph to obtain an influence action region, and carrying out quality judgment on the software through the influence action region to obtain a comprehensive quality weight; the evaluation work efficiency is improved, and the accuracy of evaluation results is improved.

Description

Technical Field

[0001] The present invention relates to the field of big data technology, and specifically to a software development quality evaluation system based on big data. Background Art

[0002] With the development of information technology, software has become an indispensable part of modern society. The number and scale of software development projects are constantly expanding, and software development quality has become a key factor determining the success of projects. Currently, software development quality evaluation mainly relies on manual review, code review, and limited test cases. These methods are usually time-consuming, costly, and difficult to cover all potential problems.

[0003] The development of big data technology provides new possibilities for collecting, analyzing, and processing massive data. By real-time collecting, integrating, and analyzing multi-source data in the development process, standardizing the data, and extracting features to comprehensively evaluate software quality, potential quality problems can be detected early, and the later repair cost can be reduced. For this reason, a software development quality evaluation system based on big data is provided herein. Summary of the Invention

[0004] The object of the present invention can be achieved by the following technical solutions: A software development quality evaluation system based on big data, including a management center, which is connected with a software collection module, a development processing module, a quality analysis module, and an intelligent prediction module; The software collection module is used to collect the software running environment and software running code; The development processing module is used to construct a simulation virtual space, simulate the operation of the software running code through the simulation virtual space, interactively execute the software running code and collect software interaction data, and perform data screening based on the interactive execution process to obtain software comprehensive information; The quality analysis module is used to extract and classify the software interaction data and software comprehensive information to obtain comprehensive influence parameters, perform modal conversion on the comprehensive influence parameters to obtain comprehensive influence signals, perform separation transformation on the comprehensive influence signals by setting screening tracking coefficients to obtain tracking coefficient components, and perform multi-dimensional extraction on the comprehensive influence signals according to the tracking coefficient components to obtain multi-dimensional influence coefficients; The intelligent prediction module is used to perform digital graph conversion on the multi-dimensional influence coefficients to obtain an influence coefficient dynamic graph, set a moving central axis, sort the endpoints of the influence coefficient dynamic graph to obtain a coefficient component value sequence, perform region conversion on the coefficient component value sequence to obtain an influence action region, and perform quality determination on the software through the influence action region to obtain a comprehensive quality weight.

[0005] Preferably, the process of interactively executing the software running code and collecting software interaction data includes: Construct a simulation virtual space according to the software operating environment, and set up a node monitoring end based on the simulation virtual space; Upload the software operation code to the simulation virtual space, send a start instruction to the obtained software operation code, and simulate the operation of the software operation code through the simulation virtual space based on the start instruction; Send a user interaction instruction to the obtained software operation code based on the simulation operation process, interactively execute the software operation code according to the user interaction instruction, and perform interactive acquisition on the interactive execution process to obtain software interaction data.

[0006] Preferably, the process of data screening based on the interactive execution process includes: Collect data from the software operation code after interactive execution through the node monitoring end to obtain software comprehensive information; Screen for defects in the obtained software comprehensive information to obtain comprehensive fault information; Perform scalar statistics on the software comprehensive information according to the software interaction data to obtain the operation throughput.

[0007] Preferably, the process of the quality analysis module extracting and classifying the software interaction data and the software comprehensive information includes: Obtain the software interaction data, perform matching and association on the software interaction data to obtain request response parameters; Screen and extract the software comprehensive information according to the comprehensive fault information to obtain the fault effect coefficient; Record the obtained operation throughput, request response parameters, and fault effect coefficient as comprehensive influence parameters.

[0008] Preferably, the process of separating and transforming the comprehensive influence signal by setting a screening tracking coefficient includes: Set the screening tracking coefficient, perform attribute transformation on the screening tracking coefficient to obtain a tracking component; Perform dynamic separation on the screening tracking coefficient according to the obtained tracking component to obtain a dynamic component; Perform separation and interception on the screening tracking coefficient according to the obtained tracking component and dynamic component to obtain a tracking coefficient component.

