Product information processing method, system, computing device, and electronic device

Abstract

The application discloses a product information processing method and system, a computing device and an electronic device, and relates to the technical field of large models and product demand management. The method can include: obtaining a task information set of a product, wherein the task information in the task information set is used to describe a development task that needs to be performed in a development process of the product; determining target task information of the product from the task information set, wherein the target task information is used to describe a target development task that has not been completed in the development process of the product; obtaining an operation event set associated with the target task information, wherein an operation event in the operation event set is used to represent an operation behavior performed on the target task information; identifying at least one key operation event from the operation event set by using a target model; and determining development information of the product based on the key operation event. The application solves the technical problem of low information processing efficiency of the product.

Description

Technical Field

[0001] This application relates to the fields of large model technology and product demand management, and more specifically, to a product information processing method, system, computing device, and electronic device. Background Technology

[0002] Currently, with the rapid development of product engineering (e.g., software, projects) and the deepening of software engineering practices, the complexity and difficulty of software development (product research and development) are constantly increasing. Software development activities involve a large amount of manpower, processes, tools, and technologies, making product management particularly important. Product management involves planning, executing, tracking, and monitoring all aspects of software development activities to ensure that the product is completed on time and meets requirements. Therefore, how to improve the efficiency of requirements management and process tracking, and how to automate and intelligently manage requirements and process tracking, are urgent problems to be solved.

[0003] In related technologies, time tracking and reporting functions have become key tools for supporting product progress and team collaboration in software development and management. Time tracking systems typically focus on recording the total time spent on work items as an evaluation metric for product progress and team efficiency. However, time tracking tools only provide static time records and cannot capture dynamic behavioral patterns during the development process. Furthermore, relying on manual input of work items and timing not only increases the burden on users (e.g., developers) but also frequently leads to inaccuracies and omissions in information recording, severely weakening the reliability of statistical information and the basis for decision-making. In addition, manually managing large amounts of data significantly reduces information processing efficiency as team size increases or product complexity grows, affecting product progress and the timeliness of decisions. Therefore, the technical problem of low information processing efficiency in products remains.

[0004] There is currently no effective solution to the above problems. Summary of the Invention

[0005] This application provides an information processing method, system, computing device, and electronic device for a product, to at least solve the technical problem of low information processing efficiency in products.

[0006] According to one aspect of the embodiments of this application, a product information processing method is provided. The method may include: acquiring a task information set of the product, wherein the task information in the task information set describes the R&D tasks that need to be performed in the product's R&D process; determining target task information of the product from the task information set, wherein the target task information describes target R&D tasks that were not completed during the product's R&D process; acquiring a set of operation events associated with the target task information, wherein the operation events in the operation event set represent operation behaviors performed on the target task information; identifying at least one key operation event from the operation event set using a target model, wherein the importance of the key operation event to the target R&D task is greater than the importance of other operation events in the operation event set (excluding the key operation event) to the target R&D task, and the target model is obtained by training a large model; and determining R&D information of the product based on the key operation event, wherein the R&D information represents different statistical information generated by the product during the R&D process.

[0007] According to another aspect of the embodiments of this application, another product information processing method is provided, applied to a product development platform. The method may include: acquiring a task information set of a smart product, wherein the task information in the task information set describes the development tasks that need to be performed in the functional development process of the smart product; determining target task information of the smart product from the task information set, wherein the target task information describes target development tasks that have not been completed during the development process of the smart product; acquiring a set of operation events associated with the target task information, wherein the operation events in the operation event set represent operation behaviors performed on the target task information; identifying at least one key operation event from the operation event set using a target model, wherein the importance of the key operation event to the target development task is greater than the importance of other operation events in the operation event set (excluding the key operation event) to the target development task, and the target model is obtained by training a large model; and determining development information of the smart product based on the key operation event, wherein the development information represents different statistical information generated by the smart product in the functional development process.

[0008] According to another aspect of the embodiments of this application, another product information processing method is provided. The method may include: displaying a set of task information for the product on an operating interface, wherein the task information in the task information set describes the R&D tasks that need to be performed in the product's R&D process; and displaying product R&D information on the operating interface in response to an information processing instruction applied to the operating interface, wherein the R&D information represents different statistical information generated by the product in the R&D process, and at least one key operation event is determined based on the target task information associated with the target task information in the task information set. The target task information describes a target R&D task that has not been completed during the product's R&D process, and the operation events in the operation event set represent operational behaviors performed on the target task information. The importance of the key operation event to the target R&D task is greater than the importance of the operation events in the operation event set other than the key operation event to the target R&D task. The key operation event is identified from the operation event set using a target model, and the target model is obtained by training a large model.

[0009] According to another aspect of the embodiments of this application, a method for storing product information is provided. The method may include: acquiring a task information set of the product, wherein the task information in the task information set describes the R&D tasks that need to be performed in the product's R&D process; determining target task information of the product from the task information set, wherein the target task information describes target R&D tasks that have not been completed during the product's R&D process; querying a set of operation events associated with the target task information from a hierarchical data storage structure, wherein the operation events in the operation event set represent the operation behaviors performed on the target task information; identifying at least one key operation event from the operation event set using a target model, and storing the key operation event in a data storage structure at the corresponding level of the hierarchical data storage structure, wherein the importance of the key operation event to the target R&D task is greater than the importance of other operation events in the operation event set (excluding the key operation event) to the target R&D task, and the target model is obtained by training a large model; determining R&D information of the product based on the key operation event, and storing the R&D information in a data storage structure at the corresponding level of the hierarchical data storage structure, wherein the R&D information represents different statistical information generated by the product during the R&D process.

[0010] According to another aspect of this embodiment of the application, an information processing system for a product is provided. The system may include: a client for transmitting a set of task information for the product, wherein the task information in the task information set describes the R&D tasks that need to be performed in the product's R&D process; a server for determining target task information for the product from the task information set, wherein the target task information describes target R&D tasks that have not been completed during the product's R&D process; obtaining a set of operation events associated with the target task information, wherein the operation events in the operation event set represent operation behaviors performed on the target task information; identifying at least one key operation event from the operation event set using a target model, wherein the importance of the key operation event to the target R&D task is greater than the importance of other operation events in the operation event set (excluding the key operation event) to the target R&D task, and the target model is obtained by training a large model; determining R&D information for the product based on the key operation event, wherein the R&D information represents different statistical information generated by the product in the R&D process; and transmitting the R&D information to the client.

