Software release method and device, computer readable storage medium and program product

Abstract

The invention relates to a software release method, a software release device, a non-transitory computer readable storage medium and a computer program product. According to one aspect of the invention, the software publishing method comprises the following steps: inputting business system information associated with software to be published into a first classification model to obtain a first software publishing strategy; inputting the service system information and the first software release strategy into a second classification model to obtain a second software release strategy; and on the basis of the second software release strategy, executing release operation on the to-be-released software, wherein training data of the first classification model and the second classification model comprises the corrected second software release strategy.

Description

Technical Field

[0001] The present application relates to the field of computers, and more particularly to a method for publishing software, an apparatus for publishing software, a non-transitory computer-readable storage medium, and a computer program product. Background Art

[0002] With the development of Internet software technology, the development and release of Internet software has formed a set of standard processes, which generally include continuous integration and continuous deployment. In software development practice, continuous deployment is the basis for building, testing and deploying software to production. The current continuous deployment release platform is generally in the form of stages or tasks, which need to be configured in the pipeline and released in the form of pipeline. The configuration is relatively complex, the release success rate is low, and it is impossible to integrate the various platforms involved in the release process to achieve one-stop release capabilities. In addition, due to the lack of personalized customized release strategies, the user experience is affected. Summary of the invention

[0003] In order to solve or at least alleviate one or more of the above problems, the following technical solutions are provided.

[0004] According to a first aspect of the present application, a method for software release is provided, the method comprising the following steps: inputting business system information associated with software to be released into a first classification model to obtain a first software release strategy; inputting the business system information and the first software release strategy into a second classification model to obtain a second software release strategy; and performing a release operation on the software to be released based on the second software release strategy, the training data of the first classification model and the second classification model including a revised second software release strategy.

[0005] According to the software publishing method described in one embodiment of the present application, the business system information associated with the software to be released includes one or more of the following: system importance, business impact, software release frequency, software release scenario and degree of automation of software release.

[0006] According to the method for software publishing according to one embodiment of the present application or any of the above embodiments, the first classification model includes a random forest model, and the second classification model includes a support vector machine model.

[0007] According to the method for software publishing according to one embodiment of the present application or any of the above embodiments, the training data of the first classification model and the second classification model are also generated based on at least preprocessing business system information and historical publishing information associated with the published software.

[0008] According to the method for software release in one embodiment or any one of the above embodiments of the present application, the business system information associated with the released software includes system importance, business impact, software release frequency, software release scenario and the degree of automation of software release, and the historical release information associated with the released software includes release time, diversion strategy, diversion ratio and release results.

[0009] According to the method for software release in one embodiment or any of the above embodiments of the present application, preprocessing of business system information and historical release information associated with the released software includes one or more of the following preprocessing operations: converting the release time into numerical features; performing one-hot encoding on the business impact; determining a moving average of the software release frequency based on a moving time window; and creating a composite feature based on at least two of the system importance, the business impact, the software release frequency, the software release scenario, and the degree of automation of the software release.

[0010] According to the software publishing method described in one embodiment of the present application or any of the above embodiments, the first software publishing strategy includes one or more of the following: a prediction result of a publishing time, a prediction result of a diversion strategy, and a prediction result of a diversion ratio.

[0011] According to the method for software release in one embodiment of the present application or any of the above embodiments, the prediction result of the diversion strategy includes grayscale release and full release, and the prediction result of the diversion ratio includes the range of the proportion of users using the software to be released.

[0012] According to the software publishing method of one embodiment of the present application or any of the above embodiments, the second software publishing strategy includes one or more of the following: publishing time, diversion strategy and diversion ratio.

[0013] According to the software publishing method described in one embodiment of the present application or any of the above embodiments, the revised second software publishing strategy is obtained in the following manner: using a neural network model to perform a revision operation on the second software publishing strategy at least based on the execution result of the publishing operation performed on the software to be released based on the second software publishing strategy.

[0014] According to the software publishing method described in one embodiment of the present application or any of the above embodiments, the correction operation includes: determining the accuracy of the second software publishing policy according to the execution result of the publishing operation performed on the software to be released based on the second software publishing policy; selecting part or all of the second software publishing policies in the second software publishing policy based on the determined accuracy of the second software publishing policy; and adjusting the weights of part or all of the second software publishing policies as training data according to the moment when the publishing operation is performed on the software to be released based on the part or all of the second software publishing policies, so as to obtain the corrected second software publishing policy.

