Page analysis method and device, electronic equipment, storage medium and program product

Abstract

The embodiment of the invention provides a page analysis method and device, electronic equipment, a storage medium and a program product, and the page analysis method comprises the steps that a burying point processing request for a target page is received, and the burying point processing request comprises page data of the target page; inputting the page data into a pre-trained page analysis model to obtain buried point processing data of the target page, the buried point processing data including buried point identifiers and operation information corresponding to target elements in the target page, the buried point identifiers being used for representing operation types for the target elements, and the operation information being used for representing operation types for the target elements; the operation information is used for representing operation content for the target element; and based on the operation type and the operation content, performing burying point configuration on the page data. According to the method, the page analysis request is automatically processed, the pre-trained model is utilized to analyze the page data and generate the burying point data, and the corresponding burying point identifier and information are added to each element in the page, so that the workload of manual configuration is reduced, and the burying point deployment speed is increased.

Description

Technical Field

[0001] Embodiments of the present invention relate to the field of computer technology, and in particular to a page analysis method, device, electronic device, storage medium, and program product. Background Art

[0002] The embedding technology is to embed codes in specific locations of the user interface to capture user behavior and identity information. This information is transmitted to the server through the network, and after analysis and processing, it provides companies with detailed reports on user behavior patterns and preferences. This method not only helps companies optimize products and services, but also allows for accurate market forecasting and user segmentation.

[0003] In the existing technology, tracking work usually requires specialized technical personnel to carry out, which not only increases the manpower cost of the enterprise, but also requires these personnel to have a deep understanding of the business and strong technical capabilities. On the other hand, the implementation of tracking work often involves coordination and communication between multiple departments. For example, the product department needs to clarify which user behavior data is valuable, and the technical department must ensure that this data can be effectively captured and transmitted. This cross-departmental cooperation often leads to an increase in communication costs, further affecting the efficiency of project promotion. Summary of the invention

[0004] In view of this, an embodiment of the present invention provides a page analysis method. One or more embodiments of the present invention also relate to a page analysis device, a computing device, a computer-readable storage medium and a computer program product to solve the technical defects existing in the prior art.

[0005] According to a first aspect of an embodiment of the present invention, a page analysis method is provided, comprising: receiving a point-burying processing request for a target page, wherein the point-burying processing request includes page data of the target page; inputting the page data into a pre-trained page analysis model to obtain point-burying processing data of the target page, wherein the point-burying processing data includes a point-burying identifier and operation information corresponding to a target element in the target page, the point-burying identifier is used to characterize the type of operation on the target element, and the operation information is used to characterize the content of the operation on the target element; and performing point-burying configuration on the page data based on the operation type and the operation content.

[0006] According to a second aspect of an embodiment of the present invention, a page analysis device is provided, comprising: a receiving module, configured to receive a point-burying processing request for a target page, wherein the point-burying processing request includes page data of the target page; an input module, configured to input the page data into a pre-trained page analysis model to obtain point-burying processing data of the target page, wherein the point-burying processing data includes a point-burying identifier and operation information corresponding to a target element in the target page, the point-burying identifier is used to characterize the type of operation for the target element, and the operation information is used to characterize the content of the operation for the target element; a configuration module, configured to perform point-burying configuration on the page data based on the operation type and the operation content.

[0007] According to a third aspect of an embodiment of the present invention, there is provided a computing device, comprising: a memory and a processor; the memory is used to store computer programs / instructions, and the processor is used to execute the computer programs / instructions, which implement the steps of the above-mentioned page method when executed by the processor.

[0008] According to a fourth aspect of an embodiment of the present invention, a computer-readable storage medium is provided, which stores a computer program / instruction, and the computer program / instruction implements the steps of the above-mentioned page analysis method when executed by a processor.

[0009] According to a fifth aspect of an embodiment of the present invention, there is provided a computer program product, comprising a computer program / instruction, which implements the steps of the above-mentioned page analysis method when executed by a processor.

[0010] An embodiment of the present invention realizes receiving a point-burying processing request for a target page, wherein the point-burying processing request includes page data of the target page; inputting the page data into a pre-trained page analysis model to obtain the point-burying processing data of the target page, wherein the point-burying processing data includes a point-burying identifier and operation information corresponding to a target element in the target page, the point-burying identifier is used to characterize the operation type for the target element, and the operation information is used to characterize the operation content for the target element; and configuring the page data for point-burying based on the operation type and the operation content. By automatically processing page analysis requests, using a pre-trained model to parse page data and generate point-burying data, and adding corresponding point-burying identifiers and information to each element in the page, the workload of manual configuration is reduced, the point-burying deployment speed is accelerated, and the error rate of manual deployment is reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 is an interactive schematic diagram of a page analysis system provided by an embodiment of the present invention; Figure 2 is a flow chart of a page analysis method provided by an embodiment of the present invention; Figure 3 is a schematic diagram of an interactive analysis interface provided by an embodiment of the present invention; Figure 4 is a schematic diagram of another interactive analysis interface provided by an embodiment of the present invention; Figure 5 is a process flow chart of a page analysis method provided by an embodiment of the present invention; Figure 6 is a structural schematic diagram of a page analysis device provided by an embodiment of the present invention; Figure 7 It is a structural block diagram of a computing device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0012] Many specific details are set forth in the following description to facilitate a full understanding of the present invention. However, the present invention can be implemented in many other ways than those described herein, and those skilled in the art can make similar generalizations without violating the connotation of the present invention, so the present invention is not limited to the specific implementation disclosed below.

[0013] The terms used in one or more embodiments of the present invention are only for the purpose of describing specific embodiments, and are not intended to limit one or more embodiments of the present invention. The singular forms of "a", "said" and "the" used in one or more embodiments of the present invention and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used in one or more embodiments of the present invention refers to and includes any or all possible combinations of one or more associated listed items.

[0014] It should be understood that, although the terms first, second, etc. may be used to describe various information in one or more embodiments of the present invention, these information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of one or more embodiments of the present invention, the first may also be referred to as the second, and similarly, the second may also be referred to as the first. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0015] In addition, it should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in one or more embodiments of the present invention are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0016] First, the terms involved in one or more embodiments of the present invention are explained.

[0017] Tracking: refers to a technical method in software development and data analysis, which is used to record the specific behavior of users in applications or websites.

[0018] Static code analysis: A technique that analyzes source code without executing the program.

[0019] AST: AbstractSyntaxTree, abstract syntax tree. It is a tree structure used to represent the syntax structure of source code.

[0020] SVM: Support Vector Machine. It is a supervised learning model mainly used for classification and regression analysis. Random Forest: An ensemble learning method that builds multiple decision trees to form a "forest" and then uses a voting mechanism to make the final prediction.

[0021] In the present invention, a page analysis method is provided. The present invention also relates to a page analysis system, a page analysis device, a computing device, a computer-readable storage medium, and a computer program product, which are described in detail one by one in the following embodiments.

[0022] See also Figure 1 , Figure 1 It is an interactive schematic diagram of a page analysis system provided by an embodiment of the present invention, wherein the page analysis system includes a user end and a server end, wherein the user end is used to initiate a point-of-view analysis request for a target page to the server end, and the server end is used to receive a point-of-view processing request for the target page, wherein the point-of-view processing request includes page data of the target page; the page data is input into a pre-trained page analysis model to obtain point-of-view processing data for the target page, wherein the point-of-view processing data includes a point-of-view identifier and operation information corresponding to a target element in the target page, wherein the point-of-view identifier is used to characterize an operation type for the target element, and the operation information is used to characterize an operation content for the target element; and the page data is configured for point-of-view based on the operation type and the operation content.

[0023] Applied to the page analysis system, automated page analysis and tracking configuration are achieved through the collaboration between the user end and the server end. The user end is responsible for initiating tracking analysis requests for the target page to the server end. The server end receives and processes these requests, including the page data of the target page. The server end inputs these data into a pre-trained page analysis model, which can automatically parse the page data and generate tracking processing data including tracking identifiers and operation information of the target element. The tracking identifier is used to characterize the operation type, while the operation information describes the operation content. Based on these automatically generated tracking data, the server end automatically configures tracking for the page data. This automated processing method significantly reduces the workload of manual configuration, increases the speed of tracking deployment, and reduces the error rate caused by manual operation.

[0024] See also Figure 2 , Figure 2 It is a flow chart of a page analysis method provided by an embodiment of the present invention, which specifically includes the following steps.

[0025] Step 202: Receive a tracking point processing request for a target page, wherein the tracking point processing request includes page data of the target page.

[0026] Among them, the point processing request refers to the request received by the system to process the user behavior data in the target page; the page data refers to all information related to the target page, including but not limited to the URL of the page, the page content, the user's operation records on the page, etc.