[0011] According to another aspect of the embodiments of this application, a computing device is also provided, including: a memory storing an executable program; and a processor for running the program, wherein the program executes the methods in various embodiments of this application when it runs.

[0012] According to another aspect of the embodiments of this application, an electronic device is also provided, including: a memory storing an executable program; and a processor connected to the memory via a bus for running the program, wherein the program executes the methods in various embodiments of this application when it runs.

[0013] According to another aspect of the embodiments of this application, a computer-readable storage medium is also provided, the computer-readable storage medium including a stored executable program, wherein, when the executable program is running, it controls the device where the computer-readable storage medium is located to perform the methods of various embodiments of this application.

[0014] According to another aspect of the embodiments of this application, a computer program product is also provided, including a computer program that, when executed by a processor, implements the methods of various embodiments of this application.

[0015] According to another aspect of the embodiments of this application, a computer program product is also provided, including a non-volatile computer-readable storage medium storing a computer program, which, when executed by a processor, implements the methods in various embodiments of this application.

[0016] According to another aspect of the embodiments of this application, a computer program is also provided, which, when executed by a processor, implements the methods of the various embodiments of this application.

[0017] In this embodiment, to determine the R&D information of a product, a task information set for the product can be obtained. This task information set allows analysis of the R&D tasks the product needs to perform in the R&D process, identifying target R&D tasks that have not yet been completed and determining the corresponding target task information. The operational behaviors required to complete the target R&D tasks can be determined, resulting in a set of operation events corresponding to the target task information. This set of operation events can be input into a target large-scale model, which analyzes each operation event in the set, identifying those with high importance to the target R&D tasks as key operation events. The product's R&D information can then be determined based on these key operation events. In this embodiment, an Intelligent Product Requirement and Process Tracking System (AIgile) is set up to provide a relatively complete solution. AIgile can not only quickly and accurately identify and manage R&D tasks through automated work item management, operation behavior recording, and intelligent analysis of large models, but also deeply analyze operation behavior, automatically extract key operation events, and generate rich statistical reports. This provides more efficient and intelligent decision support for product management, thereby achieving the technical effect of improving the information processing efficiency of products and solving the technical problem of low information processing efficiency of products.

[0018] It is worth noting that the general description above and the detailed description that follow are merely for illustrative purposes and do not constitute a limitation on this application. Attached Figure Description

[0019] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0020] Figure 1 This is a schematic diagram illustrating an application scenario of an information processing method for a product according to an embodiment of this application;

[0021] Figure 2 This is a flowchart of an information processing method for a product according to an embodiment of this application;

[0022] Figure 3 This is a flowchart of another product information processing method according to an embodiment of this application;

[0023] Figure 4 This is a flowchart of another product information processing method according to an embodiment of this application;

[0024] Figure 5 This is a flowchart of a product information storage method according to an embodiment of this application;

[0025] Figure 6 This is a schematic diagram of an information processing system for a product according to an embodiment of this application;

[0026] Figure 7 This is a flowchart of an AIgi le work item management and analysis process according to an embodiment of this application;

[0027] Figure 8 This is a flowchart illustrating an AIgi le work item list retrieval and display process according to an embodiment of this application;

[0028] Figure 9 This is a schematic diagram of an AIgi le user-initiated work item selection interface according to an embodiment of this application;

[0029] Figure 10 This is a schematic diagram of an AIgi le system event and user behavior triggering automatic creation and switching process of work items according to an embodiment of this application;

[0030] Figure 11 This is a flowchart of an AIgi le-based work item creation and switching decision process according to an embodiment of this application;

[0031] Figure 12(a) is a flowchart of an AIgi le user keyboard and mouse event capture and context information extraction process according to an embodiment of this application;

[0032] Figure 12(b) is a schematic diagram of the specific information content included in a click event information according to an embodiment of this application;

[0033] Figure 12(c) is a schematic diagram of the specific information content included in a keyboard event information according to an embodiment of this application;

[0034] Figure 13 This is a schematic diagram of an AIgi le MacOS notification center listening and notification event capture process according to an embodiment of this application;

[0035] Figure 14 This is a flowchart illustrating the AIgi le user event timeline construction and key event generation process according to an embodiment of this application;

[0036] Figure 15 This is a schematic diagram illustrating the process of identifying and recording key events in an AIgi le large model according to an embodiment of this application;

[0037] Figure 16This is a flowchart of a process for generating key events from a large model according to an embodiment of this application;

[0038] Figure 17 This is a schematic diagram of an AIgi le search event generation and index construction process according to an embodiment of this application;

[0039] Figure 18 This is a schematic diagram of an AIgi le code event generation and statistical analysis process according to an embodiment of this application;

[0040] Figure 19 This is a schematic diagram illustrating the distribution of AIgi le code development time and the proportion of Copi le interaction time according to an embodiment of this application;

[0041] Figure 20 This is a schematic diagram illustrating the code distribution and copy lot adoption rate analysis of an AIgi le code development activity according to an embodiment of this application;

[0042] Figure 21 This is a schematic diagram illustrating the distribution of AIgi le code submission files and the statistics of Copi lot acceptance rate according to an embodiment of this application;

[0043] Figure 22 This is a schematic diagram illustrating AIgi le code event generation and multi-dimensional statistics according to an embodiment of this application;

[0044] Figure 23 This is a schematic diagram illustrating an AIgi le user interaction event analysis and statistical process according to an embodiment of this application;

[0045] Figure 24 This is a flowchart of an AIgi le user interaction method with a large model according to an embodiment of this application;

[0046] Figure 25 This is a schematic diagram of an AIgi le user-large model interaction loop and event recording process according to an embodiment of this application;

[0047] Figure 26 This is a schematic diagram illustrating an AIgi le code writing and intelligent consultation event generation process according to an embodiment of this application;

[0048] Figure 27 This is a flowchart of an AIgi le-based automated update process for work items based on key events, according to an embodiment of this application.

