CN112416783B - Method, device, equipment and storage medium for determining software quality influence factors - Google Patents

Abstract

The present application relates to a method, device, computer equipment and storage medium for determining the influencing factors of software quality. According to the historical parameter values of the influencing factors of target software quality, a method for predicting the quality of target software based on the parameter values of the influencing factors of target software quality is constructed. The defect prediction model for the number of software defects, and then the preset optimization algorithm model is used to obtain the minimum value of the number of software defects output by the defect prediction model. According to the minimum value of the number of software defects, the target parameter value of the influencing factors of the target software quality is determined. . The method uses the target parameter value as the test guidance data of the target software to test the quality of the target software, which can ensure that the defect of the target software is the least. Adjustment, effectively reduce the cost of testing and improve the quality of software systems.

Description

Software quality influencing factor determination method, apparatus, equipment and storage medium

technical field

The present application relates to the field of computer technology, and in particular, to a method, apparatus, device and storage medium for determining software quality influencing factors.

Background technique

Software testing refers to the process of using manual or automatic means to run or measure the quality of a piece of software, the purpose of which is to check whether the software meets specified requirements or the difference between expected and actual results.

At present, the preparatory work for software product testing, including test resource preparation, personnel allocation, and test case preparation, is all based on past work experience or the completion of previous iteration projects as reference data. Each influencing factor of quality is tested one by one to ensure the quality of software products. Although past experience can be used as reference data, due to the complexity of each software system and the differences between functions, there will inevitably be errors based on previous experience as reference data. Adjustment will inevitably cost a certain amount of manpower and material resources.

Therefore, in the preparatory work of the existing software product testing, there is a lack of effective testing guidance data to carry out efficient software product testing.

SUMMARY OF THE INVENTION

Based on this, it is necessary to provide a method, device, device and storage medium for determining software quality influencing factors that can provide effective test guidance data to efficiently test software products.

In a first aspect, the present application provides a method for determining factors affecting software quality, the method comprising:

According to the historical parameter values of the influencing factors of the target software quality, a defect prediction model of the target software is constructed; the defect prediction model is used to predict the corresponding software defect quantity according to the parameter values of the influencing factors of the software quality;

Use a preset optimization algorithm model to obtain the minimum number of software defects output by the defect prediction model;

According to the minimum value of the number of software defects, the target parameter values of the influencing factors of the target software quality are determined.

In one embodiment, the above-mentioned construction of a defect prediction model of the target software according to the historical parameter values of the influencing factors of the target software quality includes:

Obtain the historical parameter values of the influencing factors and the number of historical software defects of the software version to which each historical parameter value belongs;

Determine the model sample data set from the historical parameter values of the influencing factors and the number of historical software defects;

Build a defect prediction model from the model sample dataset.

In one of the embodiments, the above-mentioned determination of the model sample data set from the historical parameter values of the influencing factors and the number of historical software defects includes:

Perform correlation analysis on the influencing factors to obtain the correlation value between the influencing factors and the number of software defects;

Determine the influence factor whose correlation value is greater than the preset threshold as the sample influence factor;

The historical parameter values of the sample influencing factors and the number of historical software defects of the software version to which the sample influencing factors belong are determined as the model sample data set.

In one embodiment, the above-mentioned constructing a defect prediction model according to the model sample data set includes:

Taking the historical parameter values of the sample influencing factors as input parameters, and using the historical software defect numbers of the software version to which the sample influencing factors belong as the model output, the initial network model is iteratively trained until the preset iterative training termination conditions are met, and the defect prediction model is obtained.

In one embodiment, the above-mentioned iterative training termination condition includes that the number of iterative training reaches a preset number of iterations, or the error between the output value of the initial network model and the preset standard value is less than a preset error.

In one embodiment, the above-mentioned use of a preset optimization algorithm model to obtain the minimum value of the number of software defects output by the defect prediction model includes:

The defect prediction model is used as the fitness function of the optimization algorithm model; the optimization algorithm model includes the range and moving speed of the decision variables, the decision variables are the influencing factors of software quality, and the parameter values are the influencing factors of non-fixed values;

Initialize the optimization parameters of the optimization algorithm model;

Based on the initialized optimization parameters, the range and moving speed of the decision variables are updated iteratively until the iteration converges, and the minimum number of software defects is obtained.

In one of the embodiments, the above-mentioned optimization parameters include an initial global best position and an initial individual best position;

Based on the initialized optimization parameters, the range and moving speed of the decision variables are iteratively updated until the iteration converges, and the minimum number of software defects is obtained, including:

Based on the initial global best position and the initial individual best position, iteratively update the range and moving speed of the decision variable until the iteration converges, and the target global best position is obtained;

When the target global best position is obtained, the output value in the defect prediction model is determined as the minimum value of the number of software defects.

In one embodiment, based on the initial global best position and the initial individual best position, the range and moving speed of the decision variable are iteratively updated until the iteration converges, and the target global best position is obtained, including:

Using the initial global best position and the initial individual best position as the initial values, update the range and moving speed of the decision variable;

According to the range and moving speed of the updated decision variable, calculate the output value of each updated decision variable in the defect prediction model, and compare all output values to determine the current global best position;

If the preset iterative convergence condition is reached, determine the current global best position as the target global best position;

If the iterative convergence condition is not reached, continue to obtain the next global best position until the iterative convergence condition is reached, and the target global best position is obtained.

In one of the embodiments, the above-mentioned optimization parameters further include an initial iteration index, a maximum number of iterations, an initial stop index, and a maximum stop index;

Then the iteration convergence condition includes: the initial iteration index reaches the maximum number of iterations, or the initial stop index reaches the maximum stop index.

In a second aspect, an embodiment of the present application provides a software quality influencing factor determination device, the device comprising:

a building module for constructing a defect prediction model of the target software according to the historical parameter values of the influencing factors of the target software quality; the defect prediction model is used to predict the corresponding software defect quantity according to the parameter values of the influencing factors of the software quality;

an optimization module, configured to use a preset optimization algorithm model to obtain the minimum value of the number of software defects output by the defect prediction model;

The determining module is configured to determine the target parameter value of the influencing factor of the target software quality according to the minimum value of the software defect quantity.

In a third aspect, an embodiment of the present application provides a computer device, including a memory and a processor, where the memory stores a computer program, and the processor implements the steps of any one of the methods provided in the first embodiment of the first aspect when the processor executes the computer program.

In a fourth aspect, embodiments of the present application provide a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, implements the steps of any one of the methods provided in the embodiments of the first aspect.