[0027] In actual applications, the system first needs to receive a tracking request for the target page. This request usually comes from a front-end application or web page, which contains the page data of the target page. The page data can include basic information of the page (such as the URL and title of the page), page content (such as HTML, CSS, JavaScript code, etc.), and user behavior data on the page (such as click events, scroll events, form submissions, etc.). The purpose of receiving tracking requests is to record and process user behavior on the target page for subsequent data analysis, user behavior research, performance monitoring, etc.

[0028] The specific steps are as follows: The system needs to listen to the tracking processing requests sent by the front end. These requests can be sent via the HTTP / HTTPS protocol, usually submitted in POST mode, carrying a request body containing page data. The request body contains the page data of the target page, which can be in JSON format or other structured data formats to facilitate system parsing and processing. After the system receives the request, it needs to parse the page data in the request body. This may include extracting the page URL, page content, user behavior data, etc. The user behavior data in the page data usually includes the elements clicked by the user, the scrolling position, the duration of stay, etc. These are important behavior indicators.

[0029] As for the implementation of the steps, one implementation method is that the system receives the request sent by the front end through the API interface. When the user performs an operation (such as clicking a button), the front end generates an event and encapsulates the event information into a request body and sends it to the back end. Another implementation method is to use WebSocket or other long connection protocols to achieve real-time data transmission. This method allows the front end to send data to the back end at the moment of the user's operation, improving the real-time performance of data transmission.

[0030] In a specific embodiment of the present invention, suppose there is an e-commerce website, and a user clicks the "Add to Cart" button when browsing the product details page. The front-end application will send a buried point processing request to the back-end server. The request contains the page data of the product details page, including the URL of the page, the timestamp of the user clicking the "Add to Cart" button, the click location and other information. After receiving the request, the back-end system parses out the key information in the page data, such as product ID, user ID, click time, etc., and stores these data in the database. At the same time, the system can also perform preliminary analysis on these data, such as counting the frequency of users clicking the "Add to Cart" button, and feedback the processing results to the front-end application.

[0031] Step 204: Input the page data into a pre-trained page analysis model to obtain the embedded point processing data of the target page, wherein the embedded point processing data includes the embedded point identifier and operation information corresponding to the target element in the target page, the embedded point identifier is used to characterize the operation type for the target element, and the operation information is used to characterize the operation content for the target element.

[0032] In actual applications, the system first needs to input the received page data of the target page into a pre-trained page analysis model. The page analysis model is usually a machine learning model or a deep learning model, which can identify the user's operation behavior on the page based on the features in the page data and extract useful information from it. The page data includes the basic information of the page (such as the URL and title of the page), the page content (such as HTML, CSS, JavaScript code, etc.), and the user's behavior data on the page (such as click events, scroll events, form submissions, etc.).

[0033] The specific steps are as follows: The system inputs the received page data of the target page into the pre-trained page analysis model. The page data needs to be pre-processed, such as removing irrelevant information and standardizing the data format, to ensure that the model can correctly read and process the data. The page analysis model analyzes the input page data, identifies the target elements in the page, and extracts the user operation information related to these elements. The target elements can be interactive components such as buttons, links, text boxes, etc. on the page. Based on the features in the page data, the model identifies what operations the user has performed on these elements, such as clicking, scrolling, filling out forms, etc.

[0034] The tracking point processing data is extracted from the results of the model analysis. These data include the tracking point identifier and operation information corresponding to the target element in the target page. The tracking point identifier is used to represent the type of operation on the target element, such as click, scroll, etc.; the operation information is used to represent the operation content on the target element, such as the specific location of the click, the scrolling distance, etc. For a button click event on the page, the page analysis model will recognize that this is a button element and the user clicked the button. The model will generate a tracking point identifier to indicate that this is a click operation, and record the specific location of the click and other related information.

[0035] For example, for a button click event on a page, the page analysis model will recognize that this is a button element and that the user clicked the button. The model will generate a tracking point identifier to indicate that this is a click operation, and will also record the specific location of the click, timestamp, and other operation information. The system will store these tracking point processing data for subsequent data analysis.

[0036] In a specific embodiment of the present invention, suppose there is an e-commerce website, and a user clicks the "Add to Cart" button when browsing the product details page. The front-end application sends a tracking point processing request to the back-end server, and the request contains the page data of the product details page. After receiving the request, the back-end system inputs the page data into the pre-trained page analysis model. The model recognizes the "Add to Cart" button on the page and analyzes that the user clicked on the button. The model generates a tracking point identifier, indicating that this is a click operation, and records the specific location of the click, timestamp and other operation information. The system stores these tracking point processing data for subsequent data analysis.

[0037] For example, when a user clicks the "Add to Cart" button, the front-end sends a request containing page data to the back-end. The back-end system inputs this data into the page analysis model, and the model recognizes that the user clicked the "Add to Cart" button and generates the corresponding tracking point identifier and operation information. The tracking point identifier indicates that this is a click operation, and the operation information includes detailed information such as the timestamp of the click and the click location. This data will be stored for subsequent analysis.

[0038] Step 206: Based on the operation type and operation content, configure the page data.

[0039] The operation type refers to the different forms of user interaction that trigger tracking points, such as click, scroll, form submission, etc. The operation content refers to the specific elements related to the operation type, such as button ID, link URL, etc.

[0040] In actual applications, you first need to define which user behaviors need to be tracked, including but not limited to click events, page visits, form submissions, etc. Then, determine the specific operation content for each type of operation. For example, for a button click event, the operation content can be the button ID or class name. Next, add the tracking code to the corresponding position of the page according to the defined operation type and content. The tracking code is usually a small JavaScript script that is triggered when the user performs a specified operation and collects relevant information. The tracking script needs to work with the backend service to send the collected data to the server for storage. For the tracking configuration of page data, you can use hard coding to directly write the tracking code into the HTML or JavaScript file of the page, or you can use configuration tracking to dynamically control the tracking logic through an external configuration file to reduce changes to the front-end code. For the implementation of the steps, one implementation method is to dynamically inject the tracking script every time the page is loaded. The advantage of this method is that the tracking logic can be flexibly adjusted without re-publishing the front-end code; another implementation method is to write the tracking logic into the front-end code during the development stage, which can ensure the consistency and efficiency of the tracking, but the front-end application may need to be redeployed when adjusting the tracking logic.

[0041] In a specific embodiment of the present invention, assume that it is necessary to monitor the product details page on an e-commerce website, and the goal is to track the user's behavior of clicking the "Add to Cart" button. First, define the operation type as a click event, and the operation content is the button ID, such as "add-to-cart-button". Next, find the DOM element corresponding to the button in the front-end code, and add the tracking code in its click event handler. The information collected by the tracking code may include the button ID, user ID (if logged in), click timestamp, etc. Finally, this information is sent to the back-end server via an HTTP request for storage and further processing. In addition, a global event listener can be set to capture all click events of elements with the "add-to-cart-button" class name, thereby covering all similar buttons on the page and implementing a unified tracking logic.

[0042] Furthermore, based on the operation type and the operation content, the page data is configured with a tracking point, including: obtaining the front-end code corresponding to the page data; generating the tracking point code based on the operation type and the operation content; and incorporating the tracking point code into the front-end code.

[0043] The front-end code refers to the part that implements the user interface of a web page or application, including HTML, CSS, JavaScript and other technology stacks. The tracking code is a code snippet specifically used to collect user behavior data. It is usually embedded in the front-end code to trigger and collect data when a specific user action occurs.

[0044] In actual applications, first, obtain the front-end code corresponding to the page data. This step usually involves finding and accessing the file system or version control system (such as Git repository) where the front-end code is stored, and locating the specific page file that needs to be embedded, such as HTML documents or JavaScript script files. The process of obtaining the code can be to open the relevant files in the development environment, or to check out the latest code version from the version control system.

[0045] Next, based on the operation type and operation content, the tracking code is generated. The operation type refers to the user's interactive actions on the page, such as clicking a button, sliding the page, filling out a form, etc.; the operation content includes the specific element identifiers related to these actions, such as the button ID, the name of the form field, etc. The system writes a piece of JavaScript code based on the pre-defined operation type and content. This code will run when the user performs a specific operation and collect the necessary data. For example, if the system needs to track the number of times a button is clicked, it will add a click event listener to the button, call the tracking API in the listener's callback function, and send a message containing the button ID and click time to the backend service. For different operation types, the writing of the tracking code will also be different. For click events, the system can listen to the click event of the DOM element; for page scrolling, it can listen to the scroll event, etc.