[0049] Figure 28 This is a schematic diagram of a user event analysis based on hierarchical storage according to an embodiment of this application;

[0050] Figure 29 This is a flowchart of a model training method according to an embodiment of this application;

[0051] Figure 30 This is a flowchart illustrating a method for incremental training and continuous learning based on user feedback, according to an embodiment of this application.

[0052] Figure 31 This is a schematic diagram of an information processing device for a product according to an embodiment of this application;

[0053] Figure 32 This is a schematic diagram of an information processing apparatus for another product according to an embodiment of this application;

[0054] Figure 33 This is a schematic diagram of an information processing apparatus for another product according to an embodiment of this application;

[0055] Figure 34 This is a schematic diagram of an information storage device for a product according to an embodiment of this application;

[0056] Figure 35 This is a structural block diagram of a computing device according to an embodiment of this application;

[0057] Figure 36 This is a structural block diagram of an electronic device according to an embodiment of this application. Detailed Implementation

[0058] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present application.

[0059] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0060] The technical solution provided in this application is mainly implemented using large-scale model technology. Here, "large-scale model" refers to a deep learning model with a massive number of parameters, typically containing hundreds of millions, tens of billions, hundreds of billions, trillions, or even tens of trillions of parameters. Large-scale models can also be called foundational models. They are pre-trained using large-scale unlabeled corpora to produce pre-trained models with hundreds of millions of parameters. Such models can adapt to a wide range of downstream tasks and have good generalization ability. Examples include Large Language Models (LLMs) and multimodal pre-training models.

[0061] It should be noted that, in practical applications, large models can be fine-tuned using a small number of samples to adapt them to different tasks. For example, large models can be widely applied in Natural Language Processing (NLP), computer vision, and speech processing. Specifically, they can be applied to computer vision tasks such as Visual Question Answering (VQA), Image Captioning (IC), and Image Generation, as well as NLP tasks such as text-based sentiment classification, text summarization, and machine translation. Therefore, the main application scenarios for large models include, but are not limited to, digital assistants, intelligent robots, search, online education, office software, e-commerce, and intelligent design. In this embodiment, the use of a large model to identify high-importance key operational events from a set of operational events is explained.

[0062] First, some nouns or terms that appear in the description of the embodiments of this application shall be interpreted as follows:

[0063] AIgi le is a requirements tracking and analysis system based on Large Language Models (LLM). AIgi le can automatically record and aggregate user behavior during the product development process, update requirements in real time, summarize work content, and automatically generate multi-dimensional statistical reports.

[0064] In this application embodiment, "requirement / work item" specifically refers to product requirements. Product requirements refer to the specific requirements of users or the market for product functions, performance, interface, etc. These product requirements are usually derived from market research, user interviews, competitor analysis, and other channels, aiming to guide product design and development and ensure that the product can meet the needs and expectations of users.

[0065] Requirements tracking refers to the continuous monitoring and recording of the implementation status of requirements throughout the product development process. Requirements tracking ensures that each requirement is correctly understood and implemented, while also enabling timely identification and resolution of problems, thus guaranteeing product quality and schedule.

[0066] Requirements analysis is the process of in-depth research and analysis of product requirements. Its purpose is to clarify the specific content, scope, and priority of these requirements, providing accurate guidance for subsequent design and development. Requirements analysis typically includes steps such as requirements gathering, requirements assessment, and requirements specification.

[0067] Requirements flow, or the update of requirements status, refers to the transmission and processing mechanism of requirements throughout the entire process from their inception to implementation. In Agile development, requirements flow typically involves multiple stages such as requirements initiation, review, allocation, development, and testing.

[0068] Agile development is a user-centric, iterative, and incremental product development methodology that emphasizes rapid response to change and continuous value delivery.

[0069] Keyboard events are records of user actions on the keyboard, including key press and release events, used to capture user input behavior;

[0070] Mouse events are records of user actions on the mouse, including click, move, scroll, and other events, used to capture user interaction behavior;

[0071] Semantic groups group related elements and content based on the context information of user actions to better understand and analyze user behavior;

[0072] Application context refers to the application environment in which the user operates, including information such as the application's name, window, and page, which provides background information for the operation.

[0073] Large model context refers to the contextual information that the large model relies on when processing user events, including the list of work items, user events within a time window, etc., to generate accurate analysis results.

[0074] A prompt can be an instruction or hint provided to a large model to guide it in generating specific outputs, such as whether to switch to or create a new work item.

[0075] Business intelligence (BI) is the technology and application used to analyze and report data.

[0076] Tiered storage can be a data storage architecture that organizes and manages data through multiple layers, ensuring data integrity and efficient access.

[0077] The raw data layer (Operating Data Store, or ODS for short) is used to store raw, unprocessed data. In AIgile, the ODS layer is used to store raw user events;

[0078] The Data Warehouse Layer (DWD) is a hierarchy used to store processed and refined data. In AIGILE, the DWD layer stores key events extracted from raw user events.

[0079] The Data Warehouse Service Layer (DWS) provides data access interfaces and services to upper-layer applications. In AIgile, the DWS layer stores aggregated metrics built from various key events, supporting analysis and querying from multiple dimensions.

[0080] The Application Data Service Layer (ADS) is a collection of data created according to specific application needs. After further processing and analysis, it provides highly aggregated and customized data analysis and reports. In AIgile, the ADS layer is used to store statistical analysis information such as time distribution, work item progress, work reports, AI interaction history, and search history, and supports interactive queries.

[0081] According to an embodiment of this application, an information processing method for a product is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0082] Considering the large number of model parameters in large models and the limited computing resources of mobile terminals, the information processing method for the above-mentioned products provided in this application embodiment can be applied to, for example... Figure 1 The application scenarios shown are not limited to these. In, for example... Figure 1In the application scenario shown, the large model is deployed on server 30, which can be in the cloud. Server 30 can connect to one or more terminal devices 10 via LAN, WAN, the Internet, or other types of data networks. Terminal devices 10 can be client devices, including but not limited to smartphones, tablets, laptops, PDAs, personal computers, smart home devices, and in-vehicle devices. These client devices collectively constitute the client relative to the server. A graphical user interface (GUI) for acquiring product images can be deployed on the client device; this GUI can be an e-commerce platform interface. Terminal devices 10 can interact with users through the GUI to access the large model, thereby implementing the product information processing method provided in this embodiment. Server 30 and terminal devices 10 can exchange information via network 20. The database 40 in the server can contain the large model. In this embodiment, database 40 stores target models for identifying key operation times from a set of operation events. These models can be accessed through server 30 to identify key operation events of high importance to the target R&D task from the set of operation events.

