Automatic point burying method and device, electronic equipment and storage medium
By using pre-trained semantic segmentation models to identify key nodes of the target software and dynamically insert monitoring codes in the CI/CD process, the problems of inaccurate burial points and discontinuous data acquisition in the existing technology are solved, and efficient Accurate data acquisition and normal functions of monitoring code.
Patent Information
- Application Number
- CN202411727931.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-28
- Publication Date
- 2025-02-18
AI Technical Summary
The prior art cannot accurately bury points in a timely manner, resulting in the inability to guarantee the completeness and accuracy of data acquisition, and the normal function of monitoring codes and the continuity and consistency of data acquisition.
By obtaining the requirements text data of the current business code of the target software and inputting it into the pre-trained semantic segmentation model, semantic segmentation is performed to identify key nodes. Then, during the CI/CD process, monitoring code is inserted into the target business code through dynamic code insertion to collect specified data.
It realizes automated burying points, improves the accuracy and completeness of burying points, ensures the continuity and consistency of data acquisition, and ensures the normal function of the monitoring code through testing.
Smart Images
Figure CN119473799A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of software monitoring technology, and in particular to an automated tracking method and device, an electronic device, and a storage medium. Background Art
[0002] Currently, in order to collect data from key links in the software operation process so as to analyze the performance, resource usage, and business implementation of the software when implementing the main business, it is usually necessary to embed points in the business code that implements these key links, that is, insert monitoring code to collect information on the implementation process of the key links by running the monitoring program.
[0003] Currently, one way to track points is for developers to conduct in-depth analysis of business logic based on experience and concerns about application system performance, identify key links, and manually write or modify source code to track points in the business code that implements the key links. Another way is to match key links based on preset rules and track points in the business code that implements the key links.
[0004] However, the manual tracking method is not only inefficient, but also prone to omission or misunderstanding of key links, and cannot guarantee the integrity and accuracy of tracking, which in turn leads to the integrity and accuracy of collected data. The method based on preset rules relies on the accuracy and timely update of the rules, but the business logic is diverse and complex, and it is difficult to guarantee accuracy, so the integrity and accuracy of tracking cannot be guaranteed. In addition, these two methods of tracking cannot determine whether the monitoring code is normal, and there is a certain lag in data collection. It is impossible to ensure timely adjustments as the code is updated to ensure the continuity and consistency of data collection, that is, it is impossible to guarantee the normal collection of data. Summary of the invention
[0005] Based on the above-mentioned deficiencies of the prior art, the present application provides an automated tracking method and device, an electronic device, and a storage medium to solve the problem that the prior art cannot accurately track points in a timely manner.
[0006] In order to achieve the above objectives, this application provides the following technical solutions:
[0007] The first aspect of the present application provides an automated tracking method, comprising:
[0008] Obtain the demand text data of the current business code of the target software;
[0009] The demand text data is input into a pre-trained semantic segmentation model, and the demand text data is semantically segmented by the semantic segmentation model to obtain various key nodes of the current business code of the target software; wherein the key nodes include key user behavior nodes and key business nodes; the semantic segmentation model is pre-trained using the demand text data of the historical business code and its corresponding node labels;
[0010] When the CI / CD process is performed on the current business code of the target software, whenever any target business code in the current business code of the target software is built, a monitoring code is inserted into the target business code by dynamic code insertion, so as to collect specified data generated by running the target business code when the monitoring code in the target business code is run; wherein, the target business code refers to the business code corresponding to the key node.
[0011] Optionally, in the above-mentioned automated tracking method, the following is further included:
[0012] Obtain demand text data for multiple historical business codes;
[0013] Preprocessing the demand text data of each historical business code;
[0014] Dividing the required text data of each target business code into multiple segments of text data;
[0015] Combining the text data of each segment into training samples;
[0016] A semantic segmentation model is constructed using a recurrent neural network, and the semantic segmentation model is iteratively trained using the training samples to obtain the trained semantic segmentation model.
