Method, system and equipment for detecting configuration information of AB experiment and medium

Abstract

The invention discloses a configuration information detection method, system and device for an AB experiment and a medium, and belongs to the technical field of computers. The AB experiment configuration information detection method comprises the steps of obtaining current configuration information of an AB experiment in a current version; generating a current abstract syntax tree corresponding to the current configuration information, and comparing the current abstract syntax tree with a historical abstract syntax tree to obtain configuration change content of the AB experiment; querying a preset rule corresponding to the configuration change content; judging whether the configuration change content conforms to the preset rule or not; if yes, carrying out change influence analysis according to the configuration change content to obtain an analysis result; wherein the analysis result is used for describing an influence object and an influence degree of version change; and if not, judging that the current configuration information is abnormal. The AB experiment configuration can be effectively managed, and the safety of the AB experiment is improved.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a configuration information detection method, system, device and medium for AB experiment. Background Art

[0002] With the development of internet technology, AB experiments are widely used in software testing to evaluate the effectiveness of new features or versions. However, AB experiment configuration changes can directly impact online business operations. Misoperation can easily cause online incidents and result in significant losses.

[0003] In related technologies, changes and verifications of AB experiment configurations are usually done manually, which can easily lead to chaotic AB experiment configuration management and cause online accidents.

[0004] Therefore, how to effectively manage AB experiment configuration and improve the safety of AB experiments is a technical problem that those skilled in the art currently need to solve. Summary of the Invention

[0005] The purpose of this application is to provide a configuration information detection method, system, device and medium for AB experiments, which can effectively manage AB experiment configuration and improve the security of AB experiments.

[0006] To solve the above technical problems, the present application provides a configuration information detection method for an AB experiment, comprising:

[0007] Get the current configuration information of AB experiment in the current version;

[0008] Generating a current abstract syntax tree corresponding to the current configuration information, and comparing the current abstract syntax tree with a historical abstract syntax tree to obtain configuration changes of the AB experiment; wherein the historical abstract syntax tree is generated based on historical configuration information, and the historical configuration information is configuration information of the AB experiment before the current version;

[0009] Query the preset rules corresponding to the configuration change content;

[0010] Determining whether the configuration change content complies with the preset rules;

[0011] If so, perform a change impact analysis based on the configuration change content to obtain an analysis result; wherein the analysis result is used to describe the affected objects and impact extent of the version change;

[0012] If not, it is determined that there is an abnormality in the current configuration information.

[0013] Optionally, also include:

[0014] If it is detected that the configuration information of the AB experiment has changed, the current version number of the current configuration information is generated according to the current time and the hash value;

[0015] If an experiment execution instruction is received, the user cluster participating in the AB experiment is determined, and the target user in the user cluster is removed; wherein the target user is a user bound to other version numbers;

[0016] Binding the current version number to users in the user cluster and executing the AB experiment according to the current configuration information;

[0017] After the AB experiment is completed, the users in the user cluster are unbound from the current version number.

[0018] Optionally, after obtaining the current configuration information of the AB experiment in the current version, the following is also included:

[0019] Decomposing the current configuration information into a plurality of dimension fields, and generating a structured snapshot of the current configuration information according to all the dimension fields;

[0020] If a version comparison instruction is received, the structured snapshot of the current configuration information is compared with the structured snapshot of the historical configuration information to generate a visual difference report.

[0021] Optionally, after determining that the current configuration information is abnormal, the method further includes:

[0022] The structured snapshot of the historical configuration information is used to perform version rollback.

[0023] Optionally, after obtaining the current configuration information of the AB experiment in the current version, the following is also included:

[0024] Parse the business requirement document to obtain the value constraints corresponding to multiple business requirement items;

[0025] Determining whether the current configuration information meets the value constraint condition;

[0026] If not, risk warning information is generated, and the fields in the current configuration information that do not meet the value constraint conditions are marked.