A method, device, computer equipment, and storage medium for determining a software quality influencing factor provided by the embodiments of the present application, by constructing a parameter value prediction based on the influencing factor of the target software quality according to the historical parameter value of the influencing factor of the target software quality The defect prediction model of the number of software defects of the target software, and then the preset optimization algorithm model is used to obtain the minimum number of software defects output by the defect prediction model. According to the minimum number of software defects, the influencing factors of the target software quality are determined. target parameter value. In this method, the defect prediction model is constructed for the target software, different software products are constructed with different defect prediction models, and a model specially used to predict the number of software defects of the target software according to the parameter values of the influencing factors of the target software quality is constructed. On this basis, the defect prediction model is optimized by the preset optimization algorithm model, the minimum value of the number of software defects output by the defect prediction model is obtained, and the input parameter corresponding to the minimum value is determined as the target parameter value of the target influence parameter, In this way, using the target parameter value as the test guide data of the target software to perform quality testing on the target software can ensure that the target software has the fewest defects. Refer to the target parameter value to prepare for the pre-test work, adjust the unreasonable parts of the pre-test work, effectively reduce the test cost and improve the quality of the software system.

Description of drawings

1 is an application environment diagram of a method for determining a software quality influencing factor provided by an embodiment;

2 is a schematic flowchart of a method for determining a software quality influencing factor provided by an embodiment;

3 is a schematic flowchart of a method for determining a software quality influencing factor provided by another embodiment;

4 is a schematic flowchart of a method for determining a software quality influencing factor provided by another embodiment;

5 is a schematic flowchart of a method for determining a software quality influencing factor provided by another embodiment;

6 is a schematic flowchart of a method for determining a software quality influencing factor provided by another embodiment;

7 is a flowchart of a method for determining a software quality influencing factor provided by an embodiment;

8 is a structural block diagram of an apparatus for determining software quality influencing factors provided by an embodiment;

FIG. 9 is an internal structure diagram of a computer device in an embodiment.

Detailed ways

In order to make the purpose, technical solutions and advantages of the present application more clearly understood, the present application will be described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application, but not to limit the present application.

A method for determining a software quality influencing factor provided by an embodiment of the present application can be applied to an application environment as shown in FIG. 1 , where a processor of a computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium, an internal memory. The nonvolatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the execution of the operating system and computer programs in the non-volatile storage medium. The database of the computer equipment is used to store the data in the process of determining the software quality influencing factors. The network interface of the computer device is used to communicate with other external devices through a network connection. The computer program, when executed by the processor, implements a software quality influencing factor determination method.

Embodiments of the present application provide a method, apparatus, computer device, and storage medium for determining software quality influencing factors, which can provide effective test guidance data for efficient software product testing. The technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems will be specifically described in detail below with reference to the accompanying drawings. The following specific embodiments may be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. It should be noted that, in a method for determining a software quality influencing factor provided by the present application, the execution subject of FIG. 2-FIG. 7 is a computer device, wherein the execution subject may also be a software quality influencing factor determination device, wherein the device can pass Software, hardware, or a combination of software and hardware is implemented as part or all of a computer device.

In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments It is a part of the embodiments of the present application, but not all of the embodiments.

In one embodiment, FIG. 2 provides a method for determining influencing factors of software quality. This embodiment involves a computer device, after constructing a defect prediction model of the target software according to the historical parameter values of the influencing factors of the target software quality, using a pre-defined The set optimization algorithm model obtains the minimum value of the number of software defects output by the defect prediction model, so as to determine the specific process of the target parameter value of the influencing factor of the target software quality based on the minimum value, as shown in Figure 2, the method includes: :

S101 , construct a defect prediction model of the target software according to the historical parameter values of the influencing factors of the target software quality; the defect prediction model is used to predict the corresponding software defect quantity according to the parameter values of the influencing factors of the software quality.

The target software refers to any software that currently needs to determine the quality influencing factor; the parameter value of the quality influencing factor finally determined for the target software in this embodiment is an effective test for performing efficient software testing on the target software. guidance data. Among them, the influencing factors of software quality can also be understood as the influencing factors that form software defects, such as test time, personnel allocation, software and hardware testing resources, number of requirements, function point coverage, test strategy coverage, test case coverage, Test case granularity, number of test cases, number of lines of code, degree of difference between test environments, number of integrated systems, etc.

In practical applications, many service platforms have integrated more functions of software products. For example, smart medical cloud software products have many integrated functions, and there are many defects in the development process, and effective prediction calculation defects are correct. The basis of software defect optimization, so software defects can be predicted by establishing a software defect prediction model. A defect prediction model refers to a model that can predict the number of software defects according to the parameter values of any software quality influencing factors. The model can be a neural network model, for example, a back propagation (back propagation, BP) neural network, etc., and the BP neural network has the characteristics of strong fitting ability, which can realize effective calculation of software defects.

Generally, each software will go through multiple versions of progression, and each version of the software will have certain defects. Therefore, the historical parameter values of the quality influencing factors of the target software can be collected, that is, the parameter values of the influencing factors of each historical version. The parameter value includes specific input parameters and the number of defects of the software, and then the above-mentioned defect prediction model can be constructed according to these historical parameter values. The input parameters of the successfully constructed defect prediction model are the influencing factors of any combination of the above influencing factors, and the corresponding output is the number of defects corresponding to the influencing factors of these combinations.

S102, using a preset optimization algorithm model to obtain the minimum value of the number of software defects output by the defect prediction model.

The optimization algorithm model refers to the optimization model of the defect prediction model, which is used to minimize the number of software defects output by the defect prediction model, so as to improve the software quality of the defect prediction model.

Optionally, the optimization algorithm model can be established through particle swarm optimization (Particle swarm optimization, PSO). The PSO algorithm has strong global optimization ability. Since the defect prediction model established above has many influencing factors as input parameters, there is still a complex relationship between these influencing factors. However, the PSO algorithm has fast search speed and high efficiency. The advantages of simplicity, no adjustment of many parameters, and suitability for real-valued processing can lead to better guiding results when optimizing the above defect prediction model.

Specifically, the optimization algorithm model is used to continuously iterate the defect prediction model to obtain the minimum number of defects output by the defect prediction model. In the optimization algorithm, the decision variables of the defect prediction model are set according to the above influencing factors, and the corresponding decision variable parameters are obtained. is the optimal solution to be obtained.