[0017] Optionally, in the above-mentioned automated tracking method, the step of inputting the requirement text data into a pre-trained semantic segmentation model, performing semantic segmentation on the requirement text data by the semantic segmentation model, and obtaining various key nodes of the current business code of the target software includes:
[0018] Preprocessing the demand text data;
[0019] Segmenting the preprocessed demand text data to obtain multiple segments of current text data;
[0020] Performing feature processing on the current text data of each segment to obtain a feature matrix corresponding to the current text data of each segment;
[0021] Inputting the feature matrix corresponding to each segment of the current text data into the semantic segmentation model, extracting features from the feature matrix corresponding to each segment of the current text data through the semantic segmentation model, and analyzing the labels corresponding to each segment of the current text data based on the extracted features;
[0022] Each segment of the current text data whose corresponding tag belongs to the designated key tag is determined as a key node of the current business code of the target software.
[0023] Optionally, in the above-mentioned automated tracking method, inserting the monitoring code into the target business code by means of dynamic code instrumentation includes:
[0024] Analyze each code in the target service code to obtain component feature information of the target service code;
[0025] Determine the location of the target service code based on the component feature information of the target service code;
[0026] Insert the monitoring code at the embedding point of the target business code.
[0027] The second aspect of the present application provides an automated point burying device, comprising:
[0028] A first acquisition unit is used to acquire the demand text data of the current business code of the target software;
[0029] An analysis unit is used to input the demand text data into a pre-trained semantic segmentation model, and perform semantic segmentation on the demand text data through the semantic segmentation model to obtain various key nodes of the current business code of the target software; wherein the key nodes include key user behavior nodes and key business nodes; the semantic segmentation model is pre-trained using the demand text data of the historical business code and its corresponding node labels;
[0030] A tracking unit is used to insert monitoring code into the target business code by dynamic code instrumentation when performing CI / CD process processing on the current business code of the target software, whenever any target business code in the current business code of the target software is built, so as to collect specified data generated by running the target business code when running the monitoring code in the target business code; wherein, the target business code refers to the business code corresponding to the key node.
[0031] Optionally, in the above-mentioned automated point embedding device, it further includes:
[0032] A second acquisition unit is used to acquire demand text data of multiple historical business codes;
[0033] A first preprocessing unit, used for preprocessing the demand text data of each of the historical business codes;
[0034] A division unit, used for dividing the required text data of each target service code into multiple segments of text data;
[0035] A composition unit, used for composing the text data of each segment into training samples;
[0036] The training unit is used to construct a semantic segmentation model using a recurrent neural network, and iteratively train the semantic segmentation model using the training samples to obtain the trained semantic segmentation model.
[0037] Optionally, in the above-mentioned automated tracking device, the analysis unit includes:
[0038] A second preprocessing unit, used for preprocessing the demand text data;
[0039] A segmentation unit, used for segmenting the preprocessed demand text data to obtain multiple segments of current text data;
[0040] A feature processing unit, used to perform feature processing on each segment of the current text data to obtain a feature matrix corresponding to each segment of the current text data;
[0041] A label analysis unit, used for inputting a feature matrix corresponding to each segment of the current text data into the semantic segmentation model, extracting features from the feature matrix corresponding to each segment of the current text data through the semantic segmentation model, and analyzing labels corresponding to each segment of the current text data based on the extracted features;
[0042] The node determination unit is used to determine each segment of the current text data whose corresponding tag belongs to the specified key tag as a key node of the current business code of the target software.
[0043] Optionally, in the above-mentioned automated tracking device, the tracking unit includes:
[0044] A code analysis unit, used to analyze each code in the target service code to obtain component feature information of the target service code;
[0045] A location determination unit, used to determine the location of the target service code based on the component feature information of the target service code;
[0046] An insertion unit is used to insert a monitoring code at the embedding point of the target business code.
[0047] A third aspect of the present application provides an electronic device, including:
[0048] Memory and processor;
[0049] Wherein, the memory is used to store programs;
[0050] The processor is used to execute the program, and when the program is executed, it is specifically used to implement the automated tracking method as described in any one of the above items.