[0027] Optionally, after determining whether the current configuration information meets the value constraint condition, the method further includes:

[0028] If the current configuration information does not meet the value constraint condition, using a machine learning model to generate an AB experiment prediction effect corresponding to the current configuration information;

[0029] The current configuration information is modified based on the AB experiment prediction result and the historical configuration information.

[0030] Optionally, after querying the preset rules corresponding to the configuration change content, the method further includes:

[0031] If the number of the preset rules is greater than 1, determining whether there are conflicting preset rules;

[0032] If so, the conflicting preset rules will be uploaded to the management terminal and a manual review prompt will be generated;

[0033] Accordingly, determining whether the configuration change content complies with the preset rules includes:

[0034] Determining the priority of the preset rules;

[0035] Determine whether the configuration change content complies with the preset rules in order of priority from high to low.

[0036] The present application also provides a configuration information detection system for an AB experiment, the system comprising:

[0037] Information acquisition module, used to obtain the current configuration information of AB experiment in the current version;

[0038] a change determination module, configured to generate a current abstract syntax tree corresponding to the current configuration information, and compare the current abstract syntax tree with a historical abstract syntax tree to obtain configuration changes for the AB experiment; wherein the historical abstract syntax tree is generated based on historical configuration information, and the historical configuration information is configuration information of the AB experiment before the current version;

[0039] A rule query module is used to query the preset rules corresponding to the configuration change content;

[0040] A judgment module is used to determine whether the configuration change content complies with the preset rules; if so, a change impact analysis is performed based on the configuration change content to obtain an analysis result; wherein the analysis result is used to describe the affected objects and impact degree of the version change; if not, it is determined that there is an abnormality in the current configuration information.

[0041] The present application also provides a storage medium storing a computer program, which, when executed, implements the steps of the configuration information detection method of the above-mentioned AB experiment.

[0042] The present application also provides an electronic device including a memory and a processor, wherein the memory stores a computer program, and when the processor calls the computer program in the memory, the steps of the configuration information detection method of the above-mentioned AB experiment are implemented.

[0043] The present application provides a configuration information detection method for an AB experiment. After obtaining the current configuration information of the AB experiment, the method compares the current configuration information with the abstract syntax tree of the historical configuration information to obtain the configuration change content of the AB experiment due to the version update. The present application compares the configuration change content with the corresponding preset rules, and performs a change impact analysis on the configuration change content when the configuration change content complies with the preset rules to obtain the analysis result; if the configuration change content does not comply with the preset rules, it is determined that the current configuration information is abnormal. Through the above scheme, the present application can accurately detect the configuration change content of the AB experiment, and analyze the compliance of the configuration change content and the impact of the change, so that the staff can make preparations in advance and effectively avoid risks. Therefore, the present application can effectively manage the AB experiment configuration and improve the security of the AB experiment. The present application also provides a configuration information detection system, a storage medium and an electronic device for an AB experiment, which have the above-mentioned beneficial effects and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0045] Figure 1 A flowchart of a configuration information detection method for an AB experiment provided in an embodiment of the present application;

[0046] Figure 2 A flowchart of configuration changes of an experimental version provided in an embodiment of the present application;

[0047] Figure 3 This is a structural diagram of a configuration information detection system for an AB experiment provided in an embodiment of the present application. DETAILED DESCRIPTION

[0048] To make the purpose, technical solutions, and advantages of the embodiments of this application more clear, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0049] See below Figure 1 , Figure 1 This is a flow chart of a configuration information detection method for an AB experiment provided in an embodiment of the present application.

[0050] Specific steps may include:

[0051] S101: Obtain the current configuration information of the AB experiment in the current version;

[0052] Among them, this embodiment can be applied to electronic devices with data analysis functions. Before this step, there may be an operation of receiving AB experiment version update information, and the current configuration information of the AB experiment in the current version can be obtained. The AB experiment is an experiment that evaluates new functions or designs by randomly assigning users to different versions (such as A and B) and comparing key indicators.