[0046] Then, merge the generated tracking code into the front-end code. This step requires inserting the newly written tracking code into the appropriate location of the front-end code to ensure that the tracking code can be executed correctly when the user performs the specified operation. For example, you can add tracking code inside the button click event handler, or dynamically register an event listener after the page is loaded. For some complex tracking logic, you may also need to introduce the tracking SDK to simplify the writing process of the tracking code by calling the API provided by the SDK.

[0047] In a specific embodiment of the present invention, it is assumed that the user's behavior of clicking the "Add to Cart" button on the product details page of an e-commerce website needs to be monitored. First, the system accesses the folder where the front-end code is stored, and finds the corresponding product details page HTML file and the related JavaScript file. Then, the system generates the embedding code based on the operation type of the click event and the operation content of the button ID. The system adds a click event listener to the "Add to Cart" button, and writes code in the callback function of the listener. When the button is clicked, the button ID and click time and other information are collected and sent to the back-end service. Finally, the system adds the above embedding code to the appropriate location in the JavaScript file of the page to ensure that the embedding code can be triggered and executed when the button is clicked.

[0048] Based on this, by making appropriate modifications to the front-end code corresponding to the page data and adding tracking code, the system can automatically track the user's specific operations. These operation data can help analyze user behavior, thereby improving user experience and enhancing the market competitiveness of products. The system can monitor user activities in real time, providing a solid foundation for subsequent data analysis and decision-making.

[0049] Furthermore, after configuring the page data for tracking based on the operation type and operation content, it also includes: collecting tracking information corresponding to the interactive operation in response to the front-end user's interactive operation on the target element; generating an interactive analysis interface for the target page based on the tracking information; and sending the interactive analysis interface to the page management end so that the management object performs data analysis operations.

[0050] Among them, the front-end user's interactive operation on the target element refers to the user's interactive behavior with specific UI elements (such as buttons, links, forms, etc.) during the use of web pages or applications, such as clicking, sliding, filling, etc. The embedding information refers to the data about the operation collected by the system when the user performs the above operation, including but not limited to the time, location, type of operation and object of operation.

[0051] In actual applications, in response to the front-end user's interactive operations on the target element, the tracking information corresponding to the interactive operation is collected. When the user performs a specific operation on the page, the tracking code will be triggered and start collecting data related to this operation. For example, when the user clicks a button, the tracking code will record the timestamp of the click, the button ID, and other possibly related data, such as the user's unique identifier. This information will be encapsulated into a data packet and prepared for transmission to the back-end server.

[0052] Next, based on the collected tracking information, an interactive analysis interface for the target page is generated. This process is usually completed by the backend server. After receiving the tracking information sent by the frontend, the server will process and analyze the data and generate a visual report or chart. These reports or charts will show the user's interactive behavior patterns and help understand how users interact with the page. For example, statistics can be used to determine which buttons are clicked most frequently and the distribution of user stay time on the page.

[0053] Finally, the generated interactive analysis interface is sent to the page management end. The page management end here refers to the background management system responsible for monitoring and managing websites or applications, or a specific dashboard where administrators or analysts can view and analyze user interaction data. By sending the interactive analysis interface to the management end, management objects (such as product managers, data analysts, etc.) can perform data analysis operations. Management objects can use this data to optimize page design, improve user experience, discover potential problem areas, etc. For example, if a function button is found to have a low click rate, it may mean that the function is not easy to find or users are not interested in it, prompting the team to adjust the design of this part.

[0054] In a specific embodiment of the present invention, it is assumed that an online education platform wants to understand the user's behavior on the course details page. The system sets a tracking point on the target element of the page (such as the registration button). Whenever a user clicks the registration button, the system collects tracking point information, including click time, user ID, button ID, etc. This information is then sent to the back-end server, and the server generates a detailed analysis report based on this tracking point information, which includes a heat map of button clicks, the time distribution of user button clicks, etc. Finally, this analysis report is presented to the page management end, and the administrator can view this data, evaluate the performance of the page, and make improvement decisions accordingly.

[0055] Based on this, by responding to user interactions and collecting tracking information, the system can generate an interactive analysis interface, providing administrators with an intuitive data view. This not only helps to understand and optimize the user interface design, but also improves the overall user experience and enhances the attractiveness of the product. The system can effectively use this data to make continuous improvements and improve business efficiency.

[0056] Furthermore, an interactive analysis interface for the target page is generated based on the buried point information, including: inputting the buried point information into a pre-trained classification model to obtain the interaction type of the buried point information; and generating an interactive analysis interface for the target page based on the interaction type.

[0057] In practical applications, the tracking information is input into a pre-trained classification model to obtain the interaction type of the tracking information. This means that the tracking information collected by the front end must first be organized into a format acceptable to the model, usually a structured data set. For example, if the tracking information contains information such as the timestamp of the user clicking a button, the button ID, and the user ID, then this information will be organized into a data row, with each column representing a different type of data. Next, this data is input into a pre-trained classification model. The model may be a supervised learning model that has been trained with a large amount of historical data and has learned how to determine the type of user interaction based on the input data features. For example, the model may distinguish between multiple interaction types such as "browse", "click", and "purchase".

[0058] Generate an interactive analysis interface for the target page based on the interaction type. Once the interaction type of the buried point information is obtained, the corresponding analysis interface can be generated based on these types. These interfaces usually display data in a graphical way, for example, a bar chart showing the number of different types of interactions, a heat map showing the frequency of clicks on each element on the page, etc. In addition, the analysis interface may also include tabular data, listing the specific information of each interaction in detail. The system can customize different visualization components for each interaction type to more clearly present the distribution of various interactive behaviors. For example, for click events, the system can display a click heat map that highlights the most frequently clicked areas on the page; for input operations, it can display frequency statistics of input content.

[0059] In a specific embodiment of the present invention, it is assumed that the user behavior on the product details page of an e-commerce platform needs to be analyzed. The system collects a large amount of buried point information, including operations such as users clicking on product images, adding to shopping carts, and submitting orders. These buried point information are organized into a structured data format and input into a pre-trained classification model. The model recognizes multiple interaction types such as "browsing product details", "adding to shopping cart", and "submitting order". Subsequently, based on these interaction types, the system generates an interaction analysis interface. The interface includes a heat map that shows the frequency of product images being clicked; a bar chart that shows the comparison of the number of operations of adding to the shopping cart and submitting an order; and a table that lists in detail the specific time and user ID of each interaction.

[0060] Based on this, by inputting the embedded point information into the pre-trained classification model and generating an interactive analysis interface based on the interaction type output by the model, the system can provide page managers with a comprehensive perspective to understand user behavior. These analysis results can not only help identify user preferences, but also discover potential problem areas, thereby optimizing product design and improving user experience. The system can make more accurate data-driven decisions based on these analyses.

[0061] Furthermore, an interactive analysis interface for the target page is generated based on the interaction type, including: classifying the buried point information according to the interaction type to obtain a classification processing result; and generating an interactive analysis interface for the target page based on the classification processing result.

[0062] In actual applications, the buried point information is classified according to the interaction type to obtain the classification processing result. This step first requires that the collected buried point information be classified according to its corresponding interaction type. For example, all click events are classified into one category, and form submission events are classified into another category. The system can create multiple categories, each corresponding to a specific interaction type. During the classification process, the system will check each buried point information, determine the interaction type it belongs to, and then classify it into the corresponding category. For example, for a click event, the system will check whether the event type field in the buried point information is "click". If so, the information will be classified into the "click event" category. Such classification processing can help organize and analyze different types of operation data more clearly.

[0063] Generate an interactive analysis interface for the target page based on the classification processing results. After the classification processing is completed, the system will generate an interactive analysis interface based on these classification results. These interfaces are usually displayed in a table format, with each row representing a specific user operation record and each column representing different attributes of the operation, such as operation time, operation type, operation object, etc. For example, you can create a table to display all click events, which contains information such as the time of the click, the ID of the clicked element, and the user ID. This tabular display allows page managers to quickly browse and understand user interaction behaviors. In addition to the basic table display, filtering and sorting functions can be added as needed, allowing managers to view data based on specific conditions (such as time range, operation type, etc.).

[0064] In a specific embodiment of the present invention, it is assumed that the user behavior on the product details page of an online shopping website needs to be analyzed. The system collects a large amount of buried point information, including operations such as users clicking the "Add to Cart" button, scrolling the page, filling in comments, etc. The system first classifies and processes these operations according to the types of operations. For example, all operations of clicking the "Add to Cart" button are classified as the "Add to Cart" category, and all operations of scrolling the page are classified as the "Page Scrolling" category. Then, the system generates an interactive analysis interface based on these classification processing results. A table is displayed on the interface, where each row represents an "Add to Cart" operation, and the columns contain information such as operation time, user ID, and ID of the button clicked. In addition, there is another table showing all "Page Scrolling" operations, and the columns contain information such as scrolling time and scrolling distance.