[0009] Preferably, the process of multidimensionally extracting the comprehensive influence signal according to the tracking coefficient component includes: Upload the obtained tracking coefficient component to the comprehensive influence signal, and reconstruct the interval of the comprehensive influence signal through the tracking coefficient component to obtain an influence coefficient component; Perform reconstruction statistics on the obtained influence coefficient component to obtain a reconstructed influence coefficient; Perform reconstruction extraction on the reconstructed influence coefficient according to the screening tracking coefficient to obtain a multidimensional influence coefficient.

[0010] Preferably, the process of sorting the endpoints of the dynamic graph of influence coefficients includes: Construct an original dynamic graph based on multi-dimensional influence coefficients, upload the multi-dimensional influence coefficients to the original dynamic graph, perform number-graph conversion on the multi-dimensional influence coefficients based on the original dynamic graph to obtain the dynamic graph of influence coefficients; Set a moving central axis, upload the moving central axis to the dynamic graph of influence coefficients, and perform endpoint positioning on the dynamic graph of influence coefficients through the moving central axis to obtain endpoint extreme values; Perform positioning and sorting on the dynamic graph of influence coefficients through the moving central axis to obtain a sequence of coefficient component values.

[0011] Preferably, the process of determining the quality of the software through the influence area includes: Perform internal positioning on the sequence of coefficient component values to obtain in-segment component values, perform regional screening on the endpoint extreme values and in-segment component values to obtain the influence area; Perform internal partitioning through the influence area to obtain characteristic influence blocks, perform block accounting on the characteristic influence blocks according to the multi-dimensional influence coefficients to obtain quality influence blocks; Perform classification and sorting on the obtained quality influence blocks to obtain a quality influence set, and perform weight estimation on the software through the quality influence set to obtain a comprehensive quality weight.

[0012] Compared with the prior art, the beneficial effects of the present invention are: 1. Construct a simulation virtual space to perform simulation operation on the collected software operation code. During the simulation operation process, simulate the interaction requests of users to send interaction instructions to the simulation virtual space and perform interaction execution, and collect software interaction data and software comprehensive information generated during the interaction process; Simulate the operation of software code in the virtual space, improve the operation efficiency, facilitate the accurate collection of operation data, and simulate the request operations of users on the software, which is beneficial to improving the user experience; 2. Extract and classify the software interaction data and software comprehensive information to obtain comprehensive influence parameters. By processing the comprehensive influence parameters, improve the data processing speed, facilitate the construction of a parameter change graph, analyze the parameter change graph and convert it into an area region corresponding to the parameter, display the change situation of each parameter index of software development according to the area region, and then perform quality determination on the software through the converted area region to obtain a comprehensive quality weight; Obtain the quality ranking of software development according to the obtained comprehensive quality weight, provide a quantitative evaluation for the quality of software code, help developers optimize the code structure, and improve the stability of the software. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0014] Figure 1 This is the schematic diagram of the present invention. Specific embodiments

[0015] The following will clearly and completely describe the technical solutions of the present invention in combination with the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0016] As Figure 1 shown, a software development quality evaluation system based on big data includes a management center, and the management center is connected with a software collection module, a development processing module, a quality analysis module, and an intelligent prediction module; The software collection module is used to collect the software operation environment and software operation code; The development processing module is used to construct a simulation virtual space, simulate the operation of the software operation code through the simulation virtual space, interactively execute the software operation code and collect software interaction data, and perform data screening based on the interactive execution process to obtain software comprehensive information; The quality analysis module is used to extract and classify the software interaction data and software comprehensive information to obtain comprehensive influence parameters, perform modal conversion on the comprehensive influence parameters to obtain comprehensive influence signals, perform separation transformation on the comprehensive influence signals by setting screening tracking coefficients to obtain tracking coefficient components, and perform multi-dimensional extraction on the comprehensive influence signals according to the tracking coefficient components to obtain multi-dimensional influence coefficients; The intelligent prediction module is used to perform digital graph conversion on the multi-dimensional influence coefficients to obtain an influence coefficient dynamic graph, set a moving central axis, sort the endpoints of the influence coefficient dynamic graph to obtain a coefficient component value sequence, perform region conversion on the coefficient component value sequence to obtain an influence action region, and perform quality determination on the software through the influence action region to obtain a comprehensive quality weight.