[0053] AB experiments can have multiple versions. This step retrieves the configuration information for the current version (i.e., the latest version) of the AB experiment. This configuration information can include the experiment audience, traffic configuration, group configuration, and group parameter configuration changes.

[0054] S102: Generate a current abstract syntax tree corresponding to the current configuration information, compare the current abstract syntax tree with a historical abstract syntax tree, and obtain configuration changes of the AB experiment;

[0055] After obtaining the current configuration information, this step can generate a current Abstract Syntax Tree (AST) corresponding to the current configuration information. Prior to this step, historical configuration information can also be obtained and used to generate a historical AST based on the historical configuration information. This historical configuration information refers to the configuration information of the AB experiment prior to the current version (i.e., the historical version). This historical configuration information can be the configuration information of the previous version of the AB experiment.

[0056] This application compares the current abstract syntax tree with the historical abstract syntax tree to obtain the configuration change content of the AB experiment. Specifically, this embodiment can start from the root node and compare whether the structure of the current abstract syntax tree and the historical abstract syntax tree are consistent, such as the type, order and hierarchical relationship of the nodes; this embodiment can also compare the attributes and values ​​of each corresponding node, such as whether the values ​​of fields such as the experiment ID, effective time, audience rules, and traffic distribution strategy have changed. It can be seen that this step can compare the abstract syntax tree based on structure and content, so as to accurately identify the newly added, deleted or modified parts in the current configuration information, thereby obtaining the complete configuration change content.

[0057] S103: Querying the preset rules corresponding to the configuration change content;

[0058] In this embodiment, the correspondence between the change items and the preset rules is pre-stored. After the configuration change content is determined, the preset rules corresponding to the change items involved in the configuration change content can be queried. The above preset rules are used to specify the scope and / or format requirements of the configuration change.

[0059] S104: Determine whether the configuration change content complies with the preset rules; if so, proceed to S105; if not, proceed to S106.

[0060] S105: Perform change impact analysis based on the configuration change content to obtain analysis results;

[0061] This step is based on determining that the configuration change content complies with preset rules. At this time, a change impact analysis can be performed based on the configuration change content to obtain analysis results. The above analysis results are used to describe the impact objects and impact degree of the version change (from historical version to current version).

[0062] Specifically, this embodiment can pre-store the impact of each change item on the AB experiment results, thereby determining the impact objects and impact degree of the configuration version change. In addition, this embodiment can also input the configuration change content into a machine learning model and use the machine learning model to predict the impact objects and impact degree of the version change.

[0063] S106: Determine whether the current configuration information is abnormal.

[0064] After obtaining the current configuration information of the AB experiment, this embodiment compares the current configuration information with the abstract syntax tree of the historical configuration information to obtain the configuration change content of the AB experiment due to the version update. This embodiment compares the configuration change content with the corresponding preset rules, and performs a change impact analysis on the configuration change content when the configuration change content complies with the preset rules to obtain the analysis results; if the configuration change content does not comply with the preset rules, it is determined that the current configuration information is abnormal. Through the above scheme, this embodiment can accurately detect the configuration change content of the AB experiment, and analyze the compliance of the configuration change content and the impact of the change, so that the staff can make preparations in advance and effectively avoid risks. Therefore, this embodiment can effectively manage the AB experiment configuration and improve the security of the AB experiment.

[0065] As for Figure 1 Further introduction to the corresponding embodiment, this embodiment can also achieve mutual exclusion of AB experimental versions in the following ways:

[0066] If a change in the configuration information of the AB experiment is detected, the current version number of the current configuration information is generated based on the current time and hash value; if an experiment execution instruction is received, the user cluster participating in the AB experiment is determined, and the target user in the user cluster is eliminated; wherein the target user is a user bound to another version number; the current version number is bound to the users in the user cluster, and the AB experiment is executed according to the current configuration information; after the AB experiment is executed, the users in the user cluster are unbound from the current version number. The above hash value can be generated based on the current configuration information. Through the above scheme, it can be ensured that the user can only enter one AB experiment version to avoid conflicts.