其中，y_o表示的是缺陷预测模型的输出值，即缺陷数量，BPNN指代的是以BP神经网络构建的缺陷预测模型，那么该优化过程即为求取缺陷数量最小值miny_o。这样通过优化算法模型对缺陷预测模型进行优化，求取上述公式(1)中的y_o的最小值。For example, the optimization process can be described as the formula:

Among them, _yo represents the output value of the defect prediction model, that is, the number of defects, and BPNN refers to the defect prediction model constructed by BP neural network, then the optimization process is to obtain the minimum number of defects, miny _o . In this way, the defect prediction model is optimized through the optimization algorithm model, and the minimum value of _yo in the above formula (1) is obtained.

Among them, decision variables refer to influencing factors with non-fixed values, so among the above-determined influencing factors, before the specific optimization process, if the parameters have been determined, they will not be used as decision-making variables here, for example, the requirements in the above-mentioned influencing factors The number and the number of integrated systems, that is, the actual value in the optimization algorithm model, are not used as decision variables. For the influencing factors whose unit is a percentage, the parameter range can be set to 0-100%, for example, test case coverage, function point coverage; when the remaining influencing factors are used as decision variables, the limit range is set to be greater than 0, so as to avoid When the range of decision variables is not set, the optimization process is endless and the iterative convergence cannot be achieved.

S103, according to the minimum value of the software defect quantity, determine the target parameter value of the influencing factor of the target software quality.

After obtaining the minimum value of the number of software defects by optimizing the algorithm model in the above steps, the input parameter of the defect prediction model corresponding to the minimum value is directly determined as the target parameter value of the target influence parameter of the target software. Equivalently, using the target parameter value as the test guide data of the target software to perform quality testing on the target software can ensure that the target software has the fewest defects.

In the method for determining software quality influencing factors provided by this embodiment, by constructing a defect prediction model for predicting the number of software defects of the target software according to the parameter values of the influencing factors of the target software quality according to the historical parameter values of the influencing factors of the target software quality, Then, using the preset optimization algorithm model, the minimum value of the software defect quantity output by the defect prediction model is obtained, and the target parameter value of the influencing factor of the target software quality is determined according to the minimum value of the software defect quantity. In this method, the defect prediction model is constructed for the target software, different software products are constructed with different defect prediction models, and a model specially used to predict the number of software defects of the target software according to the parameter values of the influencing factors of the target software quality is constructed. On this basis, the defect prediction model is optimized by the preset optimization algorithm model, the minimum value of the number of software defects output by the defect prediction model is obtained, and the input parameter corresponding to the minimum value is determined as the target parameter value of the target influence parameter, In this way, using the target parameter value as the test guide data of the target software to perform quality testing on the target software can ensure that the target software has the fewest defects. Refer to the target parameter value to prepare for the pre-test work, adjust the unreasonable parts of the pre-test work, effectively reduce the test cost and improve the quality of the software system.

Taking the smart medical cloud software as an example, in practical applications, this method can effectively optimize the parameter values of the influencing factors involved in the preparatory work of the smart medical cloud software project, and obtain the best parameter values for the target software test. . The optimal parameter value has an effective guide for the preliminary work of the software product testing work. For example, during the pre-allocation of resources, the formulation of test plans, and the design of test cases, the optimal parameter value can be referred to to prepare for the pre-test work, and test Adjust the unreasonable parts of the previous work, effectively reduce the test cost and improve the quality of the software system, ensure the high-quality launch and delivery of smart medical cloud software products, and meet the needs of use.

On the basis of the above embodiments, the embodiments of the present application provide an embodiment of a method for determining influencing factors of software quality, which involves constructing defect predictions of target software according to historical parameter values of influencing factors of target software quality The specific process of the model, as shown in Figure 3, the above step S101 includes:

S201: Obtain historical parameter values of the influencing factors and the number of historical software defects of the software version to which each historical parameter value belongs.

When acquiring the historical parameter values of the influencing factors of the target software, it is necessary to acquire the number of historical software defects of the software version to which each historical parameter value belongs. When obtaining historical parameter values, for service platforms that integrate more functions of software products, each software product is interdependent and can be used in an integrated manner. When constructing defect prediction models for these software products, it can collect The historical parameter values of any software product in the service platform are used as a reference. For example, there are many smart medical cloud products, and each product is interdependent and can be used in an integrated manner. Therefore, when establishing a defect prediction model for the smart medical cloud, the historical parameters The value can collect the historical development data of any software product of the smart medical cloud, that is, collect the parameter values of the quality influencing factors of each software system of the smart medical cloud when the previous versions were released, thus increasing the diversity of training data for the defect prediction model, and also The integrity of the training data is guaranteed.

The historical parameter values of the above influencing factors and the number of historical software defects of the software version to which each historical parameter value belongs are collected correspondingly to form a historical data set.

S202, a model sample data set is determined from the historical parameter values of the influencing factors and the number of historical software defects.

When building a defect prediction model, it is very important to determine the input parameters of the model. In order to ensure that the final sample data set for training the defect prediction model is more in line with the target software, the final model needs to be selected from the historical data set collected in the above steps. sample dataset.

For example, the correlation between the above-mentioned influencing factors and the number of software defects can be analyzed, for example, by Spearman rank-order correlation coefficient (SRCC) analysis, and then according to the correlation analysis results, The final required model sample data set is determined from the historical parameter values of the influencing factors and the number of historical software defects. The model sample data set may also be determined from the historical parameter values of the influencing factors and the number of historical software defects according to preset weights or big data empirical values. This embodiment of the present application does not limit this.

S203, construct a defect prediction model according to the model sample data set.

After the model sample data set is obtained, a defect prediction model is constructed according to the model sample data set.

For example, it is assumed that there are n main influencing parameters in the model sample data set, namely u ₁ , u ₂ ,...,u _n . That is, the n influencing parameters are determined as the input parameters of the defect prediction model, and the corresponding number of software defects is used as the output to train the initial defect prediction model. The initial BP neural network model can be used for training, so that due to the complex relationship between various influencing factors, for example, the number of requirements and the number of lines of code or the number of use cases and the granularity and coverage of use cases; the more requirements, the code will also The more corresponding; the more use cases, the more granularity and coverage of use cases. There is a strong coupling between the influencing parameters of software products, and it is difficult for general mathematical models to meet the requirements of modeling accuracy, while BP neural network has strong nonlinear function fitting ability, which can solve the problem of lack of software modeling.