[0015] According to the software publishing method described in one embodiment of the present application or any of the above embodiments, the correction operation includes: determining the accuracy of the second software publishing policy based on the execution result of the publishing operation on the software to be released based on the second software publishing policy; selecting part or all of the second software publishing policies in the second software publishing policy based on the determined accuracy of the second software publishing policy; and modifying one or more of the publishing time, diversion strategy and diversion ratio in the part or all of the second software publishing policies to obtain the corrected second software publishing policy.

[0016] According to the method for software release according to one embodiment of the present application or any of the above embodiments, the method further includes: using a neural network model to establish a release strategy database based on at least the execution result of the release operation on the software to be released based on the second software release strategy, the release strategy database includes the business system information, the second software release strategy determined based on the business system information, and the release recommendation degree corresponding to the business system information and the second software release strategy.

[0017] According to the software publishing method described in one embodiment of the present application or any of the above embodiments, the execution result of the publishing operation performed on the software to be released based on the second software publishing strategy includes the success rate and user satisfaction of business processing using the software to be released.

[0018] According to the method for software publishing described in one embodiment of the present application or any of the above embodiments, the method further includes: performing anomaly detection on the software to be released in response to the user's actual operation and simulated user operation after performing the publishing operation on the software to be released; and selectively stopping the execution of the publishing operation based on the anomaly detection result.

[0019] According to the method for software publishing described in one embodiment of the present application or any of the above embodiments, performing anomaly detection on the software to be released includes detecting one or more of the following in response to actual user operations and simulated user operations: the success rate of business processing using the software to be released, the time consumed for business processing using the software to be released, the user request volume of the software to be released, and the error code of the software to be released.

[0020] According to a second aspect of the present application, a software publishing device is provided, the device comprising: a memory; a processor coupled to the memory; and a computer program stored on the memory, wherein when the computer program runs on the processor, it causes the steps of the software publishing method described in the first aspect of the present application to be executed.

[0021] According to a third aspect of the present application, a non-transitory computer-readable storage medium is provided, characterized in that the non-transitory computer-readable storage medium includes instructions, and the instructions, when run, execute the steps of the method for software publishing according to the first aspect of the present application.

[0022] According to a fourth aspect of the present application, a computer program product is provided, wherein the computer program product comprises instructions, and when the instructions are executed by a processor, the steps of the method for software publishing according to the first aspect of the present application are implemented.

[0023] The software release scheme according to one or more embodiments of the present application can obtain a software release strategy suitable for the software to be released based on at least the business system information associated with the software to be released through the first classification model and the second classification model, and can train the first classification model and the second classification model by using the revised second software release strategy as training data, thereby realizing a personalized and fully automated software release process, improving the efficiency and success rate of software release, reducing the release cost, and improving the user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The above and / or other aspects and advantages of the present application will become clearer and easier to understand through the following description of various aspects in conjunction with the accompanying drawings, in which the same or similar units are represented by the same reference numerals. In the accompanying drawings:

[0025] Figure 1 A flowchart of a method for publishing software according to one or more embodiments of the present application is shown.

[0026] Figure 2 A flowchart of a method for publishing software according to one or more embodiments of the present application is shown.

[0027] Figure 3A schematic block diagram of a software publishing apparatus according to one or more embodiments of the present application is shown. DETAILED DESCRIPTION

[0028] The following describes in detail example embodiments of the present application, and examples of these embodiments are illustrated in the accompanying drawings. It should be noted that the following description is for explanation and illustration and is therefore not to be construed as limiting the present application. Without departing from the principles of the present application, those skilled in the art may make electrical, mechanical, logical and structural changes to these embodiments as needed without departing from the scope of the present application. In addition, those skilled in the art may appreciate that, for any specific application scenario or actual need, one or more features of the different embodiments described below may be combined.

[0029] Terms such as "comprising" and "including" indicate that in addition to the units and steps directly and clearly stated in the specification, the technical solution of the present application does not exclude the situation of having other units and steps that are not directly or clearly stated. Terms such as "first" and "second" do not indicate the order of units in terms of time, space, size, etc., but are only used to distinguish between units.