[0067] As for Figure 1 Further introduction to the corresponding embodiment, after obtaining the current configuration information of the AB experiment in the current version, the current configuration information can also be decomposed into multiple dimensional fields, and a structured snapshot of the current configuration information is generated based on all the dimensional fields; if a version comparison instruction is received, the structured snapshot of the current configuration information is compared with the structured snapshot of the historical configuration information to generate a visual difference report.

[0068] Through the above method, the current configuration information can be decomposed into multiple dimensional fields and stored independently, thereby realizing multi-dimensional field-level comparison.

[0069] Furthermore, after determining that the current configuration information is abnormal, the structured snapshot of the historical configuration information may be used to perform a version rollback, so as to perform an AB experiment using the historical configuration information.

[0070] As for Figure 1 Further introduction to the corresponding embodiment, after obtaining the current configuration information of the AB experiment in the current version, the business requirement document can also be parsed to obtain the value constraints corresponding to multiple business requirement items; determine whether the current configuration information meets the value constraints; if not, generate risk warning information, and mark the fields in the current configuration information that do not meet the value constraints, so as to quickly locate and repair the problem. The above business requirement document is used to describe the business requirements corresponding to the AB experiment, such as the traffic threshold is less than 50 Each business requirement item has a corresponding field in the current configuration information, such as user type. The above method can be used to determine whether the value of the above field meets the value constraint. The value constraint in this embodiment is determined based on the business requirement document, and the value constraint will change as the business changes. The preset rules in this embodiment can be pre-defined configuration information setting rules. Regardless of the business to which the AB experiment is applied, the above preset rules must be met.

[0071] Furthermore, after determining whether the current configuration information meets the value constraint conditions, if the current configuration information does not meet the value constraint conditions, a machine learning model can be used to generate an AB experiment prediction effect corresponding to the current configuration information; and the current configuration information can be corrected in combination with the AB experiment prediction effect and the historical configuration information.

[0072] Specifically, this embodiment can input the current configuration information into a machine learning model to predict the results of AB experiments under the current configuration information, such as conversion rate, retention rate, and other indicators. Historical configuration information is configuration information that meets the value constraints. This embodiment can compare the predicted results of AB experiments corresponding to the current configuration information with the actual results of AB experiments corresponding to the historical configuration information, and adjust the current configuration information based on the comparison results.

[0073] As for Figure 1 Further introduction of the corresponding embodiment, after querying the preset rules corresponding to the configuration change content, if the number of preset rules obtained by the query is greater than 1, it is determined whether there are conflicting preset rules (such as two preset rules requiring the total flow 80 and 90 ); If so, the conflicting preset rules will be uploaded to the management end and a manual review prompt will be generated.

[0074] Correspondingly, if the number of preset rules obtained by the query is greater than 1, the priority of the preset rules can be determined; and whether the configuration change content complies with the preset rules is judged in descending order of priority.

[0075] The process described in the above embodiment is explained below through an embodiment in actual application.

[0076] Currently, AB experiment configuration changes are typically performed manually. However, each AB experiment may correspond to multiple experimental versions. Experimental group configuration and policy changes generate new version numbers, and each version is mutually exclusive in time. Only one experimental version exists on a given timeline, making manual management prone to errors. Configuration changes are difficult to retrieve and highlight, making it difficult for operators to detect changes in a timely manner, increasing the risk of operational errors. Related technologies lack version snapshot functionality, making it impossible to fully compare differences between versions, such as the experimental target audience, group configuration, and traffic allocation, leading to chaotic configuration management.