[0065] Based on this, by classifying the embedded information and displaying the classification results in a table, the system can help page managers understand user behavior patterns more intuitively. These analysis results can be used to optimize page design, improve user experience, and discover possible problems. The system can effectively use this classification information to make continuous improvements and improve business efficiency.

[0066] Furthermore, an interaction analysis interface for the target page is generated based on the interaction type, including: determining the first number of interactions of each target element in the target page based on the embedding point information; generating a first interaction list based on the first number of interactions and the interaction type; and generating an interaction analysis interface for the target page based on the first interaction list.

[0067] In actual applications, the first number of interactions of each target element in the target page is determined based on the tracking information. This step first requires extracting all user operation records related to the target page from the tracking information. The system will identify each time a user interacts with an element on the page and count them according to the type of interaction (such as clicking, scrolling, filling out a form, etc.). For example, if the target page contains an "Add to Cart" button, the system will count the number of times the button is clicked. For each target element, the system will calculate the total number of its specific interaction type. For example, the system will count the number of times all buttons on the page are clicked and assign a unique identifier to each button for subsequent analysis.

[0068] Next, a first interaction list is generated based on the first number of interactions and the interaction type. In this step, the system classifies the counted number of interactions according to the interaction type and creates a list to record this information. For example, each item in the list can be represented as a key-value pair, where the key is the identifier of the target element (such as a button ID) and the value is the number of interactions of the element. Such a list can clearly show the interaction status of each element on the page. For example, the list may have an entry like this: "button-add-to-cart:120", which means that the "Add to Cart" button was clicked 120 times. The system can generate a separate list for each interaction type, or merge the data of all interaction types into a comprehensive list.

[0069] Then, an interactive analysis interface for the target page is generated based on the first interaction list. In this step, the system will design and generate an interactive analysis interface based on the first interaction list generated previously. This interface is usually presented in the form of tables, charts, etc., so that the page manager can intuitively see the number of interactions of different elements. For example, a bar chart can be used to show the number of times each button is clicked, or a table can be used to list the number of times each form field is filled in. The analysis interface can also provide filtering and sorting functions, allowing managers to view data within a specific time period as needed, or arrange elements from high to low according to the number of interactions. In addition, the interface can also include some advanced functions, such as trend analysis, comparison of data changes in different time periods, etc.

[0070] In a specific embodiment of the present invention, see Figure 3 As shown, Figure 3 It is a schematic diagram of an interactive analysis interface provided by an embodiment of the present invention. The interactive analysis interface is a page click analysis page, and the analysis object is the login page of the xxx webpage. The time period for collecting buried points is from 18:00:00 on July 8, 2024 to 20:00:00 on July 8, 2024. The click ranking list (first interactive list) ranks 5 types of click buttons by the number of clicks: ① Login button; ② Check the protocol box; ③ Password input box; ④ Other login methods; ⑤ Account input box. ① Login button and ② Check the protocol box show specific buried point information, including: ① Login button: buried point information: 10001 Click the login button, 10001 is the identifier of the click operation; Number of clicks: 10000; Average click rate: 100%; Top level: 1. ② Check the protocol box: buried point information: 10001 The protocol box is checked, 10001 is the identifier of the click operation; Number of clicks: 9700; Average click rate: 97%; Top level: 2.

[0071] Based on this, by statistically analyzing the embedded information in the target page and generating a list of interaction times, the system can provide page managers with a clear view to understand user behavior patterns. These analysis results can be used to optimize page design, improve user experience, and help identify possible problems. The system can effectively use these analysis results to make continuous improvements and improve business efficiency.

[0072] Furthermore, an interactive analysis interface for the target page is generated based on the buried point information, including: determining the interactive heat of each target element in the target page based on the buried point information; determining the heat level of each target element based on the interactive heat; and performing heat effect processing on each target element in the target page based on the heat level to obtain an interactive analysis interface for the target page.

[0073] Interaction heat refers to the frequency of user interaction with each element on the page within a certain period of time, usually measured by the number of interactions. Heat level is the result of quantitative grading of interaction heat, which is used to distinguish the degree of user interaction activity between different elements. Heat effect processing refers to presenting the interaction heat of each element in a visual way on the page according to the heat level of the element.

[0074] In actual applications, based on the tracking information, the interaction heat of each target element in the target page is determined. This step first requires extracting the user operation records related to the target page from the tracking information. The system will identify each time a user interacts with an element on the page and count the number of these interactions. For example, if the target page contains an "Add to Cart" button and a "Buy Now" button, the system will count the number of times these two buttons are clicked. For each target element, the system will calculate the total number of its specific interaction type to determine its interaction heat. For example, the "Add to Cart" button was clicked 120 times, and the "Buy Now" button was clicked 80 times.

[0075] Next, determine the heat level of each target element based on the interaction heat. In this step, the system will classify the number of interactions counted according to certain standards. For example, all elements can be divided into three heat levels: high, medium, and low. The top 20% of elements with the most interactions are marked as "high heat", the next 30% are marked as "medium heat", and the rest are marked as "low heat". The purpose of this is to more intuitively distinguish which elements on the page are most popular with users and which are less used. For example, the "Add to Cart" button may be marked as "high heat" due to the high number of clicks, while the "Buy Now" button may be marked as "medium heat".

[0076] Then, based on the heat level, each target element in the target page is processed for heat effect, and an interactive analysis interface for the target page is obtained. In this step, the system will design and generate an interactive analysis interface based on the previously determined heat level. This interface is usually presented in a visual way so that the page manager can intuitively see the interactive heat of different elements. For example, different colors can be used to represent different heat levels. In the interactive analysis page, a button with higher heat will appear darker red to indicate its higher popularity, while a button with lower heat will appear lighter. In addition, the interface can also contain other auxiliary information, such as a numeric label to show the specific number of clicks, or to display detailed data information by hovering the mouse.

[0077] In a specific embodiment of the present invention, it is assumed that a product details page of an e-commerce website needs to be analyzed. The system collects a large amount of buried point information, including operations such as users clicking the "Add to Cart" button, the "Buy Now" button, and scrolling the page. The system first counted that the "Add to Cart" button was clicked 120 times and the "Buy Now" button was clicked 80 times. Then, based on these statistical results, it was determined that the heat level of the "Add to Cart" button was "high" and the heat level of the "Buy Now" button was "medium". Finally, the system processed the heat effect of these two buttons in the interactive analysis page. The "Add to Cart" button is darker red due to its high click heat; while the "Buy Now" button is lighter red. In this way, page managers can see at a glance which elements are the most popular and which may need further optimization to improve user experience.

[0078] Based on this, by statistically analyzing the embedded information in the target page and visualizing each target element based on the heat level, the system can provide page managers with a clear view to understand user behavior patterns. These analysis results can be used to optimize page design, improve user experience, and help identify possible problems. The system can effectively use these analysis results to make continuous improvements and improve business efficiency.

[0079] Furthermore, an interaction analysis interface for the target page is generated based on the tracking information, including: based on the tracking information, analyzing the association relationship between multiple interaction operations performed by the front-end user on the target page; based on the association relationship, determining at least one interaction process among the multiple interaction operations, wherein the interaction process includes at least two interaction operations; determining a second interaction number of the interaction process, and generating a second interaction list based on the second interaction number and the interaction process; and generating an interaction analysis interface for the target page based on the second interaction list.

[0080] Among them, the association relationship of multiple interactive operations performed by front-end users on the target page refers to the logical connection or sequence between user operations. These association relationships reveal the user's behavior pattern when using the page. The interaction process refers to a series of related user operations that occur in a certain order and work together to complete a task or achieve a certain purpose. The second interaction count refers to the number of times a specific interaction process occurs within a certain period of time. The second interaction list is a list compiled based on the interaction process and its number of occurrences, which is used to generate the interaction analysis interface later.

[0081] In actual applications, based on the tracking information, the correlation between multiple interactive operations performed by front-end users on the target page is analyzed. This means first extracting all the user's operation records on the target page from the tracking information and analyzing the relationship between these operations. For example, the system will recognize that the user clicks the "Search" button, then fills in the search box, and then clicks a link in the search results. This analysis aims to find common patterns or paths between user operations. The system can use data mining techniques such as sequential pattern mining to identify these correlations, thereby revealing the typical behavior path of users on the page.

[0082] Next, at least one interaction process is determined among the multiple interaction operations based on the association relationship. In this step, the system will identify several interaction processes containing at least two operations based on the previously analyzed operation association relationship. For example, the system may identify an interaction process in which the user first clicks on a recommended product on the homepage, then browses the product details page, and then adds the product to the shopping cart. The system will record these processes and number or name them for subsequent reference.