[0083] In this embodiment, the system consisting of a terminal device, a network, and a server can perform the following steps: If a user needs to obtain R&D information about a product in the R&D process, they can perform input or selection operations on the operation interface of the corresponding terminal device 10 to upload the instructions of the product to be processed to the terminal device 10. The terminal device 10 can send the product selected by the user to the server 30 via the network 20. After receiving the product for which information processing is required, the server 30 can perform the following steps: Step S102, obtain the task information set of the product; Step S104, determine the target task information of the product from the task information set; Step S106, obtain the operation event set associated with the target task information; Step S108, identify at least one key operation event from the operation event set using the target model; Step S110, determine the R&D information of the product based on the key operation event. The R&D information generated by the target model can be output to the terminal device 10 via the network 20 to display the R&D information on the operation interface of the terminal device for the user to view.

[0084] In this embodiment, AIgi le solves the problem of low product information processing efficiency in existing technologies by automating work item management, recording operational behavior, and intelligent analysis of large models. It can not only quickly and accurately identify and manage R&D tasks, but also deeply analyze operational behavior, automatically extract key operational events, and generate rich statistical reports, thereby providing more efficient and intelligent decision support for product management. This achieves the technical effect of improving product information processing efficiency and solves the technical problem of low product information processing efficiency.

[0085] It should be noted that with the rapid development of high-performance computing units, the methods provided in this application embodiment can also be applied to model-in-machine systems in other application scenarios. In one optional embodiment, the model-in-machine system has multiple built-in models, and users can select one model to adjust as needed to obtain their own model. The high-performance computing unit built into the model-in-machine system can then directly call the adjusted model to execute the methods provided in this application embodiment. In another optional embodiment, the large model-in-machine system has a pre-trained model built-in, and the high-performance computing unit built into the model-in-machine system can then directly call that model to execute the methods provided in this application embodiment.

[0086] Furthermore, when users need to train their own models, they can upload their own datasets via the client. These datasets are then sent to the server, allowing the server to adjust the pre-trained model using the dataset to obtain the user's customized model, which can then be deployed to the production environment. To facilitate users' model adjustment needs, the server provides complete adjustment tools, development frameworks, and processes, supporting multiple adjustment strategies. This allows the adjusted model to better adapt to different application domains and achieve a high degree of customization.

[0087] Under the aforementioned operating environment, this application provides the following: Figure 2 The method for processing product information shown is illustrated. It should be noted that the product information processing method in this embodiment can be [details omitted]. Figure 1 The mobile terminal in the illustrated embodiment is executed. Figure 2 This is a flowchart of a product information processing method according to an embodiment of this application, such as... Figure 2 As shown, the method may include the following steps:

[0088] Step S202: Obtain the product's task information set.

[0089] In the technical solution provided in step S202 of this application, the product can be a software development project, or it can be referred to as software or a project. The task information set corresponding to the product can be a collection of R&D tasks (tasks) that the product needs to perform in the development process, including but not limited to function development, performance optimization, interface design, code review, testing and deployment, etc., without specific limitations. The task information in the task information set can be used to describe the R&D tasks that need to be performed in the product's R&D process. The task information set can also be called a work item or a work item list, for example, it can be an AIgi le work item. The task information in the task information set can include the task description, status information, assigned personnel, priority, deadline, dependencies, progress records, etc. of the corresponding task (e.g., R&D task), without limitations. Optionally, the task description can be used to clarify the task's goals and requirements, including task requirements, performance indicators, and interface design, etc. The status information can be used to indicate the current status of the task, such as "not started", "in progress", and "completed". The priority can be used to indicate the urgency and importance of the task. The dependency relationship is used to indicate the dependency between the task and other tasks. Progress logs can be used to represent the progress and completion status of a task, such as time distribution and amount of code.

[0090] In this embodiment, if it is necessary to statistically analyze the R&D information in the product development process, the task information set of the product can be obtained.

[0091] Optionally, in the AIgi le system, the above embodiments involve automatically collecting and integrating work items related to the product's development process (R&D process) from different sources. This operation is crucial for subsequent automated requirements management, behavior tracking, and data analysis.

[0092] Optionally, the AIgi le system can pre-integrate multiple data sources, which may include project management tools (such as Trello, Aone), version control tools, communication and collaboration platforms, and code editors. AIgi le can then retrieve pull work items from these data sources.

[0093] Optionally, work items obtained from multiple data sources can be displayed as a work item list on the user's terminal device interface. For example, multiple work items can be numbered and displayed vertically on the interface in ascending order of number. The status of each work item can also be displayed, such as pending development, pending processing, under development, and released. The assignor, priority, and project to which each work item belongs can also be displayed.

[0094] It should be noted that the above-described form and specific content of displaying the task information set on the operation interface after obtaining the task information set are merely illustrative examples and are not subject to specific limitations. Any method that can display the task information set on the operation interface, making it convenient for users to select the target task information for generating the required R&D information, is within the protection scope of this application's embodiments.

[0095] In this embodiment, by integrating multiple data sources, the challenge of information being scattered across multiple tools and platforms can be overcome, enabling unified management and display of requirements and work item information related to product development. Real-time monitoring of user actions and automatic updates to the task information set reduce manual operations and improve the efficiency and accuracy of data collection and updates. Through real-time and structured task information sets, accurate and comprehensive requirements status and progress information can be obtained, enabling more scientific decision-making and adjustments to project processes and resource allocation.

[0096] Step S204: Determine the target task information of the product from the task information set.

[0097] In the technical solution provided in step S204 of this application, the target task information can be used to describe the target R&D tasks that have not been completed during the product development process. The target task information can be a current work item actively selected by the user, such as a work item to be developed or to be processed.

[0098] In this embodiment, after obtaining the product's task information set, the target task information corresponding to the target R&D tasks that were not completed during the product R&D process can be determined from the task information set.