[0030] In the context of this application, the term "software" refers to a computer program that runs on an operating system of a terminal and can perform one or more specific functions. The term "software release" refers to the process of launching software to the market and providing it to users after it is developed.

[0031] Hereinafter, various exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings.

[0032] Figure 1 A flowchart of a method for publishing software according to one or more embodiments of the present application is shown.

[0033] like Figure 1 As shown in , in step S101, business system information associated with the software to be released is input into a first classification model to obtain a first software release strategy.

[0034] Optionally, the business system information may include system importance, business impact, software release frequency, software release scenarios, and the degree of automation of software release. For example, system importance may be represented by a score of 0-10, business impact may be represented by low, medium, and high, software release frequency may be used to indicate the number of times software is released each month, and software release scenarios may include the latest version release, minor fixes, emergency fixes, etc. for the software to be released.

[0035] Optionally, the first classification model can be implemented as a random forest model, which is used to process the business system information associated with the software to be released to obtain a first software release strategy including one or more of the prediction results of the release time, the prediction results of the diversion strategy, and the prediction results of the diversion ratio. It should be noted that the prediction results of the release time, the prediction results of the diversion strategy, and the prediction results of the diversion ratio can each be used to represent the range of the prediction. For example, the prediction result of the release time can be implemented as a time range from 10:00 on Monday to 20:00 on Friday, the prediction result of the diversion strategy can include grayscale release and full release, and the prediction result of the diversion ratio can be implemented as a range of the proportion of users using the software to be released (for example, 10%-25%, 50%-80%, etc.). Taking the release of a new version of software as an example, a full release means that all objects can use the new version of the software after it is released. A grayscale release means that while some objects continue to use the old version of the software, other objects begin to use the new version of the software. If the objects have no objection to the new version of the software, the scope of use of the new version of the software is gradually expanded until all objects use the new version of the software, thereby achieving a full update of the software version.

[0036] In one embodiment, the random forest model may include multiple decision trees, each of which may be trained independently. In one embodiment, the prediction results of each trained decision tree may be processed by voting or averaging to obtain the final prediction result. By integrating multiple decision trees, the overfitting problem can be effectively avoided and the generalization ability of the first classification model can be improved.

[0037] In step S103, the business system information and the first software release strategy are input into the second classification model to obtain the second software release strategy.

[0038] Optionally, the second classification model can be implemented as a support vector machine model, which is used to process the business system information and the first software release strategy to obtain a second software release strategy including one or more of a release time, a diversion strategy, and a diversion ratio. In one embodiment, the second software release strategy may include a release time, a diversion strategy, and a diversion ratio. Exemplarily, the second software release strategy can be implemented as a release time of 20:00 this Friday, a diversion strategy of grayscale release, and a diversion ratio of 60%, that is, the second software release strategy can be implemented as a determined release time, a diversion strategy, and a diversion ratio, which are different from the prediction results of the release time, the diversion strategy, and the diversion ratio in the first software release strategy. The first classification model can be used to roughly predict the release time, the diversion strategy, and the diversion ratio to obtain the first software release strategy, and then the second classification model can be used to accurately predict the release time, the diversion strategy, and the diversion ratio, thereby improving the prediction accuracy.

[0039] In one embodiment, the training data of the first classification model and the second classification model can be generated based on at least preprocessing the business system information and historical release information associated with the released software. In one embodiment, the business system information associated with the released software may include system importance, business impact, software release frequency, software release scenario, and the degree of automation of software release, and the historical release information associated with the released software may include release time, diversion strategy, diversion ratio, and release result (e.g., success, partial success, failure). In one embodiment, preprocessing the above business system information and historical release information may include: converting the release time into a numerical feature (e.g., day of the week, hour); performing unique hot encoding on the business impact; determining the moving average of the software release frequency based on a moving time window; and creating a composite feature based on at least two of the system importance, business impact, software release frequency, software release scenario, and the degree of automation of software release, such as creating a composite feature by multiplying the system importance and the software release frequency.

[0040] In step S105, a publishing operation is performed on the software to be published based on the second software publishing strategy, and the training data of the first classification model and the second classification model include the revised second software publishing strategy.