[0077] The following flaws can be seen in the related technologies: During configuration file version management, experimental configurations are stored in Git (an open-source distributed version control system), and manual diff comparisons are performed, resulting in low accuracy and a lack of systematic verification of experimental objectives and traffic rules. Simple diff detection tools use text comparison to mark code changes, but this does not cover logical layer changes, and text comparison cannot identify semantic-level changes (such as logical expression refactoring). During the grayscale release process, manual approval is relied upon to ensure version mutual exclusion, which is inefficient, and version conflict detection is delayed, which can easily cause online failures.

[0078] In response to the defects of the above-mentioned related technologies, this embodiment provides a security control method for AB experiment configuration changes. This method can automatically generate new version numbers, mutually manage versions, retrieve and highlight changed items, and generate version snapshots to compare the changed content, thereby avoiding online accidents caused by misoperation and improving the security of configuration changes.

[0079] The system architecture on which the above security control method relies includes: version generation module, change detection and highlighting module, version snapshot storage module and security verification module.

[0080] The function of the version generation module is to automatically trigger the generation of a new version number each time the experimental configuration changes (including changes to the experimental audience, traffic configuration, group configuration, and group parameter configuration). The above version number can be combined into a timestamp + hash value to ensure uniqueness and mutual exclusivity. Specifically, this embodiment can manage the version of the AB experiment through the following dynamic version management mechanism: automatically generate a unique version number based on the experimental configuration change, and ensure that only one version is online at the same time through a distributed lock; the version number format is: {year-month-day-hour-minute-second}-{GitHash}-seq{increasing sequence number}. Furthermore, this embodiment can implement a mutually exclusive execution mechanism for experimental versions based on distributed locks. GitHash represents the hash value in Git, and seq represents the sequence.

[0081] The function of the change detection and highlighting module is to highlight key changes (such as the flow threshold from 10 Increased to 20 Specifically, this embodiment can perform semantic-level difference analysis by comparing the abstract syntax tree nodes of the new and old versions of the configuration information, identifying logical changes (such as function call path adjustments), ignoring irrelevant code changes, and using colors to mark different types of changes (red - major adjustments, yellow - optimization suggestions).

[0082] The version snapshot storage module stores complete configuration information (experiment audience, traffic allocation, grouping parameters) and a history of changes. It also supports rapid snapshot retrieval by version number and multi-dimensional comparison, presenting audience tag differences in a table format. The full snapshot comparison engine stores snapshot fields including experiment ID, effective date, audience rules, traffic allocation strategy, and grayscale ratio. The engine supports cross-version comparison and generates visual difference reports (such as diff tables and heat maps).

[0083] Specifically, the multi-dimensional snapshot comparison engine can implement snapshot storage through the snapshot storage structure design. In this embodiment, the experimental configuration can be decomposed into multiple dimensional fields and stored independently to achieve structured snapshot storage. In the multi-dimensional comparison algorithm used by the multi-dimensional snapshot comparison engine, field-level comparison can be performed to generate difference reports by dimension (audience, traffic, grayscale); visualization enhancement processing can also be performed to display the change in traffic ratio through heat maps (such as from A:30 to A:40). →A:40 , B:70 →B:60 ), or highlight the addition or removal of audience tags through an interactive table (e.g., new tags: "new users").

[0084] Based on the structured snapshot, this embodiment can also perform change impact analysis based on the difference fields to automatically assess the impact range; for example, the traffic distribution change affects 80 Users are automatically marked as high-risk. This structured snapshot storage solution breaks the black-box nature of text snapshots and supports precise field-by-field comparison. This embodiment also integrates difference analysis results with the business implications of the experimental configuration (such as traffic overflow risk) for business semantic association, rather than simply displaying text changes. This embodiment structures snapshots by business dimension and designs a visual comparison tool for A / B experimental scenarios.