[0083] Then, the second interaction count of the interaction process is determined, and a second interaction list is generated based on the second interaction count and the interaction process. In this step, the system will count the number of times each identified interaction process occurs. For example, the system will count how many times the above-mentioned "click on the recommended product -> browse the product details page -> add to the shopping cart" process appears. The system summarizes these processes and the number of times they occur into a list, namely the second interaction list. Each item in the list contains a description of an interaction process and the number of times the process occurs.

[0084] Finally, an interactive analysis interface for the target page is generated based on the second interactive list. In this step, the system will design and generate an interactive analysis interface based on the second interactive list. This interface is usually presented in the form of tables, charts, etc., so that page managers can intuitively see the frequency of occurrence of different interactive processes. For example, a bar chart can be used to show the number of occurrences of each process, or a flowchart can be used to show the path of user operations and their frequency. The analysis interface can also provide filtering and sorting functions, allowing managers to view data within a specific time period as needed, or arrange processes from high to low according to the number of interactions. In addition, the interface can also include some advanced functions, such as trend analysis, comparison of data changes in different time periods, etc.

[0085] In a specific embodiment of the present invention, it is assumed that the product details page of an online shopping website needs to be analyzed. The system collects a large amount of buried point information, including operations such as users clicking on "recommended products", browsing "product details page", and "adding products to shopping cart". The system first analyzes the correlation between these operations and identifies a typical interaction process: "click on recommended products->browse product details page->add to shopping cart". Then, the system counts the number of times this process occurs, assuming it is 120 times. Finally, the system generates an interaction analysis interface based on this information. A bar chart is displayed on the interface, showing the number of occurrences of different interaction processes; there is also a flow chart that clearly depicts the path of user operations and their frequency.

[0086] In another specific embodiment of the present invention, see Figure 4 As shown, Figure 4 It is another interactive analysis interface schematic diagram provided by an embodiment of the present invention. The interactive analysis interface is a process analysis page, the analysis object is the login page of the xxx webpage, the embedding point collection time period is from 18:00:00 on July 8, 2024 to 20:00:00 on July 8, 2024, and the process ranking list ranks 5 interactive processes by the number of clicks: ① process 1; ② process 2; ③ process 3; ④ process 4; ⑤ process 5. When the user clicks on a process in the process ranking list, the corresponding interactive process will be played in the slide show area on the left.

[0087] Based on this, by performing correlation analysis on the user's interactive operations on the target page and generating a list of the interaction process and the number of times it occurs, the system can provide page managers with a clear view to understand the user's behavior patterns. These analysis results can be used to optimize page design, improve user experience, and help identify possible problems. The system can effectively use these analysis results to make continuous improvements and improve business efficiency.

[0088] Furthermore, after configuring the page data for tracking based on the operation type and operation content, it also includes: receiving feedback information from front-end users regarding the page data, wherein the feedback information is used to characterize the existence of tracking errors in the page data; and training a page analysis model based on the feedback information until a preset stop condition is reached.

[0089] Feedback information refers to the feedback provided by front-end users on page data when using the page, especially information indicating tracking errors in the page. Tracking errors may include but are not limited to failure of tracking code to trigger correctly, incomplete or inaccurate collected data, etc. The page analysis model is a machine learning model used to analyze and optimize tracking data. It can be continuously iterated and improved based on user feedback information to improve the accuracy and effectiveness of tracking.

[0090] In actual applications, we receive feedback from front-end users on page data. This means that we need to provide a mechanism on the front-end page so that users can easily report problems or abnormal situations they encounter. For example, we can add a feedback button or link on the page, and users can describe the problem they encounter after clicking it, such as a button click not triggering the expected data collection. These feedback messages usually include a description of the specific scenario when the user encounters the problem, possible operation steps, and any detailed information that can help locate the problem.

[0091] Next, the page analysis model is trained based on the feedback information. In this step, the system uses the collected feedback information as training data to improve the existing page analysis model. Specifically, the system processes the feedback information and extracts features related to the tracking errors, such as error type, frequency of occurrence, scope of impact, etc. These feature data are then used to train or update the page analysis model. The model training process may involve supervised learning methods, in which feedback information is used as part of a labeled data set to help the model learn to identify which tracking configurations may cause errors and learn how to optimize these configurations. For example, if multiple feedback indicates that a particular tracking code fails to trigger correctly in certain situations, the model will learn this and try to propose improvement measures, such as adjusting the code logic or adding fault tolerance mechanisms.

[0092] The training process will continue until the preset stopping condition is reached. The preset stopping condition can be various, for example, when the performance of the model no longer improves significantly, or when the predetermined number of training rounds is reached, or when the amount of feedback information drops to a low level, indicating that the buried errors have been effectively resolved. The setting of these conditions depends on the specific application scenario and technical requirements, and aims to ensure that the model can continuously absorb new knowledge while avoiding overfitting.

[0093] In a specific embodiment of the present invention, it is assumed that multiple tracking points are set on the homepage of an e-commerce website to track user behavior. However, some users reported that when they clicked on certain recommended products, the system did not record the corresponding click events. After receiving these feedback messages, the system began to train the page analysis model based on these feedbacks. By analyzing the specific scenario descriptions in the feedback information, such as the specific time when the user clicked on the recommended product and the location of the click, the model identified that the cause of the tracking point error may be that the tracking point code failed to correctly bind to the corresponding DOM element. The system repeatedly trains the model and adjusts the tracking point code according to the model's suggestions until the preset stop condition is met, that is, the tracking point errors reported by users are significantly reduced or completely eliminated.

[0094] Based on this, by receiving feedback from front-end users and using it to train the page analysis model, the system can continuously optimize the tracking configuration and reduce or even eliminate tracking errors. This approach not only improves the quality of tracking data, but also provides users with a better experience. Based on these optimizations, the system can continuously improve its data collection mechanism to ensure that the collected data is accurate and reliable.

[0095] Furthermore, after configuring the page data for tracking based on the operation type and operation content, it also includes: for the target page, when abnormal information is detected, obtaining abnormal tracking information corresponding to the abnormal information; and feeding back the abnormal tracking information to the page management end so that the management object performs exception handling operations.

[0096] Among them, abnormal information refers to the data or behavior that does not meet expectations detected by the tracking system during the interaction between the user and the page. Abnormal tracking information refers to the tracking data related to these abnormal situations, including but not limited to the failure of the tracking code to trigger correctly, incomplete or erroneous data collection, and user behavior that does not meet expectations.

[0097] In actual applications, for the target page, when abnormal information is detected, the abnormal tracking point information corresponding to the abnormal information is obtained. This means that the system needs to have the ability to monitor the tracking point data in real time, so as to promptly discover and record any unexpected situations. For example, if the system expects that the corresponding tracking point event should be triggered when the user clicks a button, but in fact no event is triggered or the wrong event is triggered, this constitutes abnormal information. The system will automatically record these abnormal situations, including the time when the abnormality occurred, the specific elements involved, the user's operation type, and so on. For example, the system may record the following information: "At 17:00 on September 7, 2024, the user clicked the button with ID 'add-to-cart', but the tracking point code was not successfully triggered." Next, the abnormal tracking information is fed back to the page management end so that the management object can perform exception handling operations. In this step, the system sends the collected abnormal tracking information to the page management end, which is usually a backend management system or a dedicated data analysis platform. The page management end can be a web interface, a mobile application, or other forms of management tools. Its main function is to help managers view, analyze, and process these abnormal information. When the abnormal tracking information reaches the management end, the management object (such as a product manager, developer, or data analyst) can view this information and take appropriate processing measures. For example, the management object may check the specific cause of the exception, check whether there is a problem with the tracking code, or confirm whether there are new user behavior patterns that need to be taken into consideration. The management object can also adjust the tracking configuration, fix known problems, and optimize the data collection strategy through the tools provided by the management end.

[0098] In a specific embodiment of the present invention, it is assumed that multiple tracking points are set on the homepage of an e-commerce website to track user behavior. The system detects that when the user clicks the "Buy Now" button, the tracking point code fails to trigger correctly. The system records this exception, including the time of occurrence, user ID, button ID and other information. Then, the system sends these abnormal tracking point information to the page management end. After the page management end receives the information, the management object (such as the developer) can view the exception details through the management interface and perform exception handling operations. For example, the developer can check whether there are logical errors in the tracking point code, or whether the tracking point fails due to certain environmental factors (such as browser compatibility issues). The developer can fix the code problem and update the tracking point configuration on the management end to prevent similar exceptions from occurring again in the future.