[0041] Exemplarily, when the second software release strategy is implemented as the release time of this Friday at 20:00, the diversion strategy is gray release, and the diversion ratio is 60%, the software to be released can be released to 60% of users for use based on gray release at 20:00 this Friday. Exemplarily, when the second software release strategy is implemented as the release time of this Wednesday at 14:00, the diversion strategy is full release, and the diversion ratio is 100%, the software to be released can be released to all users for use based on full release at 14:00 this Wednesday.

[0042] In one embodiment, the trained neural network model may be used to perform a correction operation on the second software release policy at least according to an execution result of performing a release operation on the software to be released based on the second software release policy.

[0043] In one embodiment, the correction operation may include: determining the accuracy of the second software release policy according to the execution result of the release operation on the software to be released based on the second software release policy; selecting part or all of the second software release policies in the second software release policy based on the determined accuracy of the second software release policy; and adjusting the weight of part or all of the second software release policies as training data according to the time when the release operation is performed on the software to be released based on part or all of the second software release policies, so as to obtain the corrected second software release policy. Exemplarily, the execution result of the release operation on the software to be released may include the success rate of business processing using the software to be released and the satisfaction of user feedback, and part or all of the second software release policies corresponding to the success rate greater than or equal to 95% and the satisfaction of user feedback greater than or equal to 95 points can be selected, and the weight of part or all of the second software release policies as training data is adjusted according to the time when the release operation is performed on the software to be released based on part or all of the second software release policies, so that the weight of the latest second software release policy as training data is the largest, and the weight of the second software release policy as training data gradually decreases over time, thereby improving the training effect of the classification model. For example, the NPSMeter tool can be used to implement surveys and user feedback collection within the software, and methods such as text analysis, topic analysis, association analysis, and data visualization can be used to determine the satisfaction of user feedback.

[0044] In one embodiment, the correction operation may include: determining the accuracy of the second software release policy according to the execution result of the release operation on the software to be released based on the second software release policy; selecting part or all of the second software release policies in the second software release policy based on the determined accuracy of the second software release policy; and modifying one or more of the release time, diversion strategy and diversion ratio in part or all of the second software release policies to obtain the corrected second software release policy. Exemplarily, the execution result of the release operation on the software to be released may include the success rate of business processing using the software to be released and the satisfaction of user feedback, and part or all of the second software release policies corresponding to the success rate less than 95% or the satisfaction of user feedback less than 95 points can be selected, and one or more of the release time, diversion strategy and diversion ratio in part or all of the second software release policies can be modified to obtain the corrected second software release policy. By correcting the second software release policy and using it as training data, the amount of training data can be increased and the accuracy of training data can be improved, thereby improving the training effect of the classification model.

[0045] In one embodiment, a neural network model can be used to establish a release strategy database based on at least the execution result of the release operation on the software to be released based on the second software release strategy, wherein the release strategy database includes business system information, the second software release strategy determined by the business system information, and the release recommendation degree corresponding to the business system information and the second software release strategy, that is, the association relationship between the business system information, the second software release strategy and the release recommendation degree is established through the release strategy database. In one embodiment, the second software release strategy can be determined based on the business system information associated with the software to be released through the release strategy database. In one embodiment, the release strategy database can be used as training data, which can increase the amount of training data and improve the accuracy of the training data, thereby improving the training effect of the classification model.

[0046] In one embodiment, after executing the publishing operation on the software to be published, an abnormality detection can be performed on the software to be published in response to the actual user operation and the simulated user operation, and the publishing operation can be selectively stopped based on the abnormality detection result. Exemplarily, the abnormality detection of the software to be published can include detecting the success rate of business processing using the software to be published, the time consumed for business processing using the software to be published, the user request volume of the software to be published, the error code of the software to be published, etc.

[0047] The software publishing method according to one or more embodiments of the present application can obtain a software publishing strategy suitable for the software to be released based on at least the business system information associated with the software to be released through the first classification model and the second classification model, and can train the first classification model and the second classification model by using the revised second software publishing strategy as training data, thereby realizing a personalized and fully automated software publishing process, improving the efficiency and success rate of software publishing, reducing the publishing cost, and improving the user experience.

[0048] The following will be combined Figure 2-3 The specific implementation of software release according to one or more embodiments of the present application is further described.