[0017] In actual application, the software collection module is used to collect the software operation environment and software operation code, and the specific process includes: Capture the software development environment to obtain the software operation environment; The environment capture represents capturing the operating environment of software development. The operating environment represents the devices and environmental information required for the developed software to run properly, including hardware devices, software environment, development tools, testing tools, and deployment tools. Among them, the hardware devices include development devices, server devices, and network devices, and the software environment includes operating systems, runtime environments, databases, and containerization tools; Data collection is performed on software development to obtain software operation code; The data collection represents collecting the code data of the developed software, which is the software operation code; Furthermore, the software operation code of the collected software has been tested and can run properly. It's just that the code quality is not clear and needs to process the software operation code to achieve the purpose of evaluating the software development quality.

[0018] The development processing module is used to construct a simulation virtual space, simulate the operation of the software operation code through the simulation virtual space, interactively execute the software operation code and collect software interaction data, and perform data screening based on the interactive execution process to obtain software comprehensive information. The specific process includes: Construct a simulation virtual space according to the obtained software operating environment. The simulation virtual space is a virtual generation of the software operating environment to obtain a three-dimensional virtual space where the software can run properly and realizes the same functions as in reality. That is, according to the hardware devices, software environment, development tools, testing tools, and deployment tools included in the software operating environment, the simulation virtual space also includes virtual generated hardware devices, software environment, development tools, testing tools, and deployment tools; Set up node monitoring terminals based on the simulation virtual space. The node monitoring terminals are used to monitor the software operation code in the simulation virtual space and collect the data information generated by the software operation code running in the simulation virtual space; Upload the obtained software operation code to the simulation virtual space, send a start instruction to the obtained software operation code, and simulate the operation of the software operation code through the simulation virtual space based on the start instruction; Furthermore, the start instruction represents sending a start instruction for software operation to the software operation code in the simulation virtual space. After the software receives the instruction, it controls the software operation code to perform simulation operation through the simulation virtual space, that is, starts the operation of the software operation code of the software in the simulation virtual space; Send a user interaction instruction to the obtained software operation code based on the simulation operation process, interactively execute the software operation code according to the received user interaction instruction, and perform interactive collection on the interactive execution process to obtain software interaction data; It should be further explained that, in the specific implementation process, the user interaction instruction represents a request operation initiated by a simulated user to the software, the request operation represents the data information of the user input acting on the software, the user interaction instruction is generated according to the obtained request operation, the obtained user interaction instruction is uploaded to the software running code in the simulation running, and the interaction data generated by the user interaction instruction and the software running code is collected to obtain the software interaction data; The software interaction data includes request content, response time and request time point, wherein the request content indicates the content of the user interaction instruction, and the response time indicates the time taken from the user initiating the request to the software system returning the result; for example, in a software for watching a video, the user inputs the request content, clicks a button to initiate the request, sends the obtained request to the server, the server processes the request and returns the data, the client receives the data and displays the result, then the total time from the time the user clicks the button to the time the result is displayed is the response time; Collecting data of the software running code after interactive execution through the node monitoring terminal to obtain comprehensive software information, wherein the data collection means collecting data generated during the software running process after interactive execution; The software comprehensive information includes performance data, operation log data and security data, wherein the performance data includes running time, CPU usage, memory usage, security data indicates vulnerability reports and security attack events suffered, and the operation log data includes log information generated by user interactive instructions after the software running code is simulated. At the same time, the operation log data can also monitor the software interactive data generated by interactive execution; Perform defect screening on the obtained software comprehensive information to obtain comprehensive fault information. The defect screening means screening the data generated by the running code of the started software through existing equipment and tools to obtain fault information during the running process, i.e. comprehensive fault information. Among them, "existing equipment and tools" include code review tools, test frameworks, integration test tools, performance test tools, security test tools, debugging tools, log analysis tools, version monitoring tools and monitoring tools, which exist in reality and are used to monitor the fault information generated during the running of the software, can be normally used in the simulated virtual space, and perform the same monitoring on the uploaded software running code; The comprehensive fault information includes the fault name, error code, fault occurrence time and fault repair time; Perform scalar statistics on the obtained software comprehensive information according to the obtained software interaction data to obtain the operation throughput; The scalar statistical representation records each request content corresponding to each user interaction instruction in the software interaction data as a request instruction, and sets a unit threshold time, which is a preset period of time. In this embodiment, the unit threshold time is set to be the same as the time length of the running time in the software comprehensive information, that is, the number of request instructions is counted within the running time; Based on the unit threshold time, count the number of request instructions to obtain the number of requests, and obtain the running throughput according to the obtained number of requests and the unit threshold time, where the running throughput = the number of requests ÷ the unit threshold time.