[0051] A fourth aspect of the present application provides a computer storage medium for storing a computer program, which, when executed by a processor, is used to implement an automated tracking method as described in any one of the above.
[0052] The present application provides an automated tracking method to obtain the demand text data of the current business code of the target software. The demand text data is input into a pre-trained semantic segmentation model, and the demand text data is semantically segmented by the semantic segmentation model to obtain the key nodes of the current business code of the target software. Among them, the key nodes include key user behavior nodes and key business nodes. The semantic segmentation model is pre-trained using the demand text data of the historical business code and its corresponding node labels, so that the demand text of the code is deeply analyzed by the pre-trained semantic segmentation model to identify the key links of the code, and no longer rely on the set rules, thereby effectively ensuring the accuracy of the identification, and then ensuring the accuracy of the tracking. Then, when the current business code of the target software is processed by the CI / CD process, whenever any target business code in the current business code of the target software is built, the monitoring code is inserted into the target business code by dynamic code insertion, so that when the monitoring code in the target business code is run, the specified data generated by running the target business code is collected. Among them, the target business code refers to the business code corresponding to the key node, so that the dynamic tracking process is integrated into the CI / CD process, so that in the CI / CD process processing, dynamic tracking is carried out accordingly during the construction, and the business code inserted into the monitoring code will be tested and deployed accordingly in the subsequent process, so that the tracking can be updated in time, ensuring the continuity and consistency of the collected data, and the normal function of the monitoring code can also be guaranteed through testing. BRIEF DESCRIPTION OF THE DRAWINGS
[0053] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.
[0054] Figure 1A flowchart of an automated tracking method provided in an embodiment of the present application;
[0055] Figure 2 A flowchart of a method for training a semantic segmentation model provided in an embodiment of the present application;
[0056] Figure 3 A flowchart of a method for analyzing key nodes through a semantic segmentation model provided in an embodiment of the present application;
[0057] Figure 4 A flowchart of a method for inserting monitoring code into a target business code by dynamic code insertion provided in an embodiment of the present application;
[0058] Figure 5 A schematic diagram of the architecture of an automated tracking device provided in an embodiment of the present application;
[0059] Figure 6 A schematic diagram of the architecture of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION
[0060] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.
[0061] In this application, 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 such actual relationship or order between these entities or operations. Moreover, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the statement "comprises a ..." does not exclude the presence of other identical elements in the process, method, article or device including the element.
[0062] This embodiment provides an automated tracking method. Figure 1 As shown, the specific steps include:
[0063] S101, obtaining the demand text data of the current business code of the target software.
[0064] The target software can also be any software that needs to be tracked and data collected. The current business code of the target software refers to the latest version of the target software that has not yet been deployed. The demand text data of the current business code is the description text of the demand corresponding to the current business code, and the demand is the function that the business logic needs to implement. Therefore, the demand text data based on the current business code can share the key links to be implemented in the current business code.
[0065] S102: Input the requirement text data into a pre-trained semantic segmentation model, perform semantic segmentation on the requirement text data through the semantic segmentation model, and obtain various key nodes of the current business code of the target software.
[0066] Among them, key nodes include key user behavior nodes and key business nodes. Key user behavior nodes are key behaviors in the process of users operating software to perform business, such as users collecting benefits and redeeming benefits in marketing activities. Key business nodes are key nodes in the business functions implemented by the software, such as benefit generation and benefit issuance. The semantic segmentation model is pre-trained using the demand text data of historical business codes and their corresponding node labels. Among them, the node label is the type label of the specific key node.
[0067] It should be noted that in order to no longer rely on set rules and conduct in-depth analysis to effectively ensure the accuracy of key nodes of analysis, in the embodiment of the present application, the semantic segmentation model after iterative training is used to process the demand text data, which no longer needs to rely on set rules and can handle various complex business logics. In addition, it conducts in-depth analysis of the demand text data, which can automatically identify key entities, time, and relationship features in the text data, thereby determining the various key nodes involved in the demand text data, and then correspondingly determining the various key nodes contained in the current business code of the target software.