[0099] Based on this, by monitoring the target page in real time and obtaining and feeding back abnormal tracking information when abnormal information is detected, the system can help page managers find and handle problems in a timely manner. This not only improves the accuracy of data collection, but also improves user experience. The system can effectively use this feedback information for continuous improvement to ensure the stability and reliability of the tracking system.

[0100] Furthermore, for the target page, when abnormal information is detected, abnormal point information corresponding to the abnormal information is obtained, including: for the target page, when abnormal information is detected, querying the log record corresponding to the abnormal information; based on the log record, determining the abnormal position corresponding to the abnormal information in the page data; collecting the abnormal point information corresponding to the abnormal position within a preset time period.

[0101] Among them, abnormal information refers to the data or behavior that does not meet expectations detected by the tracking system during the user's interaction with the page. Logging refers to the detailed records generated by the system during operation, including user operations, system responses, and other information that helps diagnose problems. The abnormal location refers to the specific location or element where the abnormality occurs in the page data, and the abnormal tracking information refers to the tracking data related to these abnormal situations.

[0102] In actual applications, it is necessary to query log records related to exception information. These log records include detailed information generated by the system when an exception occurs, which helps to identify the specific circumstances of the exception. According to the log records, locate the specific location of the exception information in the page data. This step requires analyzing the information in the log to determine the page element or operation link where the exception occurred. After determining the location of the exception, collect all the exception buried point information related to the exception location within the preset time period. This information includes all user operations and event data related to the exception location recorded within a specific time period, which helps to further analyze and locate the root cause of the problem.

[0103] Based on this, by monitoring the target page in real time and querying the log records when abnormal information is detected, the system can accurately locate the abnormal location and collect abnormal tracking information within the relevant period. This not only helps to quickly diagnose and solve problems, but also improves the accuracy of data collection and the stability of the system. The system can effectively use this information for continuous improvement to ensure the reliability of the tracking system and optimize the user experience.

[0104] Furthermore, the exception burial point information is fed back to the page management end so that the management object performs exception handling operations, including: obtaining the exception time corresponding to the exception information, and determining the feedback order based on the exception time; associating the log record with the exception burial point information based on the feedback order and feeding it back to the page management end.

[0105] In practical applications, it is necessary to obtain the exception time corresponding to the occurrence of the exception information. This timestamp is the key information to identify the order of occurrence of the exception events, which helps to determine the feedback order later. Based on the obtained exception time, determine the feedback order of the exception point information. Generally, exceptions that occur earlier should be fed back first so that the management object can understand and handle the exception in chronological order. Next, each exception point information needs to be associated with the corresponding log record. This association ensures that the management object can access detailed log records at the same time when viewing the exception point information, so as to have a more comprehensive understanding of the context and specific reasons for the exception. Finally, according to the determined feedback order, the associated log records and exception point information are fed back to the page management end. In this way, the management object can receive this information through the page management end and perform corresponding exception handling operations, such as analyzing the cause of the exception, repairing page errors, or adjusting the point configuration.

[0106] Based on this, it can ensure that the abnormal tracking point information is fed back in a timely, orderly and comprehensive manner, thereby improving the efficiency of page management and the accuracy of exception handling.

[0107] An embodiment of the present invention realizes receiving a point-burying processing request for a target page, wherein the point-burying processing request includes page data of the target page; inputting the page data into a pre-trained page analysis model to obtain the point-burying processing data of the target page, wherein the point-burying processing data includes a point-burying identifier and operation information corresponding to a target element in the target page, the point-burying identifier is used to characterize the operation type for the target element, and the operation information is used to characterize the operation content for the target element; and configuring the page data for point-burying based on the operation type and the operation content. By automatically processing page analysis requests, using a pre-trained model to parse page data and generate point-burying data, and adding corresponding point-burying identifiers and information to each element in the page, the workload of manual configuration is reduced, the point-burying deployment speed is accelerated, and the error rate of manual deployment is reduced.

[0108] The following combination Figure 5 , taking the application of the page analysis method provided by the present invention in the game login page as an example, the page analysis method is further described. Figure 5 It is a process flow chart of a page analysis method provided by an embodiment of the present invention, which specifically includes a development stage: steps 502 to 508, a user stage: steps 510 to 512, and an analysis stage: steps 514 to 518.

[0109] Step 502: Determine tracking point rules according to business needs.

[0110] Specifically, it is necessary to clarify which user behaviors and operations are valuable to the business, such as page browsing, click events, form submissions, etc. Next, define the specific tracking locations and trigger conditions, and formulate data collection standards and formats. This step usually involves close cooperation between business analysts and product managers to ensure that the tracking rules can accurately reflect business needs and user behaviors, laying the foundation for subsequent data collection and analysis.

[0111] Step 504: Collect training data.

[0112] Specifically, raw data is obtained from actual user interactions, including user behavior logs, operation history, and other relevant information. This data is used to train and optimize the model so that it can accurately identify and analyze target user behavior. Generally, the data collection process needs to ensure the diversity and representativeness of the data so that the model can perform well in different scenarios. At the same time, data privacy and security regulations must be followed to protect user information.

[0113] Step 506: Program development and tracking point development.

[0114] Specifically, we use supervised learning technology to train a model to accurately identify page elements and generate a unique tracking ID and its corresponding reporting information for each element. We use deep learning frameworks such as TensorFlow to optimize model performance. In addition, we apply natural language processing technology, including the use of NLP libraries such as NLTK, to automatically generate tracking documents.

[0115] We developed an automatic injection tool that can deeply analyze the front-end code. Based on the output of the AI ​​model, it can automatically insert the embedding code at the appropriate location in the code, including assigning IDs and corresponding reporting information. To achieve this function, we used technologies such as static code analysis and AST (abstract syntax tree) operations.

[0116] Step 508: Application deployment.

[0117] Specifically, the developed models and tools are deployed to the game program. The developed program is put online in the production environment. This includes deploying the front-end and back-end code to the server, configuring related services and databases, and ensuring that all tracking codes and data collection mechanisms can work properly.

[0118] Step 510: User interaction generates interaction data.

[0119] Specifically, user behavior data such as clicks, slides, and browsing will be captured and organized. This step needs to ensure that the data can accurately reflect the actual operations of users, and perform appropriate data preprocessing, such as data cleaning and format conversion, for subsequent analysis and use. At the same time, the real-time and accuracy of data collection needs to be monitored to ensure the integrity and validity of the interaction data.

[0120] Step 512: reporting interaction data.

[0121] Specifically, the front-end code sends the user's interactive data to the server through an HTTP request. This usually uses a POST request to package the data and send it to the preset API interface. This process needs to ensure the security and stability of data transmission, including the use of HTTPS encryption protocol to protect data privacy. At the same time, it is necessary to handle data loss or delay caused by network fluctuations to ensure that data can be uploaded to the server in a timely and complete manner.

[0122] Step 514: The backend receives the interaction data.

[0123] Step 516: Get the front-end code.

[0124] Specifically, get the latest front-end code from the version control system or code library. This step ensures that all tracking code and data collection logic have been integrated into the latest application version. The front-end code needs to be tested and verified to confirm that the tracking function works properly and the code is compatible with the back-end system. After obtaining the front-end code, it needs to be deployed and launched to ensure that user interaction data can be correctly collected and uploaded.

[0125] Step 518: Generate a visual analysis page.

[0126] Specifically, according to the interaction data and front-end code, the game login page is configured with tracking points. For each generated tracking point, the trained AI model is used to determine its type, such as page browsing, button clicks, form submissions, etc. In order to improve the accuracy of judgment, we use machine learning classification models, such as SVM or random forest. In actual applications, the tool will collect user feedback and data. Through online learning technology and feedback loop mechanism, the tool can continuously learn and correct errors, thereby improving the accuracy of the model and enhancing the robustness of the tool. This mechanism ensures that the system can continue to optimize in a changing business environment and provide more accurate services.

[0127] On the basis of complete reported data, the analysis page reconstructs the image and adds annotations, which are then displayed on the interface through graphics and text. For the front end, a layer of intuitive screen images is constructed to accurately annotate each reported position, and distinguish them by quantity and color blocks, so as to clearly show the quantity level and click priority. After successfully realizing the comprehensive visualization of page behavior, the page can directly support in-depth analysis of user usage habits of applications, games or programs in various fields, as well as detailed research on user behavior patterns. In addition, the platform can also assist developers in gradually analyzing unpopular functions or functions that are too deeply hidden in the page to optimize the user experience. For the back end, it can store various types of reported data transmitted by the client (i.e., the user end), and has data classification and processing functions to ensure the accuracy of the data and the precision of processing. Based on these real and reliable user behavior data, the front-end page of the tool provides information support for page annotation through intuitive data display and graphical analysis. This design is intended to help developers understand user behavior patterns more comprehensively, thereby optimizing page layout and function design, and ultimately achieving a dual improvement in user experience and application value.