[0099] Optionally, in the AIgi le system, this embodiment is a key screening and localization process, the core task of which is to determine the target task information corresponding to the currently incomplete target R&D task from the collected task information set. This process is crucial for subsequent automated requirements tracking, behavior analysis, and key event generation.

[0100] In this embodiment, there are two ways to determine the target task information from the task information set. One way is that the user can select the target task information according to their needs on the operation interface. The other way is that the AIgi le system can automatically create or select the target task information based on the task status corresponding to each task information in the task information set.

[0101] Optionally, after displaying the task information set on the user's interface, the status of each task in the task information set can be displayed, such as "Not Developed," "Under Development," "Pending Release," "Cancelled," and "Released." However, the aforementioned "Not Developed," "Under Development," and "Pending Release" indicate that the corresponding product is at a stage where a requirement or function has not yet been fully implemented.

[0102] Optionally, if the user selects the target task information based on their needs on the operation interface, the interface can display various task information and their corresponding statuses. The user can then select the target task from among these multiple task information options. For example, the user can click on a task that has not yet been completed as the target task. Alternatively, the user can enter search criteria in the search box on the operation interface to filter the task information from multiple task information options and select the task that meets the search criteria as the target task.

[0103] It should be noted that the above-described process and method for determining target task information from a set of task information independently by the user is merely an illustrative example and is not intended to impose specific limitations. Any method that can determine target task information from a large amount of task information is within the protection scope of this application's embodiments.

[0104] Optionally, if the target task information is automatically selected through AIgi le, the current status of each task information in the task information set can be analyzed to determine which task information corresponds to which R&D tasks have been completed, which R&D tasks are in progress, and which R&D tasks have not yet started.

[0105] For example, this can be achieved by parsing the status fields of individual task information in the task information set. For instance, work items with the status "Published" will be considered as completed, while work items with the status "Under Development" or "To be Developed" will be considered as target tasks to be processed.

[0106] Optionally, when automatically selecting target task information via AIgi le, task priority can also be considered. For example, high-priority tasks can be tracked and analyzed first to ensure that important R&D tasks receive timely attention and resource allocation.

[0107] It should be noted that the above-described process and method for actively determining target task information from the task information set using AIgi le are merely illustrative examples and are not intended to impose specific limitations. Any specific operation that can automate the selection of the target task information using AIgi le is within the protection scope of this application's embodiments, and will not be described in detail here.

[0108] In this embodiment, by automatically identifying currently incomplete target task information, the AIgi le system can automate the tracking of requirements, reduce manual operations, and improve the efficiency and accuracy of requirement management. Based on intelligent analysis using large model technology, the AIgi le system can automatically determine the status of work items, improving the accuracy and real-time performance of status assessment. The AIgi le system can automatically sort tasks according to their priority, ensuring that high-priority tasks are processed first. Furthermore, it can automatically filter and display target task information based on the user's current perspective and team workload, providing a personalized task management experience.

[0109] In summary, the AIgi le system automates and intelligently filters out currently incomplete target R&D tasks from the task information set, laying the foundation for subsequent automated requirements tracking and behavioral analysis. This process not only improves the efficiency of requirements management but also ensures accurate tracking of critical tasks and resource optimization.

[0110] Step S206: Obtain the set of operation events associated with the target task information.

[0111] In the technical solution provided in step S206 of this application, the operation events in the operation event set can be used to represent the operation behavior performed on the target task information. Operation events can refer to the interactive behaviors performed by the user in the working environment, including but not limited to click events, keyboard events, mouse hover, drag and drop operations, etc. This is only an example and is not a specific limitation. The operation event set can also be called a user event list. Operation behaviors reflect the specific activities of the user in performing a specific target R&D task, and can also be called user behaviors or user operations, such as writing code in a code editor, searching for technical documents in a browser, updating task status in a project management tool, etc.

[0112] In this embodiment, after determining the target task information of the product from the task information set, the set of operation events associated with the target task information can be obtained.

[0113] Optionally, in the AIgi le system's workflow, this embodiment serves as a crucial bridge connecting requirements management and user behavior analysis. Its core task is to capture and collect user action events related to each specific target R&D task. This process is essential for understanding user activities, behavioral patterns, and the context of requirements fulfillment during task execution.

[0114] Optionally, action events can be collected from the user's work environment in real time or periodically. In the AIgi le system, this can be achieved through plugins embedded in R&D tools, such as integrated development environments (IDEs), browsers, instant messaging (IM) software, or through operating system-level event listening. For example, the AIgi le system can listen for keyboard input, mouse clicks, file opening and saving operations, and record action events associated with the target task.

[0115] Optionally, after collecting operation events, the AIgi le system can associate the monitored events with specific target task information. It can analyze the user operation context, such as identifying the file, window, and application at the time of the operation, to determine whether the user operation is related to the current target development task. For example, if a user is editing a code file related to a specific work item, the AIgi le system will record the relevant keyboard and mouse events in the operation event set for that work item.

[0116] In this embodiment of the application, the AIgi le system monitors user interaction behavior to determine whether the interaction is related to a specific target R&D task. This method records and analyzes the user's specific behavior during the execution of each work item, providing detailed evidence and context for requirements tracking and project management. By correlating operation events and work item information, the progress and status of requirements can be more accurately determined, reducing errors from manual judgment. Based on the analysis results of user behavior, potential problems and bottlenecks in task execution can be identified, providing optimization suggestions and solutions.

[0117] In summary, the above embodiments represent key steps in the AIgi le system for collecting user behavior data, associating it with target R&D tasks, and supporting subsequent requirements management and project execution analysis. Through this process, more accurate requirements tracking, more comprehensive behavior analysis, and more intelligent key event generation can be provided, thereby optimizing project management processes and improving the efficiency and accuracy of requirements fulfillment.

[0118] Step S208: Using the target model, identify at least one key operation event from the set of operation events.