[0049] Figure 2 A flowchart of a method for publishing software according to one or more embodiments of the present application is shown.

[0050] like Figure 2 As shown in , in step S201, business system information associated with the software to be released is received. Optionally, business system information including system importance, business impact, software release frequency, software release scenario and software release automation level can be received in response to user input.

[0051] In step S203, the software release strategy is determined based on at least the business system information associated with the software to be released using the classification model. The classification model includes a first classification model and a second classification model, wherein the first classification model is used to determine the prediction result of the software release strategy based on the business system information associated with the software to be released, and the second classification model is used to determine the software release strategy based on the prediction result of the business system information and the software release strategy. Optionally, the software release strategy may include a release time, a diversion strategy, and a diversion ratio.

[0052] In step S205, a release operation is performed on the software to be released using the software release policy.

[0053] In step S207, after the release operation is performed on the software to be released, the software to be released is detected for abnormality in response to the actual user operation and the simulated user operation to determine whether the software to be released is abnormal. If the software to be released is determined to be abnormal, the process proceeds to step S209, otherwise the process ends after the release operation is completed.

[0054] In step S209, the publishing operation is stopped.

[0055] In step S211, the software release policy is revised, and the revised software release policy is fed back to step S203 as training data for the classification model.

[0056] Optionally, in step S211, the trained neural network model may be used to perform a correction operation on the software release policy at least according to an execution result of performing a release operation on the software to be released based on the software release policy.

[0057] Figure 3 A schematic block diagram of a software publishing apparatus according to one or more embodiments of the present application is shown.

[0058] like Figure 3 As shown in FIG, the software publishing apparatus 300 includes a memory 310, a processor 320, and a computer program 330 stored in the memory 310 and executable on the processor 320. The processor 320 executes the computer program 330 to implement a software publishing method according to one aspect of the present application.

[0059] Optionally, the software publishing device 300 may further include a display ( Figure 3The display may be used to receive business system information associated with the software to be released, which is input by the user. For example, the display may display the software release status (e.g., released, unreleased) and statistical information (e.g., release information of each software to be released, such as release time, number of released, number of unreleased, etc.). For example, the display may display each step of the release process (e.g., inspection, notification, alarm, component update operation, database operation, etc.), so that the user can edit and adjust the specific operation of each step as needed.

[0060] In addition, the present application may also be implemented as a non-transitory computer-readable storage medium, in which a program for causing a computer to execute the software publishing method according to one aspect of the present application is stored.

[0061] Here, as non-temporary computer-readable storage media, various forms of non-temporary computer-readable storage media can be used, such as disks (e.g., magnetic disks, optical disks, etc.), cards (e.g., memory cards, optical cards, etc.), semiconductor memories (e.g., ROMs, non-volatile memories, etc.), and tapes (e.g., magnetic tapes, cassettes, etc.).

[0062] In the applicable situation, hardware, software or a combination of hardware and software can be used to realize the various embodiments provided by the application. Moreover, in the applicable situation, without departing from the scope of the application, the various hardware components and / or software components set forth herein can be combined into a composite component comprising software, hardware and / or both. In the applicable situation, without departing from the scope of the application, the various hardware components and / or software components set forth herein can be divided into subcomponents comprising software, hardware or both. In addition, in the applicable situation, it is contemplated that the software component can be implemented as a hardware component, and vice versa.

[0063] Software according to the present application (such as program code and / or data) may be stored on one or more non-transitory computer-readable storage media. It is also contemplated that the software identified herein may be implemented using one or more general or special-purpose computers and / or computer systems, networked and / or otherwise. Where applicable, the order of the various steps described herein may be changed, combined into composite steps, and / or divided into sub-steps to provide the features described herein.

[0064] The embodiments and examples set forth herein are provided to best illustrate embodiments according to the present application and its specific applications, and thereby enable those skilled in the art to implement and use the present application. However, those skilled in the art will appreciate that the above description and examples are provided only for ease of illustration and example. The description set forth is not intended to cover all aspects of the present application or to limit the present application to the precise form disclosed.

Claims

1. A method for software publishing, characterized in that: The method comprises the following steps: Inputting the business system information associated with the software to be released into the first classification model to obtain a first software release strategy; Inputting the business system information and the first software release strategy into a second classification model to obtain a second software release strategy; as well as A publishing operation is performed on the to-be-published software based on the second software publishing strategy, and the training data of the first classification model and the second classification model include a revised second software publishing strategy.