[0068] Optionally, in another embodiment of the present application, a method for training a semantic segmentation model is provided, such as Figure 2 As shown, the following steps are included:
[0069] S201. Obtain demand text data of multiple historical business codes.
[0070] S202: Preprocess the demand text data of each historical business code.
[0071] Specifically, the demand text data of the historical business codes may be preprocessed by cleaning, word segmentation, stop word removal, and part-of-speech restoration, thereby ensuring the quality of the demand text data of the historical business codes.
[0072] S203: Divide the required text data of each target business code into multiple segments of text data.
[0073] Optionally, the data can be divided by sentence, that is, each sentence of text data is a piece of text data, or divided by paragraph, each paragraph of text data is a piece of text data, or divided in other ways. Since each piece of text data is usually a description of different requirements, the required text data of the target business code needs to be divided for separate analysis.
[0074] S204, assembling each segment of text data into training samples.
[0075] S205. Construct a semantic segmentation model using a recurrent neural network, and iteratively train the semantic segmentation model using training samples to obtain a trained semantic segmentation model.
[0076] In the embodiment of the present application, a recurrent neural network (RNN) is used to construct a semantic segmentation model. RNN is a neural network that is particularly suitable for processing sequence data, such as text, time series, etc. In RNN, the input of each time step takes into account the information of the previous time step, which enables RNN to capture long-term dependencies in the text, thereby enabling more accurate analysis.
[0077] Optionally, LSTM can be used as a variant of RNN, and a two-layer LSTM model can be constructed.
[0078] Specifically, during training, each segment of text data is input as an input sequence into the semantic segmentation model. Each input sequence is labeled with a corresponding label as the output target of the target model. Then, through the back propagation algorithm and the optimization algorithm, the number of adoptions of the semantic segmentation model is adjusted inappropriately, so that the model can accurately predict the labels of users in the input sequence.
[0079] Optionally, in the embodiment of the present application, a specific implementation of step S102 is as follows: Figure 3 As shown, the following steps are included:
[0080] S301. Preprocess the demand text data.
[0081] Optionally, specific preprocessing methods may include but are not limited to operations such as word segmentation and removal of stop words.
[0082] S302: Segment the preprocessed demand text data to obtain multiple segments of current text data.
[0083] Since each segment of text is used to express different requirements and thus corresponds to different implementation codes, the requirement text data needs to be segmented so that each segment of current text data can be processed separately.
[0084] S303: Perform feature processing on each segment of current text data to obtain a feature matrix corresponding to each segment of current text data.
[0085] Specifically, each word segment in each segment of current text data can be converted into a vector form to obtain a feature matrix corresponding to each segment of current text data. For example, the dimension of each word is 100, and a segment of text data includes 20 words, then the feature matrix corresponding to the segment of text data can be expressed as a 20*100 feature matrix.
[0086] S304, inputting the feature matrix corresponding to each segment of current text data into the semantic segmentation model, extracting features from the feature matrix corresponding to each segment of current text data through the semantic segmentation model, and analyzing the labels corresponding to each segment of current text data based on the extracted features.
[0087] S305: Determine each segment of current text data whose corresponding tag belongs to the designated key tag as a key node of the current business code of the target software.
[0088] S103. When the CI / CD process is performed on the current business code of the target software, whenever any target business code in the current business code of the target software is built, the monitoring code is inserted into the target business code by dynamic code insertion, so as to collect the specified data generated by running the target business code when the monitoring code in the target business code is run.
[0089] The target business code refers to the business code corresponding to the key node. Since the key node is implemented by the business code, after the key node is determined, the code that implements the key node can be determined and used as the business code corresponding to the key node.
[0090] It should be noted that the CI / CD process refers to continuous integration and continuous deployment / delivery, which is a software development process model that aims to improve the efficiency and quality of software development and release through automated and continuous building, testing, deployment and delivery processes. Therefore, in order to ensure timely and dynamic adjustment of tracking points, deployment and delivery, thereby ensuring the continuity and consistency of data collection, and also to ensure the accuracy of monitoring code input through testing, in the embodiment of the present application, the tracking point process is integrated into the CI / CD process.