[0085] The function of the security verification module is to verify in real time whether the changed content complies with the preset rules (such as the total traffic does not exceed 100 , audience labels are mutually exclusive and do not conflict). If a high-risk operation is detected (such as full traffic splitting), the security verification module can push an early warning to manual confirmation, and intercept if a rejection instruction is received. The above-mentioned security verification module can implement the following multi-level security protection: (1) Pre-check rule engine, built-in verification rule library, so as to implement experimental traffic threshold limit and label deduplication; (2) record the operator, operation time, change content and approval status through the change traceability chain; (3) provide an emergency rollback interface, so as to restore to the specified historical version snapshot with one click. This embodiment can realize a dynamic security verification rule library through the verification rule extension interface. This embodiment supports setting priorities for rules and executing them according to the priority in case of conflict. This embodiment can automatically detect rule contradictions and trigger manual review.

[0086] See Figure 2 , Figure 2 The configuration change flow chart of an experimental version provided in the embodiment of the present application is as follows:

[0087] After the experimental repository snapshot is generated, it is determined whether it represents a new version deployment. If so, a full-stack snapshot is collected (i.e., to detect incremental changes). If not, a baseline snapshot is generated (i.e., to establish an initial baseline). The snapshots are compared across multiple dimensions, and a change report is generated. The change report is manually reviewed. If the manual review passes, the version status is updated and a notification is pushed. If the manual review fails, the rejection reason is recorded and the change log is archived. The repository is marked as released, the knowledge base is updated, and version status monitoring is performed.

[0088] The above solution can be applied to the large-scale iteration of AB experiments of Internet products (such as e-commerce recommendation strategies and advertising), the iteration of risk control rules in financial systems, and the grayscale release of firmware for IoT devices.

[0089] This embodiment can automatically generate a new experimental version number and manage each experimental version in a time-exclusive manner, ensuring that there is only one experimental version on the same timeline, avoiding the problem of manual management prone to errors, and achieving standardization of version management. When the experimental configuration changes, this embodiment can automatically retrieve the changed items and highlight the changed content, so that operators can discover the changes in time, reducing the risk of misoperation. This embodiment can automatically generate version snapshots and comprehensively compare the changes of each version relative to the previous version, including the target audience of the experiment, group configuration, traffic division, etc., to achieve transparency in configuration management and avoid the problem of chaotic configuration management. Through the above measures, online accidents caused by misoperation are effectively avoided, thereby improving the security of AB experiment configuration changes.

[0090] See Figure 3 , Figure 3This is a schematic diagram of the structure of a configuration information detection system for an AB experiment provided in an embodiment of the present application. The system may include:

[0091] The information acquisition module 301 is used to obtain the current configuration information of the AB experiment in the current version;

[0092] A change determination module 302 is configured to generate a current abstract syntax tree corresponding to the current configuration information, compare the current abstract syntax tree with a historical abstract syntax tree, and obtain configuration changes for the AB experiment; wherein the historical abstract syntax tree is generated based on historical configuration information, which is configuration information of the AB experiment before the current version;

[0093] A rule query module 303 is used to query the preset rules corresponding to the configuration change content;

[0094] The judgment module 304 is used to determine whether the configuration change content complies with the preset rules; if so, a change impact analysis is performed based on the configuration change content to obtain an analysis result; wherein the analysis result is used to describe the affected objects and impact degree of the version change; if not, it is determined that there is an abnormality in the current configuration information.

[0095] After obtaining the current configuration information of the AB experiment, this embodiment compares the current configuration information with the abstract syntax tree of the historical configuration information to obtain the configuration change content of the AB experiment due to the version update. This embodiment compares the configuration change content with the corresponding preset rules, and performs a change impact analysis on the configuration change content when the configuration change content complies with the preset rules to obtain the analysis results; if the configuration change content does not comply with the preset rules, it is determined that the current configuration information is abnormal. Through the above scheme, this embodiment can accurately detect the configuration change content of the AB experiment, and analyze the compliance of the configuration change content and the impact of the change, so that the staff can make preparations in advance and effectively avoid risks. Therefore, this embodiment can effectively manage the AB experiment configuration and improve the security of the AB experiment.