[0128] For example, by filtering pages and adding date tags, the target page can be accurately located. On this basis, the embedded data of page clicks is used to realize the visualization of the button clicks on each page, and the click rankings of each button are listed in detail, while the button location and click count are clearly marked. In addition, in order to meet different business needs, a combination solution with other scenarios is also provided, such as integrating various types of data to form a total ranking of click counts, thereby providing more comprehensive and in-depth data analysis support for data analysis. When the front end combines multiple visual analysis charts logically, it can be constructed into a process analysis page, which is designed to provide users with an intuitive display of process analysis operations. In addition, a ranking function from high to low by number of behaviors is also provided on the right side of the page. When clicking an entry in the ranking list, the display area on the left will present the click order and the corresponding screen changes in the form of a slide, providing a clear and coherent analysis of user click behavior. When submitting a report, only the specific location information of the player's click (another application of the cross coordinate axis) needs to be included, and then the regional distribution of the player's clicks can be accurately displayed on the visual analysis tool. For areas that are highly clicked, the system will use red to highlight their importance, while areas with low click frequency will be marked in gray for easy distinction. This feature is designed to assist UI design or usability analysis to ensure that button design is more reasonable and more in line with user habits.

[0129] In addition, the page analysis system also includes an error reporting and processing function, which includes: First, the error log reporting function captures the exception information in the front-end code and sends the exception information to the server through HTTP requests. When implementing this function, you can use XMLHttpRequest or Fetch API in JavaScript to send a POST request to submit the exception information to the server-side interface in the form of data payload. This method can transmit error information in the application in real time, allowing the development team to quickly locate and solve problems, thereby improving the stability of the application.

[0130] Tracking reporting involves inserting tracking code into the front-end code that needs to track user behavior, such as click events and page views. When these events occur, the system automatically sends the corresponding tracking data to the server. Similar to error log reporting, tracking data is also transmitted to the server through HTTP requests. Doing so not only helps track user behavior, but also provides a basis for subsequent data analysis.

[0131] On the server side, error log reception and tracking data reception are completed by providing a RESTful API interface. The server-side interface receives POST requests from the front-end and stores error logs and tracking data in the database or performs further processing and analysis. These interfaces usually use back-end frameworks such as Node.js / Express, Java / Spring, or Python / Django, etc., and cooperate with related libraries or modules to process HTTP POST requests and parse JSON data, thereby achieving effective reception and management of front-end data.

[0132] In order to effectively manage and display the reported data, the front-end production and display of the point-of-use log background management tool is particularly important. The front-end page is usually built using HTML, CSS, and JavaScript, and the displayed data can be presented through elements such as tables and charts. To simplify the development process, open source front-end frameworks such as React or Vue.js can be used. These pages obtain log data by sending RESTful API GET requests to the server, and use chart libraries such as D3.js or ECharts to visualize the data, such as user behavior distribution and visit trends. This visual display helps users understand the data more intuitively, supporting business decisions and performance optimization.

[0133] Applied to the processing of this page analysis, AI automatic injection technology has shown significant advantages in tracking point management. Compared with the traditional tracking point method, it greatly simplifies the tracking point process. The traditional method requires detailed definition of tracking point content and complex reporting rules, involving special document writing and coding implementation. AI automatic injection technology realizes the automatic injection of tracking points by accessing API call tools, without manual intervention, reducing tedious operations. This automated method not only reduces manual workload, but also improves development efficiency and accuracy, making tracking point updates for new functions more efficient and avoiding additional manual work.

[0134] Through artificial intelligence, a unique identifier (ID) and corresponding fields, reporting instructions are generated, and the tracking document is output as a reference. The AI ​​tool can automatically identify all buttons, page refreshes and jump operations on the front-end page, and perform tracking injection in an all-round way. After completion, the tracking data can be automatically reported to the corresponding platform for management to ensure the accuracy and completeness of the data. The iteration efficiency of AI technology is extremely high. After the function is added, no additional manual supplementation is required, ensuring the continued effectiveness of tracking and comprehensive coverage of data.

[0135] The refined adjustment of AI technology further improves the capture accuracy of tracking data. For example, when a refresh button is clicked, the system originally only records one click event, but AI technology can correct the underreporting phenomenon and timely supplement the tracking of shortcut key usage. This self-correction function ensures the accuracy of tracking points at all shortcut key locations and comprehensively optimizes future connected projects, improving the comprehensiveness and accuracy of tracking data.

[0136] In addition, the tool also has a variety of practical functions. It integrates the tracking point record and log reporting information, and presents it on a visual interface in the form of a timeline, which is convenient for developers to directly observe the operation process and error information. The tool is equipped with a background warning and error viewing mechanism, which can accurately classify error information and provide severity warning functions. When the system detects frequent or high-frequency errors, it sends alarm information in a variety of ways to help developers quickly locate and handle problems. At the same time, the system supports detailed retrospective analysis of repeated errors, allowing users to select multiple reported data for comprehensive processing, ensuring that the problem is completely resolved.

[0137] Corresponding to the above method embodiment, the present invention also provides a page analysis device embodiment. Figure 6 FIG. 1 is a schematic diagram of a page analysis device provided by an embodiment of the present invention. Figure 6 As shown, the device includes: a receiving module 602, configured to receive a point-burying processing request for a target page, wherein the point-burying processing request includes page data of the target page; an input module 604, configured to input the page data into a pre-trained page analysis model to obtain point-burying processing data of the target page, wherein the point-burying processing data includes a point-burying identifier and operation information corresponding to a target element in the target page, the point-burying identifier is used to characterize the type of operation for the target element, and the operation information is used to characterize the content of the operation for the target element; a configuration module 606, configured to perform point-burying configuration on the page data based on the operation type and the operation content.

[0138] Optionally, the configuration module 606 is further configured to obtain the front-end code corresponding to the page data; generate the embedding code based on the operation type and the operation content; and incorporate the embedding code into the front-end code.

[0139] Optionally, the page analysis device also includes an interface generation module, which is configured to respond to the front-end user's interactive operations on the target elements, collect the buried point information corresponding to the interactive operations; generate an interactive analysis interface for the target page based on the buried point information; and send the interactive analysis interface to the page management end so that the management object performs data analysis operations.

[0140] Optionally, the interface generation module is further configured to input the buried point information into a pre-trained classification model to obtain the interaction type of the buried point information; and generate an interaction analysis interface for the target page based on the interaction type.

[0141] Optionally, the interface generation module is further configured to classify the buried point information according to the interaction type to obtain a classification processing result; and generate an interactive analysis interface for the target page based on the classification processing result.

[0142] Optionally, the interface generation module is further configured to determine the first number of interactions of each target element in the target page based on the tracking point information; generate a first interaction list based on the first number of interactions and the interaction type; and generate an interaction analysis interface for the target page based on the first interaction list.

[0143] Optionally, the interface generation module is further configured to determine the interaction heat of each target element in the target page based on the tracking point information; determine the heat level of each target element based on the interaction heat; based on the heat level, perform heat effect processing on each target element in the target page to obtain an interactive analysis interface for the target page.

[0144] Optionally, the interface generation module is further configured to analyze the association between multiple interactive operations performed by the front-end user on the target page based on the tracking point information; determine at least one interaction process among the multiple interaction operations based on the association relationship, wherein the interaction process includes at least two interaction operations; determine the second interaction number of the interaction process, and generate a second interaction list based on the second interaction number and the interaction process; and generate an interaction analysis interface for the target page based on the second interaction list.

[0145] Optionally, the page analysis device also includes a feedback retraining module, which is configured to receive feedback information from front-end users regarding page data, wherein the feedback information is used to characterize the presence of embedded point errors in the page data; and train the page analysis model based on the feedback information until a preset stop condition is met.

[0146] Optionally, the page analysis device also includes an exception reporting module, which is configured to obtain exception burial point information corresponding to the exception information for the target page when exception information is detected; and feed back the exception burial point information to the page management end so that the management object performs exception handling operations.

[0147] Optionally, the exception reporting module is further configured to query the log records corresponding to the abnormal information for the target page when abnormal information is detected; determine the abnormal position corresponding to the abnormal information in the page data based on the log records; and collect the abnormal point information corresponding to the abnormal position within a preset time period.

[0148] Optionally, the exception reporting module is further configured to obtain the exception time corresponding to the exception information, and determine the feedback order based on the exception time; based on the feedback order, the log record is associated with the exception burial point information and fed back to the page management end.