[0091] In addition, in order to achieve dynamic tracking, in the embodiment of the present application, dynamic code insertion is used for tracking. Among them, the dynamic code insertion method can dynamically insert or modify the code when the program is running, without intruding the source code of the program, so as to monitor, debug or enhance the program. Specifically, the dynamic code insertion technology analyzes and modifies the binary code or bytecode of the target program, and inserts the monitoring code into the key business functions or methods.
[0092] So specifically, when the developer submits a new business code, that is, submits the current business code of the target software for CI / CD process processing, it will automatically trigger the build, test, and deployment processes. During the build process, the tracking code will be automatically inserted into the new business code, so that in the subsequent testing process, the business code containing the tracking code will be functionally tested and performance tested to ensure its correctness and stability. And when the test passes, the business code containing the tracking code will be automatically deployed to the production environment, thereby realizing real-time monitoring and data collection of user behavior. That is, after the monitoring code is inserted and deployed, when the target program is running, whenever it runs to each section of the target business code, the monitoring code in the target business code will be executed, thereby collecting the data generated during the execution of the target business code. This means that whenever a user performs a key business operation, the monitoring code will be triggered to collect relevant user behavior data.
[0093] Optionally, the collected data can be sent to a back-end server, where the data is further processed and analyzed to extract the required information.
[0094] Optionally, in another embodiment of the present application, a monitoring code is inserted into the target business code by dynamic code insertion. Figure 4 As shown, the following steps are included:
[0095] S401. Analyze each code in the target business code to obtain component feature information of the target business code.
[0096] Specifically, the binary code and bytecode in the target business code are analyzed to determine the structure, functions, methods and the calling relationship between them of the code, so as to obtain the composition feature information of the target business code.
[0097] S402: Determine the location of the target business code based on the component feature information of the target business code.
[0098] After analyzing the target program, the next key step is to determine the location of the code insertion. These locations are usually key functions or methods in the business logic, such as browsing activities, participating in activities, claiming benefits, and canceling benefits. Selecting these locations can ensure that the collected user behavior data is representative and accurate. Therefore, the composition feature information of the target business code can be used to determine the key functions or methods in the target business code, and then select the corresponding locations of the key functions or methods as the embedding locations.
[0099] S403. Insert the monitoring code at the embedding location of the target business code.
[0100] The embodiment of the present application provides an automated tracking method to obtain the demand text data of the current business code of the target software. The demand text data is input into a pre-trained semantic segmentation model, and the demand text data is semantically segmented by the semantic segmentation model to obtain the key nodes of the current business code of the target software. Among them, the key nodes include key user behavior nodes and key business nodes. The semantic segmentation model is pre-trained using the demand text data of the historical business code and its corresponding node labels, so that the demand text of the code is deeply analyzed by the pre-trained semantic segmentation model to identify the key links of the code, and no longer rely on the set rules, thereby effectively ensuring the accuracy of the identification, and then ensuring the accuracy of the tracking. Then, when the current business code of the target software is processed by the CI / CD process, whenever any target business code in the current business code of the target software is built, the monitoring code is inserted into the target business code by dynamic code insertion, so that when the monitoring code in the target business code is run, the specified data generated by running the target business code is collected. Among them, the target business code refers to the business code corresponding to the key node, so that the dynamic tracking process is integrated into the CI / CD process, so that in the CI / CD process processing, dynamic tracking is carried out accordingly during the construction, and the business code inserted into the monitoring code will be tested and deployed accordingly in the subsequent process, so that the tracking can be updated in time, ensuring the continuity and consistency of the collected data, and the normal function of the monitoring code can also be guaranteed through testing.
[0101] Another embodiment of the present application provides an automated point-burying device, such as Figure 5 As shown, including:
[0102] The first acquisition unit 501 is used to acquire the requirement text data of the current business code of the target software.
[0103] The analysis unit 502 is used to input the requirement text data into a pre-trained semantic segmentation model, perform semantic segmentation on the requirement text data through the semantic segmentation model, and obtain various key nodes of the current business code of the target software.