[0019] The quality analysis module is used to extract and classify the software interaction data and the software comprehensive information to obtain comprehensive influence parameters, and perform conversion and extraction on the comprehensive influence parameters to obtain multi-dimensional influence coefficients. The specific process includes: Obtain the software interaction data, perform matching and association on the software interaction data to obtain request response parameters, which means associating the response time in the software interaction data with the corresponding request content, and recording the associated response time as the request response parameter; According to the obtained fault comprehensive information, screen and extract the software comprehensive information to obtain a fault effect coefficient, where the fault effect coefficient includes the defect resolution time and the failure rate; Further, the screening and extraction means counting the fault names in the fault comprehensive information according to the running time in the software comprehensive information to obtain the number of faults within the running time, marking the obtained number of faults as the number of defects, and obtaining the failure rate according to the obtained number of defects and the running time, where the failure rate = the number of defects ÷ the running time; Obtain the defect resolution time according to the fault occurrence time and the fault repair time in the fault comprehensive information, and the defect resolution time is the time difference between the fault repair time and the fault occurrence time; Record the obtained running throughput, request response parameters, and fault effect coefficients as comprehensive influence parameters; Perform modal conversion on the obtained comprehensive influence parameters to obtain a comprehensive influence signal; The modal conversion means converting the obtained comprehensive influence parameters into a signal form. According to the running throughput, request response parameters, and fault effect coefficients included in the comprehensive influence parameters, the comprehensive influence signal includes a running throughput signal, a request response signal, and a fault effect signal; Set a screening and tracking coefficient, and the manifestation form of the screening and tracking coefficient is in the form of a function; Perform attribute transformation on the obtained screening and tracking coefficient to obtain a tracking component; The above-mentioned attribute transformation represents controlling the screening tracking coefficient to perform stretching and translation transformations in the time dimension and the frequency dimension, and statistically analyzing the distances of the stretching and translation transformations to obtain tracking components; Dynamically separating the screening tracking coefficient according to the obtained tracking components to obtain dynamic components. The dynamic separation represents statistically analyzing the distances between adjacent two tracking components in the screening tracking coefficient to obtain dynamic components; Separating and intercepting the screening tracking coefficient according to the obtained tracking components and dynamic components to obtain tracking coefficient components; The process of the separation and interception includes: Obtaining the number of intercepted segments according to the obtained tracking components and dynamic components, and marking the obtained number of intercepted segments as JQ, where, , S represents the length of the screening tracking coefficient, D represents the dynamic component, b represents the number of dynamic components in the screening tracking coefficient, z represents the tracking component, and rounding down the calculated number of intercepted segments, that is, the finally obtained number of intercepted segments is a positive integer; Intercepting and segmenting the screening tracking coefficient according to the obtained number of intercepted segments to obtain tracking coefficient components. The intercepting and segmenting represents equally segmenting the screening tracking coefficient according to the number of intercepted segments to obtain equally long tracking coefficient components, and the number of tracking coefficient components is equal to the number of intercepted segments; Performing multi-dimensional extraction on the obtained comprehensive influence signal according to the obtained tracking coefficient components to obtain multi-dimensional influence coefficients; It should be further noted that in the specific implementation process, the process of the multi-dimensional extraction includes: Uploading the obtained tracking coefficient components to the comprehensive influence signal, and reconstructing the interval of the comprehensive influence signal through the tracking coefficient components to obtain influence coefficient components. The uploading order of "uploading the obtained tracking coefficient components to the comprehensive influence signal" is carried out according to the order of the intercepting and segmenting; The interval reconstruction represents convolving the tracking coefficient components with the comprehensive influence signal to obtain influence coefficient components, that is, convolving each tracking coefficient component with the comprehensive influence signal respectively according to the order of the intercepting and segmenting to obtain corresponding influence coefficient components; Performing reconstruction statistics on the obtained influence coefficient components to obtain reconstructed influence coefficients; The reconstruction statistics represents summing up the obtained influence coefficient components based on the order of the intercepting and segmenting to obtain reconstructed influence coefficients; Performing reconstruction extraction on the reconstructed influence coefficients according to the obtained screening tracking coefficient to obtain multi-dimensional influence coefficients; The reconstruction extraction means multiplying the obtained screening tracking coefficient by the reconstruction influence coefficient to obtain a screening reconstruction coefficient, and performing a discrete Fourier transform on the obtained screening reconstruction coefficient to obtain a multi-dimensional influence coefficient; Specifically, according to the operating throughput signal, request response signal, and fault effect signal included in the comprehensive influence signal, the multi-dimensional influence coefficient includes an operating throughput coefficient, a request response coefficient, and a fault effect coefficient.