For example, the defect prediction model (BP neural network) is:

In the above formula (2): {u _i , i=1, 2, L n} are the input parameters of the defect prediction model, where n is the number of main influencing parameters in the model sample data set; m is the hidden value in the neural network The number of layer neurons, which is generally obtained by trial and error; w _ij is the weight between the i-th input neuron and the j-th hidden layer neuron in the neural network; w _jo is the j-th neuron in the neural network The weight between the hidden layer neuron and the output neuron; b _j and b _o are the thresholds of the hidden layer and output layer in the neural network respectively; _yo is the output value of the defect prediction model and the number of software defects.

During training, the model sample data set can be divided into training sample data set and prediction data, use the training sample data set to train the initial defect prediction model, obtain better thresholds and weights through continuous iterative convergence, and obtain the final defect prediction Model. Specifically, the number of iterations can be set to M, the iteration error target can be set to 0.01, and the model is trained by back-propagation algorithm to obtain w _jo , w _ij , b _j and b _o . Optionally, the iterative training termination condition includes that the number of iterative training reaches a preset number of iterations, or the error between the output value of the initial network model and the preset standard value is less than a preset error.

The training process is as follows: (1) The input layer is responsible for receiving the input parameters u _i , and passing the parameter values to the neurons in the middle hidden layer; the hidden layer is responsible for internal information processing and exchange, and the hidden layer is passed to the neurons in the output layer The information, after processing, completes a learning forward propagation process, output y _o by the output layer; (2) when the value of the following formula (3) is greater than the preset threshold, such as 0.01, the back propagation of the error stage. The error passes through the output layer, corrects the weights between layers according to the method of error gradient descent, and transfers back to the hidden layer and the input layer layer by layer. The process of forward propagation and error back propagation is continuously performed, and the weights of each layer are continuously adjusted. This is the process of learning and training of BP neural network. When the number of iterations is greater than M or the value of formula (3) is less than 0.01, the training stops. w _jo , w _ij , b _j and _bo are the thresholds and weights of the final training model.

Among them, E in formula (3) represents the error, _yo is the output value of the defect prediction model, the number of software defects, and y is the real number of software defects in the historical version of the software. Since the number of software defects in historical software versions is the real data that has occurred, the acquired historical software defect numbers can be used as standard data to verify the accuracy of the defect prediction model.

In this embodiment, by obtaining the historical parameter values of the influencing factors and the historical software defect quantity of the software version to which each historical parameter value belongs, the model sample data set is determined from the historical parameter values of the influencing factors and the historical software defect quantity, and then according to the model samples The dataset builds a defect prediction model. The defect prediction model to be constructed belongs to the target software, and the sample data set used is also selected from the historical parameter values of the influencing factors of the target software and the number of historical software defects of the software version to which each historical parameter value belongs. The product builds different defect prediction models to ensure that the target parameter values to be determined are more suitable for the target software, and then the target parameter values are used as a reference for the preparatory work, which can greatly reduce costs and ensure software product quality.

As shown in FIG. 4 , in one embodiment, taking correlation analysis as an example, the above process of determining the model sample data set from the historical parameter values of the influencing factors and the number of historical software defects is described in detail. This embodiment includes:

S301, perform a correlation analysis on the influencing factors, and obtain a correlation value between the influencing factors and the number of defects of the software.

First, normalize the historical data set collected in the above step S201 to remove the differences between the data. After normalization, the correlation analysis is carried out on each influencing factor in the historical data set and the corresponding number of software defects.

The historical data sets collected above are collected from the historical parameter values of the influencing factors and the number of historical software defects of the software version to which each historical parameter value belongs. Each version corresponds to a set of data, so when calculating the correlation, each version The correlation between each influencing factor (parameter value) and the number of software defects in this version is calculated separately, and the correlation of each influencing factor is obtained.

Specifically, taking SRCC as an example, the correlation calculation process is as follows: assuming that the data sequences of parameters a and b are [a ₁ , a ₂ , K, a _n ] and [b ₁ , b ₂ , K, b _n ] respectively, then The SRCC correlation between parameters a and b is:

和

为平均值，δ为参数a和b相关值，范围为[-1，1],δ为相关性值，其绝对值越大表示相关性越大。Among them, in formula (4), a _i and b _i are actual values,

and

is the average value, δ is the correlation value of parameters a and b, the range is [-1, 1], δ is the correlation value, and the larger the absolute value, the greater the correlation.

For any influencing factor, the influencing factor and the number of software defects of the version to which the influencing factor belongs are taken as parameters a and b in formula (4), respectively, then through formula (4), the influencing factor and the number of software defects to which the influencing factor belongs can be obtained. The correlation between the number of software defects in the version. According to this process, the correlation value of each influencing factor is calculated.

S302: Determine the influence factor whose correlation value is greater than the preset threshold as the sample influence factor.

After the correlation value of each influencing factor is obtained, the influencing factor whose correlation value is greater than the preset threshold is determined as the sample influencing factor. For example, the influencing factors with a correlation greater than 0.5 are retained as the sample influencing factors, and the influencing factors with a correlation less than 0.5 are eliminated.

S303: Determine the historical parameter values of the sample influencing factors and the historical software defect quantity of the software version to which the sample influencing factors belong as a model sample data set.

After the sample influencing factors are obtained, the historical parameter values of each sample influencing factor and the historical software defect quantity of the software version to which the sample influencing factors belong are collected correspondingly to form the final model sample data set. Optionally, the historical parameter values of the sample influencing factors are used as input parameters, and the number of historical software defects of the software version to which the sample influencing factors belong is used as the model output, and the initial network model is iteratively trained until a preset iterative training termination condition is met, and a defect prediction is obtained. Model.

In this embodiment, by performing correlation analysis on the influencing factors, the correlation value between the influencing factor and the number of defects in the software is obtained, and the influencing factor whose correlation value is greater than the preset threshold is determined as the sample influencing factor, and then the sample influence factor is determined. The historical parameter values of the factors and the number of historical software defects of the software version to which the sample influencing factors belong are determined as the model sample data set. In this way, the accurate correlation value is used as the basis for screening the model sample data set, and the quantitative result makes the screening result more accurate.

An embodiment is provided below to describe the above process of using a preset optimization algorithm model to obtain the minimum value of the number of software defects output by the defect prediction model. As shown in FIG. 5 , in one embodiment, the above S102 includes the following: step:

S401, the defect prediction model is used as the fitness function of the optimization algorithm model; the optimization algorithm model includes the range and moving speed of the decision variables, the decision variables are the influencing factors of software quality, and the parameter values are non-fixed value influencing factors.