[0104] Among them, key nodes include key user behavior nodes and key business nodes. The semantic segmentation model is pre-trained using the demand text data of historical business codes and their corresponding node labels.
[0105] The tracking unit 503 is used to insert monitoring code into the target business code by dynamic code instrumentation when the CI / CD process is performed on the current business code of the target software. When any target business code in the current business code of the target software is built, the monitoring code is collected in the target business code when the monitoring code in the target business code is run. The specified data generated by running the target business code.
[0106] The target business code refers to the business code corresponding to the key node.
[0107] Optionally, in another embodiment of the present application, the automated tracking device further includes:
[0108] The second acquisition unit is used to acquire demand text data of multiple historical business codes.
[0109] The first preprocessing unit is used to preprocess the demand text data of each historical business code.
[0110] The division unit is used to divide the required text data of each target business code into multiple segments of text data.
[0111] The composition unit is used to combine each segment of text data into training samples.
[0112] The training unit is used to construct a semantic segmentation model using a recurrent neural network, and iteratively train the semantic segmentation model using training samples to obtain a trained semantic segmentation model.
[0113] Optionally, in the automated tracking device provided in another embodiment of the present application, the analysis unit includes:
[0114] The second preprocessing unit is used to preprocess the demand text data.
[0115] The segmentation unit is used to segment the preprocessed demand text data to obtain multiple segments of current text data.
[0116] The feature processing unit is used to perform feature processing on each segment of current text data to obtain a feature matrix corresponding to each segment of current text data.
[0117] The label analysis unit is used to input the feature matrix corresponding to each segment of current text data into the semantic segmentation model, extract features from the feature matrix corresponding to each segment of current text data through the semantic segmentation model, and analyze the label corresponding to each segment of current text data based on the extracted features.
[0118] The node determination unit is used to determine each segment of current text data whose corresponding tag belongs to the specified key tag as a key node of the current business code of the target software.
[0119] Optionally, in the automated tracking device provided in another embodiment of the present application, the tracking unit includes:
[0120] The code analysis unit is used to analyze each code in the target business code to obtain component feature information of the target business code.
[0121] The location determination unit is used to determine the location of the target business code based on the component feature information of the target business code.
[0122] The insertion unit is used to insert the monitoring code at the tracking point position of the target business code.
[0123] It should be noted that the specific working process of each unit provided in the above embodiments of the present application can refer to the implementation process of the corresponding steps in the above method embodiments, and will not be repeated here.
[0124] Another embodiment of the present application provides an electronic device, such as Figure 6 As shown, including:
[0125] Memory 601 and processor 602 .
[0126] The memory 601 is used to store programs.
[0127] The processor 602 is used to execute the program stored in the memory 601. When the program is executed, it is specifically used to implement the automated tracking method provided in any of the above embodiments.
[0128] Another embodiment of the present application provides a computer storage medium for storing a computer program, which, when executed by a processor, is used to implement any of the automated tracking methods described above.
[0129] Computer storage media includes permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include temporary computer-readable media (transitory media), such as modulated data signals and carrier waves.
[0130] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.
[0131] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. An automated tracking method, characterized in that: include: Obtain the demand text data of the current business code of the target software; The demand text data is input into a pre-trained semantic segmentation model, and the demand text data is semantically segmented by the semantic segmentation model to obtain various key nodes of the current business code of the target software; wherein the key nodes include key user behavior nodes and key business nodes; the semantic segmentation model is pre-trained using the demand text data of the historical business code and its corresponding node labels; When the CI / CD process is performed on the current business code of the target software, whenever any target business code in the current business code of the target software is built, a monitoring code is inserted into the target business code by dynamic code insertion, so as to collect specified data generated by running the target business code when the monitoring code in the target business code is run; wherein, the target business code refers to the business code corresponding to the key node.