[0020] The intelligent prediction module is used to perform a digital-to-graph conversion on the multi-dimensional influence coefficient to obtain an influence coefficient dynamic graph, perform a region conversion on the influence coefficient dynamic graph to obtain an influence effect region, and determine the quality of the software through the influence effect region to obtain a comprehensive quality weight. The specific process includes: Construct an original dynamic graph based on the multi-dimensional influence coefficient, upload the obtained multi-dimensional influence coefficient to the original dynamic graph, and perform a digital-to-graph conversion on the multi-dimensional influence coefficient based on the original dynamic graph to obtain an influence coefficient dynamic graph, where the influence coefficient dynamic graph includes influence coefficient curves; The constructed original dynamic graph is a two-dimensional rectangular coordinate system. The digital-to-graph conversion means converting the multi-dimensional influence coefficient into the original dynamic graph, that is, generating an influence coefficient curve according to the multi-dimensional influence coefficient, uploading the obtained influence coefficient curve to the original dynamic graph to obtain an influence coefficient dynamic graph. Among them, according to the operating throughput coefficient, request response coefficient, and fault effect coefficient included in the multi-dimensional influence coefficient, the influence coefficient curve includes a throughput coefficient curve, a request response coefficient curve, and a fault effect coefficient curve; Furthermore, the influence coefficient dynamic graph includes a throughput coefficient curve, a request response coefficient curve, and a fault effect coefficient curve. According to the fault effect coefficient including defect resolution time and failure rate, the influence coefficient curve includes a throughput coefficient curve, a request response coefficient curve, a defect resolution coefficient curve, and a failure rate coefficient curve. Correspondingly, there are complete throughput coefficient curves, request response coefficient curves, defect resolution coefficient curves, and failure rate coefficient curves in the influence coefficient dynamic graph; Set a moving central axis, upload the obtained moving central axis to the influence coefficient dynamic graph, and perform endpoint positioning on the influence coefficient dynamic graph through the moving central axis to obtain endpoint extrema, where the endpoint extrema include a front maximum and a rear minimum; The moving central axis is a straight line perpendicular to the horizontal axis and moves left and right on the horizontal axis. By translating the moving central axis, the intersection points of the moving central axis and the influence coefficient curve are obtained. The values at the intersection points have corresponding multi-dimensional influence coefficients, and the multi-dimensional influence coefficients at the intersection points are marked as node influence coefficients; The endpoint positioning means obtaining the intersection points at the highest and lowest points of the influence coefficient curve by translating the central axis in the dynamic influence coefficient diagram, marking the node influence coefficient of the intersection point at the highest point as the front maximum value, and marking the node influence coefficient of the intersection point at the lowest point as the rear minimum value; Positioning and sorting the dynamic influence coefficient diagram by moving the central axis to obtain a coefficient component value sequence; The internal positioning means obtaining the values of the intersection points of the influence coefficient curve, that is, the node influence coefficients, by translating the central axis in the dynamic influence coefficient diagram, and sorting the obtained node influence coefficients in descending order to obtain a coefficient component value sequence; Performing internal positioning on the obtained coefficient component value sequence to obtain end-in component values, where the end-in component values include a first coefficient quantile value, a second coefficient quantile value, and a central coefficient quantile value; The internal positioning means obtaining the node influence coefficient at the 25% position of the coefficient component value sequence in the obtained coefficient component value sequence and denoting the obtained node influence coefficient as the first coefficient quantile value, obtaining the node influence coefficient at the 75% position of the coefficient component value sequence in the obtained coefficient component value sequence and denoting the obtained node influence coefficient as the second coefficient quantile value, obtaining the node influence coefficient at the 50% position of the coefficient component value sequence in the obtained coefficient component value sequence and denoting the obtained node influence coefficient as the central coefficient quantile value; Performing region screening on the obtained endpoint extreme values and in-segment component values to obtain an influence region; The display effect diagram means first constructing a two-dimensional rectangular coordinate system, marking the central coefficient quantile value at the corresponding position on the horizontal axis, drawing a straight line perpendicular to the horizontal axis at the central quantile value, denoted as the central quantile axis, marking the obtained front maximum value and rear minimum value at the corresponding positions on the vertical axis of the two-dimensional rectangular coordinate system, drawing a straight line parallel to the horizontal axis at the front maximum value, denoted as the maximum quantile axis, drawing a straight line parallel to the horizontal axis at the rear minimum value, denoted as the minimum quantile axis, marking the first coefficient quantile value and the second coefficient quantile value at the corresponding positions on the horizontal axis of the two-dimensional rectangular coordinate system, drawing a straight line perpendicular to the horizontal axis at the first