The established defect prediction model is added to the optimization algorithm (particle swarm optimization algorithm) as the fitness function of the optimization algorithm model, and the range and moving speed of the decision variables are set in the optimization algorithm, where the decision variables represent the influencing factors of software quality. The parameter value of is a non-fixed value influencing factor.

For example, after taking the defect prediction model as the fitness function of the optimization algorithm model, the formula including the decision variables is obtained as follows:

V(k+1)=ψ(k)V(k)+c ₁ r ₁ [pbest(k)-Z(k)]+c ₂ r ₂ [gbest(k)-Z(k)] (5)

Z(k+1)=Z(k)+V(k+1) (6)

Among them, in formula (5), c ₁ and c ₂ are iterative factors, usually taking the value of c ₁ =c ₂ =2; gbest(k) is the global best position, pbest(k) is the individual best position, Z =[z ₁ ,z ₂ ,...,z _D ] ^T is the range of decision variables, and the size range of _zi is determined according to the range of the set decision variables; V=[v ₁ ,v ₂ ,...,v _D ] ^T is The moving speed of the decision variable, usually v _i ∈ [-0.5, 0.5]; ψ is the weighting coefficient, where the weighting coefficient can be determined according to formula (7), usually, ψ _max and ψ _min in formula (7) are 0.9 and 0.4, respectively , k _Imax is the maximum iteration parameter.

S402, initialize the optimization parameters of the optimization algorithm model.

Based on the above functions to be optimized, before starting to update the range and moving speed of the decision variables, the optimization parameters of the optimization algorithm model need to be initialized. Among them, the optimization parameters include iteration index, maximum number of iterations, stop index, maximum stop index, global best position, and individual best position; wherein, the initial value of the iteration index is k=0, and in each iteration, the iteration index is incremented by 1, Until the maximum number of iterations k _Imax is reached; wherein, the initial value of the stop index is n _s =0, and the stop index is incremented by 1 for each stop until the stop index reaches the maximum stop index N _Smax ; Similarly, the initial global best position is: gbest(0 ), and the initial individual best position is pbest(0). Then, to initialize the above optimization parameters is to classify all the optimization parameters as initial values.

S403 , based on the initialized optimization parameters, iteratively update the range and moving speed of the decision variable until the iteration converges, and obtain the minimum value of the number of software defects.

After initializing the optimization parameters, the process of iteratively updating the range and moving speed of the decision variables is started until the iteration converges and the minimum number of software defects is obtained. At this time, the range and moving speed of the decision variables corresponding to the convergence of the iteration are the influence The optimal value of the parameter value for the factor.

Optionally, based on the initial global best position and the initial individual best position, iteratively update the range and moving speed of the decision variable until iteratively converges to obtain the target global best position; when the target global best position is obtained, the defect prediction model The output value in is determined as the minimum number of software defects.

The range and moving speed of the decision variables are updated for the first time with the initial global best position and the initial individual best position, and the latest global best position and individual best position are determined. After the first update, it is detected whether the iterative convergence condition is reached, If not, continue to update the range and moving speed of the decision variable according to the latest global best position and individual best position until the iterative convergence condition is reached, determine the global best position obtained during the iteration convergence as the target global best position, and The output value in the defect prediction model at this time is determined as the minimum value of the number of software defects.

In this embodiment, the defect prediction model is used as the fitness function of the optimization algorithm model. The optimization algorithm model includes the range and moving speed of the decision variable, and the decision variable is the influence factor of the software quality. The parameter value is the influence of the non-fixed value. factor, and then initialize the optimization parameters of the optimization algorithm model, and based on the initial global best position and the initial individual best position, iteratively update the range and moving speed of the decision variables until the iteration converges to obtain the target global best position, and finally the target will be obtained. When the global best position is reached, the output value in the defect prediction model is determined as the minimum value of the number of software defects. Because in this method, the defect prediction model is optimized by an optimization algorithm, and the target parameter value of the influencing factors of the target software is obtained by calculating the minimum value of the output of the defect prediction model. Through the target parameter value, the preparatory work of the target software test can be carried out. Reasonable planning and resource allocation, adjust the unreasonable parts of the pre-test work, reduce the test cost of software test products, and improve the quality of software products.

In one embodiment, as shown in FIG. 6 , the above-mentioned process of iteratively updating the range and moving speed of the decision variable based on the initial global best position and the initial individual best position until the iteration converges, and obtaining the target global best position includes:

S501, update the range and moving speed of the decision variable with the initial global best position and the initial individual best position as initial values.

In the above optimization parameters, the initial global best position is gbest(0), and the initial individual best position is pbest(0). Starting from the initial value, the range and moving speed of the decision variables are updated.

S502: Calculate the output value of each updated decision variable in the defect prediction model according to the updated range and moving speed of the decision variable, and compare all output values to determine the current global best position.

After updating the range and moving speed of the decision variables, calculate the output value of each updated decision variable in the defect prediction model, compare the output values of these decision variables in the defect prediction model, and determine the new global optimum according to the comparison results. location and individual optimal location. That is, when the initial global best position and the initial individual best position, K=0 in pbest(k) and pbest(k), after the new global best position and the individual best position are determined once updated, pbest(k) and k=1 in pbest(k), then, and so on, update k=2 after the second time until K=k _Imax .

S503, if the preset iterative convergence condition is reached, determine the current global best position as the target global best position; if the iterative convergence condition is not met, continue to obtain the next global best position until the iterative convergence condition is reached, and obtain Target global best position. Optionally, the iteration convergence condition includes: the initial iteration index reaches the maximum number of iterations, or the initial stop index reaches the maximum stop index.

When the above K=k _Imax , it means that the set iterative convergence condition is reached, that is to say, the iterative convergence condition is that the initial iteration index reaches the maximum number of iterations. When the iterative optimization parameter includes a stop index, it can also be judged whether the iterative convergence condition is reached by the initial stop index reaching the maximum stop index. Then after each new global best position is obtained, judge whether the new global best position is equal to the previous global best position, for example gbest(k)=gbest(k-1), if not, stop the index adding 1, _ns = _ns +1, but if they are equal, stop the index and continue to take 0.

When judging whether the iterative convergence condition is reached, k<k _Imax and _ns < _NSmax , if either condition is not met, the iteration is stopped. After stopping the iteration, it is determined that the global best position gbest( _k ) obtained when the iteration is stopped is the target global best position, that is, the minimum value of the software defect yo, and the value of the decision variable corresponding to the target global best position is the target software. The target parameter value of the influencing factor.