[0149] The page analysis device is applied to optimize the point-burying configuration process by automatically processing page analysis requests. The receiving module 602 is responsible for receiving the point-burying processing request for the target page and obtaining the page data. Then, the input module 604 inputs these page data into the pre-trained page analysis model to generate the point-burying processing data, which contains the point-burying identification and operation information of the target element. The identification is used to characterize the operation type, and the operation information is used to characterize the operation content. Finally, the configuration module 606 automatically configures the point-burying for the page data according to the operation type and operation content. This process effectively reduces the workload of manual configuration, speeds up the deployment of point-burying, and reduces the error rate of manual deployment. Through automation and model-driven processing, efficient and accurate point-burying management is achieved.

[0150] The above is a schematic scheme of a page analysis device of this embodiment. It should be noted that the technical scheme of the page analysis device and the technical scheme of the above-mentioned page analysis method belong to the same concept, and the details not described in detail in the technical scheme of the page analysis device can be referred to the description of the technical scheme of the above-mentioned page analysis method.

[0151] Figure 7 The block diagram of a computing device 700 according to an embodiment of the present invention is shown. The components of the computing device 700 include but are not limited to a memory 710 and a processor 720. The processor 720 is connected to the memory 710 via a bus 730, and the database 750 is used to store data.

[0152] The computing device 700 also includes an access device 740 that enables the computing device 700 to communicate via one or more networks 760. Examples of these networks include a public switched telephone network (PSTN), a local area network (LAN), a wide area network (WAN), a personal area network (PAN), or a combination of communication networks such as the Internet. The access device 740 may include one or more of any type of network interface (e.g., a network interface card (NIC)) of wired or wireless, such as an IEEE 802.11 wireless local area network (WLAN) wireless interface, a world-wide interoperability for microwave access (Wi-MAX) interface, an Ethernet interface, a universal serial bus (USB) interface, a cellular network interface, a Bluetooth interface, and a near field communication (NFC).

[0153] In one embodiment of the present invention, the above components of the computing device 700 and Figure 7 Other components not shown in the figure may also be connected to each other, for example, via a bus. It should be understood that Figure 7 The computing device structure block diagram shown is only for the purpose of illustration, and is not intended to limit the scope of the present invention. Those skilled in the art may add or replace other components as needed.

[0154] The computing device 700 may be any type of stationary or mobile computing device, including a mobile computer or mobile computing device (e.g., a tablet computer, a personal digital assistant, a laptop computer, a notebook computer, a netbook, etc.), a mobile phone (e.g., a smart phone), a wearable computing device (e.g., a smart watch, smart glasses, etc.), or other types of mobile devices, or a stationary computing device such as a desktop computer or a personal computer (PC). The computing device 700 may also be a mobile or stationary server.

[0155] The processor 720 is used to execute the following computer program / instructions, which implement the steps of the above-mentioned page analysis method when executed by the processor.

[0156] Each embodiment of the present invention is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the computing device embodiment, since it is basically similar to the page analysis method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the page analysis method embodiment.

[0157] An embodiment of the present invention further provides a computer-readable storage medium storing a computer program / instruction, which implements the steps of the above-mentioned page analysis method when executed by a processor.

[0158] Each embodiment of the present invention is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the computer-readable storage medium embodiment, since it is basically similar to the page analysis method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the page analysis method embodiment.

[0159] An embodiment of the present invention further provides a computer program product, including a computer program / instruction, which implements the steps of the above-mentioned page analysis method when executed by a processor.

[0160] The above is a schematic scheme of a computer program product of this embodiment. It should be noted that the technical scheme of the computer program product and the technical scheme of the above-mentioned page analysis method belong to the same concept, and the details not described in detail in the technical scheme of the computer program product can be referred to the description of the technical scheme of the above-mentioned page analysis method.

[0161] The above describes specific embodiments of the present invention. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0162] The computer instructions include computer program codes, which may be in source code form, object code form, executable files or some intermediate forms, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the content contained in the computer-readable medium may be appropriately increased or decreased according to the requirements of patent practice. For example, in some regions, according to patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0163] It should be noted that, for the convenience of description, the aforementioned method embodiments are all described as a series of action combinations, but those skilled in the art should be aware that the embodiments of the present invention are not limited by the described action sequence, because according to the embodiments of the present invention, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the embodiments of the present invention.

[0164] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0165] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The optional embodiments do not describe all the details in detail, nor do they limit the invention to the specific implementation methods described. Obviously, many modifications and changes can be made according to the content of the embodiments of the present invention. The present invention selects and specifically describes these embodiments in order to better explain the principles and practical applications of the embodiments of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A page analysis method, characterized in that: include: Receiving a tracking point processing request for a target page, wherein the tracking point processing request includes page data of the target page; Input the page data into a pre-trained page analysis model to obtain the embedded point processing data of the target page, wherein the embedded point processing data includes the embedded point identifier and operation information corresponding to the target element in the target page, the embedded point identifier is used to characterize the operation type for the target element, and the operation information is used to characterize the operation content for the target element; Based on the operation type and the operation content, the page data is configured with a buried point.

2. The method according to claim 1, characterized in that: The configuring the page data to be embedded based on the operation type and the operation content includes: Obtain the front-end code corresponding to the page data; Generate a tracking code based on the operation type and the operation content; Incorporate the embedding code into the front-end code.

3. The method according to claim 1, characterized in that After configuring the page data based on the operation type and the operation content, the method further includes: In response to the front-end user's interactive operation on the target element, collecting the tracking point information corresponding to the interactive operation; Generate an interactive analysis interface for the target page based on the embedding information; The interactive analysis interface is sent to the page management end so that the management object performs data analysis operations.

4. The method according to claim 3, characterized in that The generating an interactive analysis interface for the target page based on the embedding information includes: Inputting the buried point information into a pre-trained classification model to obtain the interaction type of the buried point information; An interaction analysis interface for the target page is generated based on the interaction type.

5. The method according to claim 4, characterized in that The generating an interaction analysis interface for the target page based on the interaction type includes: Classify the buried point information according to the interaction type to obtain a classification result; An interactive analysis interface for the target page is generated based on the classification processing result.

6. The method according to claim 4, characterized in that The generating an interaction analysis interface for the target page based on the interaction type includes: Based on the embedding information, determining a first interaction number of each target element in the target page; generating a first interaction list based on the first number of interactions and the interaction type; An interaction analysis interface for the target page is generated based on the first interaction list.

7. The method according to claim 3, characterized in that The generating an interactive analysis interface for the target page based on the embedding information includes: Based on the embedding information, determining the interaction heat of each target element in the target page; Determine the heat level of each target element based on the interaction heat; Based on the heat level, heat effect processing is performed on each target element in the target page to obtain an interactive analysis interface for the target page.

8. The method according to claim 3, characterized in that The generating an interactive analysis interface for the target page based on the embedding information includes: Based on the embedding information, analyzing the correlation relationship between multiple interactive operations performed by the front-end user on the target page; Determining at least one interaction process among the plurality of interaction operations based on the association relationship, wherein the interaction process includes at least two interaction operations; Determining a second number of interactions of the interaction process, and generating a second interaction list based on the second number of interactions and the interaction process; An interaction analysis interface for the target page is generated based on the second interaction list.

9. The method according to claim 1, characterized in that: After configuring the page data based on the operation type and the operation content, the method further includes: For the target page, when abnormal information is detected, query the log record corresponding to the abnormal information; Based on the log record, determining an abnormal position corresponding to the abnormal information in the page data; Collecting abnormal buried point information corresponding to the abnormal position within a preset time period; Acquire the abnormal time corresponding to the abnormal information, and determine the feedback order based on the abnormal time; Based on the feedback order, the log record is associated with the abnormal tracking point information and fed back to the page management end.

10. A page analysis device, characterized in that: include: A receiving module is configured to receive a burying point processing request for a target page, wherein the burying point processing request includes page data of the target page; An input module is configured to input the page data into a pre-trained page analysis model to obtain the embedded point processing data of the target page, wherein the embedded point processing data includes an embedded point identifier and operation information corresponding to a target element in the target page, the embedded point identifier is used to characterize the operation type for the target element, and the operation information is used to characterize the operation content for the target element; The configuration module is configured to perform point-burying configuration on the page data based on the operation type and the operation content.

11. A computing device, characterized in that: include: Memory and processor; The memory is used to store computer programs / instructions, and the processor is used to execute the computer programs / instructions. When the computer program / instructions are executed by the processor, the steps of the page analysis method according to any one of claims 1 to 9 are implemented.

12. A computer-readable storage medium, characterized in that: It stores a computer program / instruction, which, when executed by a processor, implements the steps of the page analysis method described in any one of claims 1-9.

13. A computer program product, characterized in that It includes a computer program / instruction, which, when executed by a processor, implements the steps of the page analysis method described in any one of claims 1 to 9.

Images