[0096] Furthermore, it also includes:

[0097] A version mutual exclusion module is used to generate a current version number of the current configuration information based on the current time and hash value if a change in the configuration information of the AB experiment is detected; it is also used to determine the user cluster participating in the AB experiment and eliminate the target user in the user cluster if an experiment execution instruction is received; wherein the target user is a user bound to another version number; it is also used to bind the current version number to the users in the user cluster and execute the AB experiment according to the current configuration information; it is also used to unbind the users in the user cluster from the current version number after the AB experiment is executed.

[0098] Furthermore, it also includes:

[0099] The snapshot management module is used to decompose the current configuration information of the AB experiment in the current version into multiple dimensional fields after obtaining the current configuration information, and generate a structured snapshot of the current configuration information based on all the dimensional fields; it is also used to compare the structured snapshot of the current configuration information with the structured snapshot of the historical configuration information if a version comparison instruction is received, and generate a visual difference report.

[0100] Furthermore, it also includes:

[0101] The rollback module is used to perform version rollback using the structured snapshot of the historical configuration information after determining that the current configuration information is abnormal.

[0102] Furthermore, it also includes:

[0103] The risk detection module is used to parse the business requirement document after obtaining the current configuration information of the AB experiment in the current version to obtain the value constraints corresponding to multiple business requirement items; it is also used to determine whether the current configuration information meets the value constraints; if not, generate risk warning information and mark the fields in the current configuration information that do not meet the value constraints.

[0104] Furthermore, it also includes:

[0105] A configuration correction module is used to determine whether the current configuration information meets the value constraints. If the current configuration information does not meet the value constraints, the module uses a machine learning model to generate an AB experiment prediction effect corresponding to the current configuration information; and is also used to correct the current configuration information in combination with the AB experiment prediction effect and the historical configuration information.

[0106] Furthermore, it also includes:

[0107] A rule conflict detection module is used to, after querying the preset rules corresponding to the configuration change content, determine whether there are conflicting preset rules if the number of the preset rules is greater than 1; if so, upload the conflicting preset rules to the management terminal and generate a manual review prompt message;

[0108] Correspondingly, the process of the judgment module judging whether the configuration change content complies with the preset rules includes: determining the priority of the preset rules; and judging whether the configuration change content complies with the preset rules in descending order of priority.

[0109] Since the embodiments of the system part correspond to the embodiments of the method part, please refer to the description of the embodiments of the method part for the embodiments of the system part, and will not be repeated here.

[0110] This application also provides a storage medium having a computer program stored thereon, which, when executed, can implement the steps provided in the above embodiments. The storage medium may include: a USB flash drive, a mobile hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, among other media capable of storing program code.

[0111] The present application also provides an electronic device that may include a memory and a processor, wherein the memory stores a computer program, and when the processor calls the computer program in the memory, the steps provided in the above embodiment can be implemented. Of course, the electronic device may also include various network interfaces, a power supply, and other components.

[0112] The various embodiments in the specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same and similar parts between the various embodiments can be referred to each other. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description. It should be pointed out that for ordinary technicians in this technical field, without departing from the principles of this application, several improvements and modifications can be made to this application, and these improvements and modifications also fall within the scope of protection of this application.

[0113] It should also be noted that, in this specification, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus comprising the element.

Claims

1. A configuration information detection method for AB experiment, characterized in that: include: Get the current configuration information of AB experiment in the current version; Generating a current abstract syntax tree corresponding to the current configuration information, and comparing the current abstract syntax tree with a historical abstract syntax tree to obtain configuration changes of the AB experiment; wherein the historical abstract syntax tree is generated based on historical configuration information, and the historical configuration information is configuration information of the AB experiment before the current version; Query the preset rules corresponding to the configuration change content; Determining whether the configuration change content complies with the preset rules; If so, perform a change impact analysis based on the configuration change content to obtain an analysis result; wherein the analysis result is used to describe the affected objects and impact extent of the version change; If not, it is determined that there is an abnormality in the current configuration information.