In this embodiment, the initial global best position and the initial individual best position are used as initial values, the range and moving speed of the decision variable are updated, and each updated decision variable is calculated according to the range and moving speed of the updated decision variable. The output value in the defect prediction model is compared, and all output values are compared to determine the current global best position. If the preset iterative convergence condition is reached, the current global best position is determined as the target global best position; Continue to obtain the next global best position until the iterative convergence condition is reached, and obtain the target global best position. Taking the initial global best position and the initial individual best position as the initial values, the range and moving speed of the decision variables are repeatedly updated until the update iteration converges, and the target parameter values of the influencing factors of the target software are obtained, so as to obtain the best parameter values. , using the target parameter value as the test guidance data of the target software to test the quality of the target software, which can ensure that the target software has the fewest defects, refer to the best parameter value for the preliminary preparation of the test work, and carry out the pre-test work for the unreasonable places. Adjustments to effectively reduce testing costs and improve software system quality

As shown in FIG. 7 , an embodiment of a method for determining a software quality influencing factor is also provided, and the embodiment includes:

S1. Obtain historical parameter values of the influencing factors and the number of historical software defects of the software version to which each historical parameter value belongs.

S2. Perform a correlation analysis on the influencing factors, and obtain the correlation value between the influencing factors and the number of defects of the software.

S3. Determine the influence factor whose correlation value is greater than the preset threshold as the sample influence factor.

S4. Determine the historical parameter value of the sample influencing factor and the historical software defect quantity of the software version to which the sample influencing factor belongs as the model sample data set.

S5. Build a defect prediction model according to the model sample data set.

S6. Use the defect prediction model as the fitness function of the optimization algorithm model; the optimization algorithm model includes the range and moving speed of the decision variables, the decision variables are the influencing factors of software quality, and the parameter values are non-fixed value influencing factors.

S7, initialize the optimization parameters of the optimization algorithm model.

S8. Update the range and moving speed of the decision variable with the initial global best position and the initial individual best position as initial values.

S9. Calculate the output value of each updated decision variable in the defect prediction model according to the range and moving speed of the updated decision variable, and compare all the output values to determine the current global best position.

S10. If a preset iterative convergence condition is reached, determine the current global best position as the target global best position.

S11. If the iterative convergence condition is not reached, continue to obtain the next global best position until the iterative convergence condition is reached, and the target global best position is obtained.

S12. When the target global best position is obtained, the output value in the defect prediction model is determined as the minimum value of the number of software defects.

S13, according to the minimum value of the software defect quantity, determine the target parameter value of the influencing factor of the target software quality.

The implementation principles and technical effects of the steps in the method for determining a software quality influencing factor provided by this embodiment are similar to those in the previous embodiments of the software quality influencing factor determining method, and are not repeated here. The implementation manner of each step in the embodiment of FIG. 7 is only an example, and each implementation manner is not limited, and the sequence of each step can be adjusted in practical application, as long as the purpose of each step can be achieved.

It should be understood that although the steps in the flowcharts of FIGS. 2-7 are shown in sequence according to the arrows, these steps are not necessarily executed in the sequence shown by the arrows. Unless explicitly stated herein, the execution of these steps is not strictly limited to the order, and these steps may be performed in other orders. Moreover, at least a part of the steps in FIGS. 2-7 may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed and completed at the same time, but may be executed at different times. These sub-steps or stages are not necessarily completed at the same time. The order of execution of the steps is not necessarily sequential, but may be performed alternately or alternately with other steps or at least a part of sub-steps or stages of other steps.

In one embodiment, as shown in FIG. 8, a software quality influencing factor determination device is provided, including: a building module 10, an optimization module 11 and a determination module 12, wherein,

The building module 10 is used to construct a defect prediction model of the target software according to the historical parameter values of the influencing factors of the target software quality; the defect prediction model is used to predict the corresponding software defect quantity according to the parameter values of the influencing factors of the software quality ;

An optimization module 11, configured to use a preset optimization algorithm model to obtain the minimum value of the number of software defects output by the defect prediction model;

The determining module 12 is configured to determine the target parameter value of the influencing factor of the target software quality according to the minimum value of the software defect quantity.

In one embodiment, the building blocks 10 described above include:

an acquisition unit, used to acquire the historical parameter values of the influencing factors and the number of historical software defects of the software version to which each historical parameter value belongs;

A determination unit for determining a model sample data set from historical parameter values and historical software defect counts of influencing factors;

The building unit is used to build a defect prediction model based on the model sample data set.

In one embodiment, the above-mentioned determining unit includes:

The analysis subunit is used to perform correlation analysis on the influencing factors, and obtain the correlation value between the influencing factors and the number of defects in the software;

The factor determination subunit is used to determine the influence factor whose correlation value is greater than the preset threshold as the sample influence factor;

The data set determination subunit is used to determine the historical parameter value of the sample influencing factor and the historical software defect quantity of the software version to which the sample influencing factor belongs as the model sample data set.

In one embodiment, the above-mentioned optimization module 11 includes:

The function determination unit is used to use the defect prediction model as the fitness function of the optimization algorithm model; the optimization algorithm model includes the range and moving speed of the decision variables, and the decision variables are the influencing factors of software quality. factor;

The parameter initialization unit is used to initialize the optimization parameters of the optimization algorithm model;

The updating unit is used to iteratively update the range and moving speed of the decision variables based on the initialized optimization parameters, until the iteration converges, and the minimum value of the number of software defects is obtained.

In one embodiment, the above-mentioned optimization parameters include an initial global optimal position and an initial individual optimal position; then the above-mentioned updating unit includes:

The update subunit is used to iteratively update the range and moving speed of the decision variable based on the initial global best position and the initial individual best position, until the iteration converges, and the target global best position is obtained;

The minimum value determination subunit is used to determine the output value in the defect prediction model as the minimum value of the software defect quantity when the target global optimal position is obtained.

In one embodiment, the above-mentioned update subunit is specifically used to update the range and moving speed of the decision variable with the initial global best position and the initial individual best position as initial values; according to the updated range and moving speed of the decision variable , calculate the output value of each updated decision variable in the defect prediction model, and compare all output values to determine the current global best position; if the preset iterative convergence condition is reached, determine the current global best position is the target global best position; if the iterative convergence condition is not reached, continue to obtain the next global best position until the iterative convergence condition is reached, and the target global best position is obtained.

In one embodiment, the above optimization parameters further include an initial iteration index, a maximum number of iterations, an initial stop index, and a maximum stop index; then the iteration convergence conditions include: the initial iteration index reaches the maximum number of iterations, or the initial stop index reaches the maximum stop index.