[0119] In the technical solution provided in step S208 of this application, the importance of key operational events to the target R&D task is greater than the importance of operational events other than key operational events in the operational event set to the target R&D task. The target model can also be called the target large model. The target model is trained on the large model, which can be trained through fine-tuning or other training methods, such as supervised learning, reinforcement learning, zero-shot learning, etc. The choice of training method depends on the application scenario of the AIgi le system, the type and amount of available data, and the requirements for model performance and resource consumption. For example, if the goal of the AIgi le system is to improve accuracy in highly specific scenarios, fine-tuning may be necessary. If the goal is to adapt quickly to constantly changing environments, incremental learning or zero-shot learning may be more suitable. The base model for training the target large model can be, in addition to the large model, a medium-sized model, a small model, a basic feature extraction model, a self-supervised model, etc., without specific limitations here.

[0120] Optionally, the large model can be a large-scale language model. The target large model can be used to identify and understand the importance and impact of operational events associated with the target R&D task. Key operational events can also be called critical events. A critical event can refer to an operation or event that has a particularly significant impact on the success or failure of a task during its completion. For different domains, a corresponding list of critical events can be predefined. This list can be understood as anchor points provided to the large model, helping it to extract information accurately and with high performance from massive amounts of raw events.

[0121] Optionally, the way key events are defined can include supervised and unsupervised methods.

[0122] For example, regarding monitoring methods, domain experts (such as senior project managers or project management platform standard setters) can be engaged to define critical events. This approach ensures that the information extracted from the large model is standardized and meets requirements. The advantage of this method is that it ensures the standardization and consistency of information extraction, and because experts possess rich experience and professional knowledge, they can accurately identify critical events. However, this method is costly due to its heavy reliance on experts' time and management resources, and may be subject to some subjective judgment bias.

[0123] For example, in the unsupervised approach, key events can be automatically defined using machine learning algorithms based on historical demand data. These machine learning algorithms can include clustering, key rule mining, sequence pattern mining, or deep learning models based on transformers. The advantages of this approach are that it requires no human intervention, saving time and costs, and its data-driven nature reduces subjective bias. However, this approach can be complex due to algorithm selection and parameter tuning, and the automatically extracted key events may not be as accurate as those defined by experts.

[0124] It should be noted that the above definition of key events is for illustrative purposes only and is not intended to impose any specific restrictions.

[0125] Optionally, the target large model can be a text-based large model. In this embodiment, (semi-)structured text input such as prompts and context can be used, and the target large model generates corresponding operation instructions, event summaries, and other information. Therefore, the main capability of the target large model is text-to-text generation, used to process and output structured information. Among them, operation instructions are used to assist users in performing automated operations, such as supplementing requirement descriptions, commenting on requirements, and facilitating requirement flow (e.g., switching the status from development to pending release), etc., without specific limitations. The essence of event summaries is a compression process, that is, converting the original event into a summary and highly readable information. For example, the original event can be a record of the user's operation of elements or input of text in the software during a time window. The event summary can be a discussion of user experience improvement in a certain software from 14:00 to 14:30. The final storage structure can be another manifestation of compression. Data from user interaction Q&A in the ODS layer can be retained for a long time.

[0126] It's worth noting that visual models can also be introduced in certain scenarios, such as when processing webpage screenshots or image comments. For example, the requirements and current webpage screenshots can be input into a visual model, which can then output structured information to determine whether the functions mentioned in the work item have been completed.

[0127] In this embodiment, after obtaining the set of operation events associated with the target task information, the target large model can be used to identify the set of key operation events with high importance from the set of operation events.

[0128] Optionally, in the AIgi le system's workflow, this embodiment represents a crucial intelligent screening and analysis step. Its core task is to utilize a pre-trained target large model to identify events from the collected set of operational events that have high importance or influence on the target R&D task. These critical operational events become the focus of subsequent automated requirements tracking and intelligent behavioral analysis, helping the AIgi le system and project managers more accurately understand task progress and key activities in the R&D process.

[0129] Optionally, before using the target large model to analyze the importance of operational events in the operational event set, the large model can be trained using training data. The training data may include a large set of historical operational events, corresponding R&D task descriptions, task status change records, task completion times, task-related code commit records, etc. This data is used to guide the model to understand which operational events are critical and how they affect the progress and status of tasks.

[0130] Optionally, the main function of the target big model is to identify key operational events. By analyzing the context information of these operational events (e.g., the application environment in which the event occurs, the event type, the event content, and its relevance to the task), it determines the importance of each operational event to the target R&D task. The goal of the target big model is to improve the efficiency and accuracy of requirements tracking and project management, ensuring the timely capture and analysis of key behaviors.

[0131] Optionally, critical operational events refer to those events that significantly contribute to or directly impact the completion of the target R&D task. Examples include the first code commit, completion of unit tests, resolution of major issues, and significant design changes. The target big model performs in-depth analysis on each operational event, considering its context and content to determine whether it qualifies as a critical operational event. The target big model can assess factors such as the type of operational event, its background, its relevance to the task, and its subsequent impact. Based on the judgment of importance, the target big model will filter out critical operational events, which may include, but are not limited to, code writing events, code testing events, problem-solving events, design change events, and code review events.

[0132] Optionally, for each identified critical operational event, the target big model can also assess its impact on the target R&D task, such as whether it significantly advances requirement development, solves key problems, or affects project schedule. The target big model can also rank critical operational events according to their priority and impact for team members' reference and subsequent automated requirement tracking.

[0133] In this embodiment, the AIgi le system utilizes a pre-trained target big model to understand and analyze the semantics and context of operational events, identifying key operational events. The target model can analyze the context in which operational events occur. Based on the type and content of operational events, the target big model classifies operational events and assigns scores or importance judgments to them. The AIgi le system can automatically filter out key operational events, reducing the manual screening work for team members and improving the efficiency of requirements tracking.

[0134] By employing the methods described above, identifying critical operational events ensures the automatic capture of important R&D activities within the target R&D task, improving the accuracy and timeliness of requirements tracking. This helps team members and project managers understand which activities are most critical to task progress, thereby optimizing resource allocation. Based on the identification of critical operational events, the system can intelligently analyze the behavioral patterns and efficiency of R&D personnel, providing insights and recommendations for team and project management.

[0135] In summary, the above embodiments represent key steps in the AIgi le system for intelligent filtering and identification of operational events using large model technology. Through this process, the AIgi le system can automatically identify and focus on operational behaviors that have a significant impact on the target R&D tasks, providing accurate information and insights for subsequent requirements tracking and project management. These steps not only improve the intelligence and efficiency of requirements tracking but also provide strong support for team collaboration and project decision-making.