2. The method according to claim 1, wherein the business system information associated with the software to be released includes one or more of the following: system importance, business impact, software release frequency, software release scenario and degree of automation of software release.

3. The method according to claim 1, wherein the first classification model comprises a random forest model and the second classification model comprises a support vector machine model.

4. The method according to claim 1, wherein the training data of the first classification model and the second classification model are also generated based on at least preprocessing business system information and historical release information associated with the released software.

5. The method according to claim 4, wherein the business system information associated with the released software includes system importance, business impact, software release frequency, software release scenario and the degree of automation of software release, and the historical release information associated with the released software includes release time, diversion strategy, diversion ratio and release results.

6. The method according to claim 5, wherein preprocessing the business system information and historical release information associated with the released software comprises one or more of the following preprocessing operations: converting the release time into a numerical feature; One-hot encoding the business impact; Determine a moving average of the software release frequency based on a moving time window; as well as A composite feature is created based on at least two of the system importance, the business impact, the software release frequency, the software release scenario, and the automation level of the software release.

7. The method according to claim 1, wherein the first software release strategy comprises one or more of the following: a prediction result of a release time, a prediction result of a diversion strategy, and a prediction result of a diversion ratio.

8. The method according to claim 7, wherein the prediction result of the diversion strategy includes grayscale release and full release, and the prediction result of the diversion ratio includes the range of the proportion of users using the software to be released.

9. The method according to claim 1, wherein the second software release strategy comprises one or more of the following: release time, diversion strategy and diversion ratio.

10. The method according to claim 1, wherein the modified second software release policy is obtained by: The neural network model is used to perform a correction operation on the second software release strategy at least according to an execution result of performing a release operation on the to-be-released software based on the second software release strategy.

11. The method of claim 10, wherein the corrective operation comprises: determining the accuracy of the second software release strategy according to an execution result of performing a release operation on the to-be-released software based on the second software release strategy; selecting part or all of the second software release policies in the second software release policy based on the determined accuracy of the second software release policy; as well as According to the time when the release operation is performed on the software to be released based on the part or all of the second software release policy, the weight of the part or all of the second software release policy as training data is adjusted to obtain the modified second software release policy.

12. The method of claim 10, wherein the corrective operation comprises: determining the accuracy of the second software release strategy according to an execution result of performing a release operation on the to-be-released software based on the second software release strategy; selecting part or all of the second software release policies in the second software release policy based on the determined accuracy of the second software release policy; as well as Modify one or more of the release time, the diversion strategy, and the diversion ratio in part or all of the second software release strategy to obtain the modified second software release strategy.

13. The method according to claim 10, wherein the method further comprises: A release strategy database is established using a neural network model at least according to the execution result of the release operation on the software to be released based on the second software release strategy. The release strategy database includes the business system information, the second software release strategy determined based on the business system information, and the release recommendation degree corresponding to the business system information and the second software release strategy.

14. The method according to any one of claims 10 to 13, wherein the result of executing the release operation on the software to be released based on the second software release strategy comprises a success rate of business processing using the software to be released and user satisfaction.

15. The method according to claim 1, wherein the method further comprises: After performing a release operation on the software to be released, performing anomaly detection on the software to be released in response to an actual user operation and a simulated user operation; as well as The publishing operation is selectively stopped based on the anomaly detection result.

16. The method according to claim 15, wherein performing anomaly detection on the software to be released includes detecting one or more of the following in response to actual user operations and simulated user operations: the success rate of business processing using the software to be released, the time consumed for business processing using the software to be released, the user request volume of the software to be released, and the error code of the software to be released.

17. A software publishing device, characterized in that: The device is arranged on the merchant terminal device to verify the legitimacy of the user terminal device, and the device comprises: Memory; a processor coupled to the memory; and A computer program stored on the memory, which, when executed on the processor, causes the following operations: Inputting the business system information associated with the software to be released into the first classification model to obtain a first software release strategy; Inputting the business system information and the first software release strategy into a second classification model to obtain a second software release strategy; and A publishing operation is performed on the software to be released based on the second software release strategy, and the training data of the first classification model and the second classification model include a revised second software release strategy.