coefficient quantile value, denoted as the first quantile axis, and drawing a straight line perpendicular to the horizontal axis at the second coefficient quantile value, denoted as the second quantile axis, Marking the region enclosed by the central quantile axis, the maximum quantile axis, the minimum quantile axis, the first quantile axis, and the second quantile axis in the two-dimensional rectangular coordinate system, denoted as the influence region, that is, the region enclosed by the intersection of the four straight lines; Specifically, according to the influence coefficient curves including the throughput coefficient curve, the request response coefficient curve, the defect resolution coefficient curve, and the failure rate coefficient curve, the influence areas include the throughput influence area, the response influence area, the defect resolution influence area, and the failure rate influence area, that is, the throughput influence area corresponding to the throughput coefficient curve, the response influence area corresponding to the request response coefficient curve, the defect resolution influence area corresponding to the defect resolution coefficient curve, and the failure rate influence area corresponding to the failure rate coefficient curve; Perform internal partitioning on the obtained influence areas to obtain characteristic influence blocks; The internal partitioning means equally dividing the area of the obtained influence areas into four equal parts, that is, obtaining four characteristic influence blocks with the same area; Perform block accounting on the characteristic influence blocks according to the obtained multi-dimensional influence coefficients to obtain quality influence blocks, and the quality influence blocks include the first quality block and the second quality block; It should be further noted that in the specific implementation process, the process of block accounting includes: According to the operating throughput, request response parameters, defect resolution time, and failure rate included in the comprehensive influence parameters, divide the request response parameters, defect resolution time, and failure rate into one group, denoted as the first influence set, and denote the operating throughput as the second influence set; According to each comprehensive influence parameter corresponding to four characteristic influence blocks, select one characteristic influence block as a representative respectively, and jointly form the first quality block, that is, arbitrarily select one from the characteristic influence blocks obtained by the request response parameters, arbitrarily select one from the characteristic influence blocks obtained by the defect resolution time, and arbitrarily select one from the characteristic influence blocks obtained by the failure rate. These three characteristic response blocks are jointly combined to form the first quality block of the three blocks, that is, the area included in the first quality block is the sum of the areas of the three characteristic influence blocks; Select one block from the characteristic response blocks obtained by the operating throughput as the second quality block; Classify and sort the obtained quality influence blocks to obtain a quality influence set, and the quality influence set includes the first quality set and the second quality set; Furthermore, the process of classification and sorting includes: According to each software having a corresponding quality influence block, calculate the area of the obtained quality influence blocks to obtain quality influence weights, that is, the size of the quality influence weights is equal to the area of the quality influence blocks, and the quality influence weights include the first quality weight and the second quality weight; Sort the first quality weights in ascending order to obtain the first quality set, and mark the ranking of each first quality weight in the first quality set. Sort the second quality weights in descending order to obtain the second quality set, and mark the ranking of each second quality weight in the second quality set; In particular, for the comprehensive impact parameter of each software, the smaller the request response parameter, the better the software quality; the smaller the defect resolution time, the better the software quality; the smaller the failure rate, the better the software quality; and the larger the operating throughput, the better the software quality; Estimate the weight of the software through the quality impact set to obtain the comprehensive quality weight, and associate the obtained comprehensive quality weight with the corresponding software; The weight estimation means that in the quality impact set, obtain the sorting of the first quality weights and the sorting of the second quality weights corresponding to the same software, obtain the ranking of the sorting, sum up the rankings of the first quality weights and the second quality weights to obtain the comprehensive quality weight, which represents the quality evaluation weight of software development. The smaller the value of the comprehensive quality weight, the better the quality of software development; Furthermore, according to the fact that each developed software has a corresponding comprehensive quality weight, sort the obtained comprehensive quality weights in ascending order to obtain the quality weight sequence, which is the quality score sorting for evaluating the software. Perform quality level division through the quality weight sequence. Record the evaluation quality of the software development ranked in the top m as the first quality level, record the evaluation quality of the software development ranked from the (m + 1)-th to the n-th as the second quality level, and record the evaluation quality of the software development ranked from the (n + 1)-th to the p-th as the unqualified quality level, where m < n < p. Upload the software operation code of the software corresponding to the unqualified quality level to the management center, and issue a modification instruction through the management center to modify and optimize the code of the software with the unqualified quality level to improve the quality level of software development.