[0136] It should be noted that the model training process in this application embodiment can adopt conventional training methods, mainly focusing on applications in specific fields.

[0137] Step S210: Based on key operational events, determine the product's R&D information.

[0138] In the technical solution provided in step S210 of this application, the R&D information can be used to represent different statistical information generated by the product during the R&D process. This information can be statistical information, such as multi-dimensional statistical information. The statistical information can be a multi-dimensional statistical report, or simply a statistical report.

[0139] In this embodiment, after identifying key operational events from the set of operational events using a target large model, product development information can be determined based on these key operational events. Development information can include statistics on dimensions such as time, code, and file distribution, and also covers indicators related to Artificial Intelligence (AI) interaction, search behavior, and code review, which are crucial for understanding the development process, evaluating work progress, and optimizing team collaboration. It should be noted that the information recorded in the above statistics, and the dimensions of the statistics, are only illustrative examples and are not specifically limited here.

[0140] Optionally, in the AIgi le system, the main task of this embodiment is to generate and update multi-dimensional statistical information of the product in the R&D process based on previously identified key operational events.

[0141] Optionally, the AIgi le system can integrate the identified key operational events in chronological order to generate multi-dimensional statistical information for the R&D process, including but not limited to: time distribution, code distribution, file distribution, AI interaction statistics, and search statistics. Time distribution can be used to represent the time distribution of key operational events during the R&D process, helping to understand the R&D cycle and time management. Code distribution can be used to represent statistics on new code, modified code, and deleted code, for assessing workload and code quality. File distribution can be used to represent file lists and file-level operational statistics, for file management and code review. AI interaction statistics can be used to reflect the effectiveness of AI assistance by tracking the number of interactions with the AI ​​assistant, question types, response times, etc. Search statistics can be used to represent the number of searches, search keywords, and visited web pages, for analyzing information acquisition behavior.

[0142] Optionally, based on key operational events, the AIgile system automatically updates the progress and status of each work item, such as changing "not developed" to "under development" and "pending testing" to "test completed," ensuring the real-time nature and accuracy of the status. By analyzing key operational events, the AIgile system can provide in-depth insights, including identifying efficiency bottlenecks, resource allocation issues, and team collaboration patterns, providing data support for project management decisions. Based on key operational events and multi-dimensional statistical information, the AIgile system can automatically generate R&D information (e.g., work reports), including detailed progress of each work item, problems encountered, solutions, and analysis of AI and search behavior, providing documentation support for team communication and project auditing.

[0143] In this embodiment, the AIgi le system can automatically generate R&D information, including descriptions of key operational events, intelligent assessments of work item progress, and in-depth insights into R&D activities. Through these methods, automated progress and status updates reduce manual operations and improve the accuracy and speed of information updates. Multi-dimensional statistical information and in-depth analysis reports help project managers and team members understand the R&D process and identify potential problems and optimization points. Automatically generated work reports can reflect project progress in a timely manner, promoting communication and collaboration among team members. Detailed statistical information and work records provide comprehensive data support for project auditing and post-project review, ensuring project transparency and traceability.

[0144] In summary, the AIgi le system generates multi-dimensional R&D information through key operational events, automatically generating and updating various statistical and analytical information related to the R&D process, providing strong data support for project management, team collaboration, and decision support. These steps not only improve the intelligence and efficiency of requirements tracking and R&D management but also ensure the integrity and accessibility of project data, which is of great significance for improving R&D quality and project success rate.

[0145] Through steps S202 to S210 of this application, if it is necessary to determine the R&D information of a certain product, the task information set of the product can be obtained. The task information set can be used to analyze the R&D tasks that the product needs to perform in the R&D process, and to identify the target R&D tasks that have not yet been completed in the R&D process, and to determine the target task information corresponding to these target R&D tasks. The operational behaviors required to complete the target R&D tasks can be determined, and the operation event set corresponding to the target task information can be obtained. The operation event set can be input into the target large model, and the target large model can be used to analyze each operation event in the operation event set, identifying the operation events with high importance to the target R&D tasks as key operation events. The product's R&D information can be determined based on the key operation events. In the embodiments of this application, AIgi le solves the problem of low product information processing efficiency in the prior art through automated work item management, operation behavior recording, and intelligent analysis of the large model. It can not only quickly and accurately identify and manage R&D tasks, but also deeply analyze operational behaviors, automatically extract key operation events, and generate rich statistical reports, thereby providing more efficient and intelligent decision support for project management. This resulted in improved information processing efficiency for the product, solving the technical problem of low information processing efficiency.

[0146] The method described in this embodiment will be further described below.

[0147] As an optional implementation, step S208, using the target model, identifies at least one key operation event from the operation event set, including: aggregating the operation events in the operation event set to obtain an event aggregation result; and using the target model to identify the key operation event from the event aggregation result.

[0148] In this embodiment, during the process of identifying key operation times from the set of operation events using a target large model, the operation events in the set can be aggregated to obtain an event aggregation result. The target large model can then be used to identify key operation events from the event aggregation result. The event aggregation result can be used to represent the result of classifying and aggregating recorded user events.

[0149] It should be noted that, in the embodiments of this application, aggregation processing can refer to rule-based processing, such as aggregating user operations by time or work item granularity, or aggregating user changes by code file granularity, etc., without specific limitations.

[0150] Optionally, this embodiment involves aggregating operational events to identify key operational events.

[0151] Optionally, in the AIgi le system workflow, this implementation further refines and enhances the identification process of critical operational events. Its core idea is to classify and aggregate user operational events to form more structured and meaningful event aggregation results, and then use a target large model to intelligently identify critical operational events based on these aggregation results. This method not only improves the accuracy and efficiency of event identification but also captures the correlations and patterns between operational events, providing deeper insights for subsequent requirements tracking and project management.

[0152] Optionally, the AIgi le system can classify operation events according to their type and context information. For example, it can classify click events, keyboard events, and notification events separately, or classify them according to the application environment in which the operation occurs (e.g., IDE, browser, IM software, etc.).