2. The method according to claim 1, characterized in that Also includes: Obtain demand text data for multiple historical business codes; Preprocessing the demand text data of each historical business code; Dividing the required text data of each target business code into multiple segments of text data; Combining the text data of each segment into training samples; A semantic segmentation model is constructed using a recurrent neural network, and the semantic segmentation model is iteratively trained using the training samples to obtain the trained semantic segmentation model.
3. The method according to claim 1, characterized in that The step of inputting the requirement text data into a pre-trained semantic segmentation model, performing semantic segmentation on the requirement text data through the semantic segmentation model, and obtaining various key nodes of the current business code of the target software includes: Preprocessing the demand text data; Segmenting the preprocessed demand text data to obtain multiple segments of current text data; Performing feature processing on the current text data of each segment to obtain a feature matrix corresponding to the current text data of each segment; Inputting the feature matrix corresponding to each segment of the current text data into the semantic segmentation model, extracting features from the feature matrix corresponding to each segment of the current text data through the semantic segmentation model, and analyzing the labels corresponding to each segment of the current text data based on the extracted features; Each segment of the current text data whose corresponding tag belongs to the designated key tag is determined as a key node of the current business code of the target software.
4. The method according to claim 1, characterized in that: The inserting of the monitoring code into the target business code by dynamic code insertion includes: Analyze each code in the target service code to obtain component feature information of the target service code; Determine the location of the target service code based on the component feature information of the target service code; Insert the monitoring code at the embedding point of the target business code.
5. An automated point embedding device, characterized in that: include: A first acquisition unit is used to acquire the demand text data of the current business code of the target software; An analysis unit is used to input the demand text data into a pre-trained semantic segmentation model, and perform semantic segmentation on the demand text data through the semantic segmentation model to obtain various key nodes of the current business code of the target software; wherein the key nodes include key user behavior nodes and key business nodes; the semantic segmentation model is pre-trained using the demand text data of the historical business code and its corresponding node labels; A tracking unit is used to insert monitoring code into the target business code by dynamic code instrumentation when performing CI / CD process processing on the current business code of the target software, whenever any target business code in the current business code of the target software is built, so as to collect specified data generated by running the target business code when running the monitoring code in the target business code; wherein, the target business code refers to the business code corresponding to the key node.
6. The device according to claim 5, characterized in that Also includes: A second acquisition unit is used to acquire demand text data of multiple historical business codes; A first preprocessing unit, used for preprocessing the demand text data of each of the historical business codes; A division unit, used for dividing the required text data of each target service code into multiple segments of text data; A composition unit, used for composing the text data of each segment into training samples; The training unit is used to construct a semantic segmentation model using a recurrent neural network, and iteratively train the semantic segmentation model using the training samples to obtain the trained semantic segmentation model.
7. The device according to claim 5, characterized in that The analysis unit comprises: A second preprocessing unit, used for preprocessing the demand text data; A segmentation unit, used for segmenting the preprocessed demand text data to obtain multiple segments of current text data; A feature processing unit, used to perform feature processing on each segment of the current text data to obtain a feature matrix corresponding to each segment of the current text data; A label analysis unit, used for inputting a feature matrix corresponding to each segment of the current text data into the semantic segmentation model, extracting features from the feature matrix corresponding to each segment of the current text data through the semantic segmentation model, and analyzing labels corresponding to each segment of the current text data based on the extracted features; The node determination unit is used to determine each segment of the current text data whose corresponding tag belongs to the specified key tag as a key node of the current business code of the target software.
8. The device according to claim 5, characterized in that The embedding unit comprises: A code analysis unit, used to analyze each code in the target service code to obtain component feature information of the target service code; A location determination unit, used to determine the location of the target service code based on the component feature information of the target service code; An insertion unit is used to insert a monitoring code at the embedding point of the target business code.
9. An electronic device, characterized in that: include: Memory and processor; Wherein, the memory is used to store programs; The processor is used to execute the program, and when the program is executed, it is specifically used to implement the automated tracking method as described in any one of claims 1 to 4.
10. A computer storage medium, characterized in that: Used to store a computer program, which, when executed by a processor, is used to implement the automated tracking method as described in any one of claims 1 to 4.