[0021] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the present invention to the specific implementation manners. Obviously, many modifications and variations can be made according to the content of this specification. These embodiments are selected and specifically described in this specification to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and utilize the present invention well. The present invention is only limited by the claims and their full scope and equivalents.

Claims

1. A software development quality assessment system based on big data, including a management center, characterized in that: The management center is connected to a software acquisition module, a development processing module, a quality analysis module and an intelligent estimation module; The software acquisition module is used to acquire the software running environment and software running code; The development processing module is used to construct a simulated virtual space, simulate the software running code through the simulated virtual space, interactively execute the software running code and collect software interactive data, and screen the data based on the interactive execution process to obtain comprehensive software information; The quality analysis module is used to extract and classify software interaction data and software comprehensive information to obtain comprehensive impact parameters, perform modal conversion on the comprehensive impact parameters to obtain comprehensive impact signals, separate and transform the comprehensive impact signals by setting screening tracking coefficients to obtain tracking coefficient components, and perform multi-dimensional extraction on the comprehensive impact signals according to the tracking coefficient components to obtain multi-dimensional impact coefficients; The intelligent estimation module is used to convert the multidimensional influence coefficient into a digital graph, obtain a dynamic graph of the influence coefficient, set a moving central axis, sort the endpoints of the dynamic graph of the influence coefficient, obtain a coefficient component value sequence, perform region conversion on the coefficient component value sequence, obtain the influence area, judge the quality of the software through the influence area, and obtain a comprehensive quality weight.