For the specific limitation of the software quality influencing factor determination device, please refer to the above limitation on the software quality influencing factor determination method, which will not be repeated here. Each module in the above software quality influencing factor determination device may be implemented in whole or in part by software, hardware and combinations thereof. The above modules can be embedded in or independent of the processor in the computer device in the form of hardware, or stored in the memory in the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

In one embodiment, a computer device is provided, and the computer device may be a terminal, and its internal structure diagram may be as shown in FIG. 9 . The computer equipment includes a processor, memory, a network interface, a display screen, and an input device connected by a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium, an internal memory. The nonvolatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the execution of the operating system and computer programs in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal through a network connection. The computer program, when executed by the processor, implements a software quality influencing factor determination method. The display screen of the computer equipment may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer equipment may be a touch layer covered on the display screen, or a button, a trackball or a touchpad set on the shell of the computer equipment , or an external keyboard, trackpad, or mouse.

Those skilled in the art can understand that the structure shown in FIG. 9 is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer equipment to which the solution of the present application is applied. Include more or fewer components than shown in the figures, or combine certain components, or have a different arrangement of components.

In one embodiment, a computer device is provided, including a memory and a processor, a computer program is stored in the memory, and the processor implements the following steps when executing the computer program:

According to the historical parameter values of the influencing factors of the target software quality, a defect prediction model of the target software is constructed; the defect prediction model is used to predict the corresponding software defect quantity according to the parameter values of the influencing factors of the software quality;

Use a preset optimization algorithm model to obtain the minimum number of software defects output by the defect prediction model;

According to the minimum value of the number of software defects, the target parameter values of the influencing factors of the target software quality are determined.

In one embodiment, the processor implements the following steps when executing a computer program:

Obtain the historical parameter values of the influencing factors and the number of historical software defects of the software version to which each historical parameter value belongs;

Determine the model sample data set from the historical parameter values of the influencing factors and the number of historical software defects;

Build a defect prediction model from the model sample dataset.

In one embodiment, the processor implements the following steps when executing a computer program:

Perform correlation analysis on the influencing factors to obtain the correlation value between the influencing factors and the number of software defects;

Determine the influence factor whose correlation value is greater than the preset threshold as the sample influence factor;

The historical parameter values of the sample influencing factors and the number of historical software defects of the software version to which the sample influencing factors belong are determined as the model sample data set.

In one embodiment, the processor implements the following steps when executing a computer program:

The defect prediction model is used as the fitness function of the optimization algorithm model; the optimization algorithm model includes the range and moving speed of the decision variables, the decision variables are the influencing factors of software quality, and the parameter values are the influencing factors of non-fixed values;

Initialize the optimization parameters of the optimization algorithm model;

Based on the initialized optimization parameters, the range and moving speed of the decision variables are updated iteratively until the iteration converges, and the minimum number of software defects is obtained.

In one embodiment, the processor implements the following steps when executing a computer program:

Based on the initial global best position and the initial individual best position, iteratively update the range and moving speed of the decision variable until the iteration converges, and the target global best position is obtained;

When the target global best position is obtained, the output value in the defect prediction model is determined as the minimum value of the number of software defects.

In one embodiment, the processor implements the following steps when executing a computer program:

Using the initial global best position and the initial individual best position as the initial values, update the range and moving speed of the decision variable;

According to the range and moving speed of the updated decision variable, calculate the output value of each updated decision variable in the defect prediction model, and compare all output values to determine the current global best position;

If the preset iterative convergence condition is reached, determine the current global best position as the target global best position;

If the iterative convergence condition is not reached, continue to obtain the next global best position until the iterative convergence condition is reached, and the target global best position is obtained.

In one embodiment, the above-mentioned optimization parameters further include an initial iteration index, a maximum number of iterations, an initial stop index, and a maximum stop index;

Then the iteration convergence condition includes: the initial iteration index reaches the maximum number of iterations, or the initial stop index reaches the maximum stop index.

The implementation principle and technical effect of the computer device provided by the above-mentioned embodiment are similar to those of the above-mentioned method embodiment, which will not be repeated here.

In one embodiment, a computer-readable storage medium is provided on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

According to the historical parameter values of the influencing factors of the target software quality, a defect prediction model of the target software is constructed; the defect prediction model is used to predict the corresponding software defect quantity according to the parameter values of the influencing factors of the software quality;

Use a preset optimization algorithm model to obtain the minimum number of software defects output by the defect prediction model;

According to the minimum value of the number of software defects, the target parameter values of the influencing factors of the target software quality are determined.

In one embodiment, the computer program, when executed by a processor, implements the following steps:

Obtain the historical parameter values of the influencing factors and the number of historical software defects of the software version to which each historical parameter value belongs;

Determine the model sample data set from the historical parameter values of the influencing factors and the number of historical software defects;

Build a defect prediction model from the model sample dataset.

In one embodiment, the computer program, when executed by a processor, implements the following steps:

Perform correlation analysis on the influencing factors to obtain the correlation value between the influencing factors and the number of software defects;

Determine the influence factor whose correlation value is greater than the preset threshold as the sample influence factor;

The historical parameter values of the sample influencing factors and the number of historical software defects of the software version to which the sample influencing factors belong are determined as the model sample data set.

In one embodiment, the computer program, when executed by a processor, implements the following steps:

The defect prediction model is used as the fitness function of the optimization algorithm model; the optimization algorithm model includes the range and moving speed of the decision variables, the decision variables are the influencing factors of software quality, and the parameter values are the influencing factors of non-fixed values;

Initialize the optimization parameters of the optimization algorithm model;

Based on the initialized optimization parameters, the range and moving speed of the decision variables are updated iteratively until the iteration converges, and the minimum number of software defects is obtained.

In one embodiment, the computer program, when executed by a processor, implements the following steps:

Based on the initial global best position and the initial individual best position, iteratively update the range and moving speed of the decision variable until the iteration converges, and the target global best position is obtained;

When the target global best position is obtained, the output value in the defect prediction model is determined as the minimum value of the number of software defects.

In one embodiment, the computer program, when executed by a processor, implements the following steps:

Using the initial global best position and the initial individual best position as the initial values, update the range and moving speed of the decision variable;

According to the range and moving speed of the updated decision variable, calculate the output value of each updated decision variable in the defect prediction model, and compare all output values to determine the current global best position;

If the preset iterative convergence condition is reached, determine the current global best position as the target global best position;

If the iterative convergence condition is not reached, continue to obtain the next global best position until the iterative convergence condition is reached, and the target global best position is obtained.