18. The apparatus according to claim 17, wherein the business system information associated with the software to be released includes one or more of the following: system importance, business impact, software release frequency, software release scenario, and degree of automation of software release.

19. The apparatus of claim 17, wherein the first classification model comprises a random forest model, and the second classification model comprises a support vector machine model.

20. The apparatus according to claim 17, wherein the training data of the first classification model and the second classification model are also generated based on at least preprocessing business system information and historical release information associated with the released software.

21. The device according to claim 20, wherein the business system information associated with the released software includes system importance, business impact, software release frequency, software release scenario and degree of automation of software release, and the historical release information associated with the released software includes release time, diversion strategy, diversion ratio and release results.

22. The apparatus according to claim 21, wherein when the computer program is executed on the processor, the preprocessing of the business system information and the historical release information associated with the released software comprises one or more of the following preprocessing operations: converting the release time into a numerical feature; One-hot encoding the business impact; Determine a moving average of the software release frequency based on a moving time window; as well as A composite feature is created based on at least two of the system importance, the business impact, the software release frequency, the software release scenario, and the automation level of the software release.

23. The apparatus according to claim 17, wherein the first software release strategy comprises one or more of the following: a prediction result of a release time, a prediction result of a diversion strategy, and a prediction result of a diversion ratio.

24. The device according to claim 23, wherein the prediction result of the diversion strategy includes grayscale release and full release, and the prediction result of the diversion ratio includes the range of the proportion of users using the software to be released.

25. The apparatus according to claim 17, wherein the second software release strategy comprises one or more of the following: release time, diversion strategy, and diversion ratio.

26. The apparatus of claim 17, wherein the computer program, when executed on the processor, causes the revised second software release policy to be obtained by: The neural network model is used to perform a correction operation on the second software release strategy at least according to an execution result of performing a release operation on the to-be-released software based on the second software release strategy.

27. The apparatus of claim 26, wherein the computer program when executed on the processor causes the corrective operation to include: determining the accuracy of the second software release strategy according to an execution result of performing a release operation on the to-be-released software based on the second software release strategy; selecting part or all of the second software release policies in the second software release policy based on the determined accuracy of the second software release policy; as well as According to the time when the release operation is performed on the software to be released based on the part or all of the second software release policy, the weight of the part or all of the second software release policy as training data is adjusted to obtain the modified second software release policy.

28. The apparatus of claim 26, wherein the computer program, when executed on the processor, causes the corrective operation to include: determining the accuracy of the second software release strategy according to an execution result of performing a release operation on the to-be-released software based on the second software release strategy; selecting part or all of the second software release policies in the second software release policy based on the determined accuracy of the second software release policy; as well as Modify one or more of the release time, the diversion strategy, and the diversion ratio in part or all of the second software release strategy to obtain the modified second software release strategy.

29. The apparatus of claim 26, wherein the computer program when executed on the processor further causes: A release strategy database is established using a neural network model at least according to the execution result of the release operation on the software to be released based on the second software release strategy. The release strategy database includes the business system information, the second software release strategy determined based on the business system information, and the release recommendation degree corresponding to the business system information and the second software release strategy.

30. The apparatus according to any one of claims 26 to 29, wherein the result of executing the release operation on the software to be released based on the second software release strategy comprises a success rate of business processing using the software to be released and user satisfaction.

31. The apparatus of claim 17, wherein the computer program when executed on the processor further causes: After performing a release operation on the software to be released, performing anomaly detection on the software to be released in response to an actual user operation and a simulated user operation; and The publishing operation is selectively stopped based on the anomaly detection result.

32. An apparatus according to claim 31, wherein when the computer program runs on the processor, it causes anomaly detection to be performed on the software to be released, including detecting one or more of the following in response to actual user operations and simulated user operations: the success rate of business processing using the software to be released, the time consumed for business processing using the software to be released, the user request volume for the software to be released, and the error code of the software to be released.

33. A non-transitory computer-readable storage medium, characterized in that The non-transitory computer-readable storage medium includes instructions, which, when executed, execute the method for software publishing according to any one of claims 1-16.

34. A computer program product, characterized in that The computer program product comprises instructions, and when the instructions are executed by a processor, the method for software publishing according to any one of claims 1 to 16 is implemented.

Images