2. According to the big data-based software development quality assessment system of claim 1, it is characterized in that: The process of interactively executing software running code and collecting software interaction data includes: Construct a simulated virtual space according to the software running environment, and set a node monitoring terminal based on the simulated virtual space; Uploading the software running code to the simulated virtual space, issuing a start instruction to the obtained software running code, and simulating the software running code through the simulated virtual space based on the start instruction; Based on the simulation running process, a user interaction instruction is issued to the obtained software running code, the software running code is interactively executed according to the user interaction instruction, and the interactive execution process is interactively collected to obtain software interaction data.

3. A software development quality assessment system based on big data according to claim 2, characterized in that: The process of data screening based on interactive execution process includes: The node monitoring terminal collects data on the software running code after interactive execution to obtain comprehensive software information; Perform defect screening on the obtained comprehensive software information to obtain comprehensive fault information; The software comprehensive information is scalared and counted based on the software interaction data to obtain the operation throughput.

4. A software development quality assessment system based on big data according to claim 3, characterized in that: The process of extracting and classifying software interaction data and software comprehensive information by the quality analysis module includes: Obtain software interaction data, match and correlate the software interaction data, and obtain request response parameters; Screen and extract software comprehensive information according to comprehensive fault information to obtain fault effect coefficient; The obtained operation throughput, request response parameter and fault effect coefficient are recorded as comprehensive impact parameters.

5. The software development quality assessment system based on big data according to claim 1, characterized in that: The process of separating and transforming the comprehensive impact signal by setting the screening tracking coefficient includes: Set the screening tracking coefficient, perform attribute transformation on the screening tracking coefficient, and obtain the tracking component; Dynamically separate the screening tracking coefficient according to the obtained tracking components to obtain dynamic components; The screening tracking coefficient is separated and intercepted according to the obtained tracking component and dynamic component to obtain the tracking coefficient component.

6. The software development quality assessment system based on big data according to claim 1, characterized in that: The process of multi-dimensional extraction of the comprehensive impact signal according to the tracking coefficient components includes: The obtained tracking coefficient component is uploaded to the comprehensive influence signal, and the comprehensive influence signal is reconstructed by interval through the tracking coefficient component to obtain the influence coefficient component; Performing reconstruction statistics on the obtained influence coefficient components to obtain a reconstruction influence coefficient; The reconstruction influence coefficient is reconstructed and extracted according to the screening tracking coefficient to obtain the multidimensional influence coefficient.

7. The software development quality assessment system based on big data according to claim 1, characterized in that: The process of sorting endpoints of the influence coefficient dynamic graph includes: Construct an original dynamic graph based on the multidimensional influence coefficient, upload the multidimensional influence coefficient to the original dynamic graph, perform digital-to-graphic conversion on the multidimensional influence coefficient based on the original dynamic graph, and obtain the influence coefficient dynamic graph; Set a moving central axis, upload the moving central axis to the influence coefficient dynamic graph, locate the endpoints of the influence coefficient dynamic graph through the moving central axis, and obtain the endpoint extreme value; The influence coefficient dynamic diagram is positioned and sorted by moving the central axis to obtain the coefficient component value sequence.

8. A software development quality assessment system based on big data according to claim 7, characterized in that: The process of determining the quality of software by impact area includes: Perform internal positioning on the coefficient component value sequence to obtain the end component value, perform regional screening on the endpoint extreme value and the segment component value to obtain the impact area; The impact area is partitioned internally to obtain characteristic impact blocks, and the characteristic impact blocks are divided into blocks according to the multi-dimensional impact coefficient to obtain quality impact blocks; The obtained quality impact blocks are classified and sorted to obtain the quality impact set. The software weight is estimated through the quality impact set to obtain the comprehensive quality weight.

Images