In one embodiment, the above-mentioned optimization parameters further include an initial iteration index, a maximum number of iterations, an initial stop index, and a maximum stop index;

Then the iteration convergence condition includes: the initial iteration index reaches the maximum number of iterations, or the initial stop index reaches the maximum stop index.

The implementation principle and technical effect of the computer-readable storage medium provided by the above-mentioned embodiments are similar to those of the above-mentioned method embodiments, and details are not described herein again.

Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be implemented by instructing relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage In the medium, when the computer program is executed, it may include the processes of the above-mentioned method embodiments. Wherein, any reference to memory, storage, database or other medium used in the various embodiments provided in this application may include non-volatile and/or volatile memory. Nonvolatile memory may include read only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous chain Road (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, all possible combinations of the technical features in the above embodiments are not described. However, as long as there is no contradiction in the combination of these technical features, all It is considered to be the range described in this specification.

The above-mentioned embodiments only represent several embodiments of the present application, and the descriptions thereof are specific and detailed, but should not be construed as a limitation on the scope of the invention patent. It should be pointed out that for those skilled in the art, without departing from the concept of the present application, several modifications and improvements can be made, which all belong to the protection scope of the present application. Therefore, the scope of protection of the patent of the present application shall be subject to the appended claims.

Claims (10)

1. a software quality influencing factor determination method, is characterized in that, described method comprises: Correlation analysis is performed on the influencing factors of the target software quality, the correlation value between the influencing factor and the number of defects in the software is obtained, and the correlation value, the historical parameter value of the influencing factor of the target software quality and The number of historical software defects of the software version to which each historical parameter value belongs to determine the model sample data set; Based on the model sample data set, a defect prediction model of the target software is constructed; the defect prediction model is used to predict the corresponding software defect quantity according to the parameter values of the influencing factors of software quality; A preset optimization algorithm model is used to obtain the minimum value of the number of software defects output by the defect prediction model; the optimization algorithm model is an optimization model of the defect prediction model; According to the minimum value of the software defect quantity, determine the target parameter value of the influencing factor of the target software quality; in,

; where the parameter a is the influencing factor, the parameter b is the number of software defects of the version to which the influencing factor a belongs,

is the actual value in parameter a,

is the actual value in parameter b,

is the mean value of parameter a,

is the mean value of parameter b,

is the correlation value of parameters a and b. 2. The method according to claim 1, wherein, constructing the defect prediction model of the target software according to the historical parameter values of the influencing factors of the target software quality, comprising: Obtain the historical parameter values of the influencing factors and the number of historical software defects of the software version to which each historical parameter value belongs; determining a model sample data set from the historical parameter values of the influencing factors and the historical software defect quantity; The defect prediction model is constructed according to the model sample data set. 3. The method according to claim 2, wherein the determining the model sample data set from the historical parameter values of the influencing factors and the historical software defect quantity comprises: performing a correlation analysis on the influencing factors to obtain a correlation value between the influencing factors and the number of defects in the software; Determine the influence factor whose correlation value is greater than the preset threshold as the sample influence factor; The historical parameter value of the sample influencing factor and the historical software defect quantity of the software version to which the sample influencing factor belongs are determined as the model sample data set. 4. The method according to any one of claims 1-3, characterized in that, by using a preset optimization algorithm model to obtain the minimum value of the number of software defects output by the defect prediction model, comprising: The defect prediction model is used as the fitness function of the optimization algorithm model; the optimization algorithm model includes the range and moving speed of the decision variable, and the decision variable is the parameter value in the influencing factors of the software quality. Influencing factors of fixed values; Initializing the optimization parameters of the optimization algorithm model; Based on the initialized optimization parameters, the range and moving speed of the decision variables are updated iteratively until the iteration converges, and the minimum value of the software defect quantity is obtained. 5. The method according to claim 4, wherein the optimization parameters comprise an initial global best position and an initial individual best position; Then, based on the initialized optimization parameters, iteratively update the range and moving speed of the decision variable until the iteration converges, and obtain the minimum value of the number of software defects, including: Based on the initial global best position and the initial individual best position, iteratively update the range and moving speed of the decision variable until iteratively converges, and obtain the target global best position; When the target global optimal position is obtained, the output value in the defect prediction model is determined as the minimum value of the software defect quantity. 6. The method according to claim 5, characterized in that, based on the initial global best position and the initial individual best position, iteratively updates the range and moving speed of the decision variable until iteratively converges to obtain the target Global sweet spots, including: Using the initial global best position and the initial individual best position as initial values, update the range and moving speed of the decision variable; Calculate the output value of each updated decision variable in the defect prediction model according to the range and moving speed of the updated decision variable, and compare all the output values to determine the current global best position; If a preset iterative convergence condition is reached, determining that the current global best position is the target global best position; If the iterative convergence condition is not reached, continue to obtain the next global best position until the iterative convergence condition is reached, and the target global best position is obtained. 7. The method according to claim 6, wherein the optimization parameter further comprises an initial iteration index, a maximum number of iterations, an initial stop index, and a maximum stop index; Then, the iteration convergence condition includes: the initial iteration index reaches the maximum number of iterations, or the initial stop index reaches the maximum stop index. 8. A software quality influencing factor determination device, characterized in that the device comprises: The building module is used to perform correlation analysis on the influencing factors of the target software quality, obtain the correlation value between the influencing factors and the number of defects of the software, and obtain the correlation value between the influencing factors and the number of defects of the software, and obtain the correlation value and the influencing factors of the target software quality through the correlation value and the target software quality. Determine the model sample data set based on the historical parameter values of the historical parameter values and the number of historical software defects of the software version to which each historical parameter value belongs, and build the defect prediction model of the target software based on the model sample data set; the defect prediction model is used according to the software The parameter values of the influencing factors of quality predict the corresponding number of software defects; an optimization module, configured to use a preset optimization algorithm model to obtain the minimum value of the number of software defects output by the defect prediction model; a determining module, configured to determine the target parameter value of the influencing factor of the target software quality according to the minimum value of the software defect quantity; in,

; where the parameter a is the influencing factor, the parameter b is the number of software defects of the version to which the influencing factor a belongs,

is the actual value in parameter a,

is the actual value in parameter b,

is the mean value of parameter a,

is the mean value of parameter b,

is the correlation value of parameters a and b. 9. A computer device, comprising a memory and a processor, wherein the memory stores a computer program, wherein the processor implements the method according to any one of claims 1 to 7 when the processor executes the computer program. step. 10. A computer-readable storage medium on which a computer program is stored, characterized in that, when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images