2. The configuration information detection method for AB experiment according to claim 1, characterized in that: Also includes: If it is detected that the configuration information of the AB experiment has changed, the current version number of the current configuration information is generated according to the current time and the hash value; If an experiment execution instruction is received, the user cluster participating in the AB experiment is determined, and the target user in the user cluster is removed; wherein the target user is a user bound to other version numbers; Binding the current version number to users in the user cluster and executing the AB experiment according to the current configuration information; After the AB experiment is completed, the users in the user cluster are unbound from the current version number.

3. The configuration information detection method of the AB experiment according to claim 1, characterized in that: After obtaining the current configuration information of the AB experiment in the current version, it also includes: Decomposing the current configuration information into a plurality of dimension fields, and generating a structured snapshot of the current configuration information according to all the dimension fields; If a version comparison instruction is received, the structured snapshot of the current configuration information is compared with the structured snapshot of the historical configuration information to generate a visual difference report.

4. The configuration information detection method of AB experiment according to claim 3, characterized in that: After determining that the current configuration information is abnormal, the method further includes: The structured snapshot of the historical configuration information is used to perform version rollback.

5. The configuration information detection method of AB experiment according to claim 1, characterized in that: After obtaining the current configuration information of the AB experiment in the current version, it also includes: Parse the business requirement document to obtain the value constraints corresponding to multiple business requirement items; Determining whether the current configuration information meets the value constraint condition; If not, risk warning information is generated, and the fields in the current configuration information that do not meet the value constraint conditions are marked.

6. The configuration information detection method for AB experiment according to claim 5, characterized in that: After determining whether the current configuration information meets the value constraint condition, the method further includes: If the current configuration information does not meet the value constraint condition, using a machine learning model to generate an AB experiment prediction effect corresponding to the current configuration information; The current configuration information is modified based on the AB experiment prediction result and the historical configuration information.

7. The configuration information detection method for AB experiment according to any one of claims 1 to 6, characterized in that: After querying the preset rules corresponding to the configuration change content, the following steps are also included: If the number of the preset rules is greater than 1, determining whether there are conflicting preset rules; If so, the conflicting preset rules will be uploaded to the management terminal and a manual review prompt will be generated; Accordingly, determining whether the configuration change content complies with the preset rules includes: Determining the priority of the preset rules; Determine whether the configuration change content complies with the preset rules in order of priority from high to low.

8. A configuration information detection system for AB experiment, characterized in that: include: Information acquisition module, used to obtain the current configuration information of AB experiment in the current version; a change determination module, configured to generate a current abstract syntax tree corresponding to the current configuration information, and compare the current abstract syntax tree with a historical abstract syntax tree to obtain configuration changes for the AB experiment; wherein the historical abstract syntax tree is generated based on historical configuration information, and the historical configuration information is configuration information of the AB experiment before the current version; A rule query module is used to query the preset rules corresponding to the configuration change content; A judgment module is used to determine whether the configuration change content complies with the preset rules; if so, a change impact analysis is performed based on the configuration change content to obtain an analysis result; wherein the analysis result is used to describe the affected objects and impact degree of the version change; if not, it is determined that there is an abnormality in the current configuration information.

9. An electronic device, characterized in that: The method comprises a memory and a processor, wherein a computer program is stored in the memory, and when the processor calls the computer program in the memory, the steps of the configuration information detection method of the AB experiment according to any one of claims 1 to 7 are implemented.

10. A storage medium, characterized in that: The storage medium stores computer-executable instructions, which, when loaded and executed by a processor, implement the steps of the configuration information detection method for an AB experiment according to any one of claims 1 to 7.

Images