[0153] Optionally, based on the classification of the above-mentioned operation events, operation events categorized into the same group can be further aggregated to extract common features and patterns. For example, event aggregation algorithms can be used to aggregate multiple modifications to the same code file into code editing events, or a series of searches for different keywords but pointing to the same webpage can be aggregated into in-depth research events. Event aggregation not only reduces the number of events but also enhances the richness and information density of event descriptions. It should be noted that the above aggregation methods are only illustrative and no specific limitations are imposed here.

[0154] Optionally, when aggregating operation events, contextual information such as the user behavior timeline, code changes before and after, and changes in search keywords can be added to enhance the coherence and semantic information of the event aggregation results, providing a more comprehensive perspective for subsequent key operation event identification.

[0155] In this embodiment, the AIgi le system employs an event aggregation algorithm (e.g., rule-based methods or machine learning models) to automatically identify and aggregate similar events, while supplementing necessary contextual information to form structured event aggregation results. A pre-trained large model is used to perform deep analysis on the event aggregation results, assessing the criticality of events, identifying key operational events, and generating detailed event descriptions. When aggregating operational events, the system is able to perceive the context in which the operational events occur, including applications, code files, search keywords, AI interaction content, etc., which helps improve the accuracy of event identification and the richness of event descriptions. Through the above methods, event aggregation reduces the number of events processed and improves the efficiency and speed of identifying key operational events. Through contextual supplementation and event aggregation, event descriptions are richer and more accurate, reflecting the true meaning of the events and the details of the R&D process.

[0156] In summary, by optimizing event aggregation and target large-scale model identification, the process of identifying key operational events has been further improved, providing the AIgile system with more accurate R&D information and requirements tracking data. This not only enhances the intelligence and efficiency of project management but also provides strong data support for team collaboration and decision support. Through this process, the AIgile system can intelligently identify and focus on operational behaviors that have a significant impact on target R&D tasks, providing crucial information for subsequent requirements tracking and project management.

[0157] As an optional implementation, the target model is used to identify key operation events from the event aggregation results, including: generating event prompt information from the task information set, target task information, operation event set, and context information of the target task information, wherein the event prompt information is used to describe the key operation events; and guiding the target model to determine the operation type according to the event prompt information of the key operation, and identifying the key operation events under the operation type from the event aggregation results.

[0158] In this embodiment, during the process of identifying key operation events from the event aggregation results using the target large model, event prompt information can be generated from the task information set, target task information set, operation event set, and contextual information of the target task information. The event prompt information can guide the target large model to determine the operation type, and key event operations under the aforementioned operation type can be identified from the event aggregation results. The event prompt information can be used to describe the key operation events, and can also be referred to as key event descriptions.

[0159] Optionally, the process of identifying key operational events from the event aggregation results using the target large model can be further refined and structured to ensure the accuracy and efficiency of intelligent identification. The steps include generating event prompts, using these prompts to guide the target large model in determining the operation type, and identifying key operational events under a specific operation type from the event aggregation results.

[0160] Optionally, during the generation of event prompts, the task information set, target task information, operation event set, and contextual information of the target task information can be integrated. That is, the work item list, the current status and description of the work items, the timeline of user operations, event type, event content, the environment in which the event occurred (such as application, file, page, etc.), and the contextual information before and after the event can be integrated. Based on the above information, event prompts are constructed. These prompts can describe key operation events, including the event type, background, content, timestamp, and relevance to the target R&D task. The goal of event prompts is to provide specific and detailed event context for the target big model, enabling it to accurately identify and understand the nature and impact of the event.

[0161] Optionally, when constructing event notification information, the context in which the operation event occurs can be perceived, such as editing code in an IDE, searching for information in a browser, or having a team discussion in an IM software. This contextual information helps the target large model determine the importance of the event to the R&D task.

[0162] Optionally, the target big data model receives the integrated event prompts as input for its analysis and judgment. Based on its training data and algorithms, the target big data model can understand and analyze the event details and context described by the event prompts, thereby determining the operation type of the event. By analyzing the event prompts, the target big data model identifies the operation type associated with the event, such as "code development," "requirements analysis," "documentation writing," and "problem solving." The identification of operation types is based on the content, background, and degree of impact on the target R&D task of the event, which helps in the intelligent screening of subsequent key operation events. When identifying operation types, the target big data model can also consider the context in which the operation event occurs, which helps to more accurately understand the nature and importance of the event and avoid misjudging events with irrelevant background information or little impact on the task as key operation events.

[0163] Optionally, in identifying key operational events from the event aggregation results, based on a defined operational type, the AIgi le system can filter out key operational events belonging to that operational type from the event aggregation results. For example, if the target big model determines the operational type to be "code development," the AIgi le system will identify events related to code writing, code editing, and code submission from the event aggregation results, and further filter out events with high importance, i.e., key operational events. The target big model is then used to assess the importance of the filtered operational events, identifying those events that significantly contribute to the development task. Factors that the target big model can consider include: the frequency and duration of the operational events, the correlation between the operational events and the development task, and the impact of the operational events on the task status.

[0164] Optionally, for identified critical operational events, the system will record detailed event information, including: event type, time, location, content, and an assessment of its impact on the R&D task. This information will be used for subsequent requirements tracking, project management, and intelligent behavioral analysis.

[0165] In this embodiment, within the AIgi le system, the target large model can understand the semantics and context of event prompts, accurately identifying the operation type of the event, which is the foundation for identifying key operation events. Employing a context-aware algorithm, it can understand and analyze the background environment in which the event occurs, improving the accuracy of event identification and description. It can automatically identify and classify the types of operation events, providing a basis for the screening of key events. Through the above methods, event prompts are generated and the target large model is guided to identify operation types, improving the accuracy and efficiency of key event screening. The identification of key operation events provides intelligent support for requirements tracking, ensuring the automatic capture and recording of important R&D activities. By identifying and recording key operation events, the AIgi le system can provide deep insights, including identifying efficiency bottlenecks, resource allocation issues, team collaboration patterns, etc., providing data support for project management decisions. Detailed recording and intelligent analysis of key events facilitate information sharing and collaboration among team members, improving the efficiency and effectiveness of communication. The identification and analysis of key operation events provides data-driven decision support for project management, improving the level of intelligence in project management.

[0166] In summary, by generating event notifications, utilizing the target large model to determine operation types, and intelligently identifying key operation events from event aggregation results, the critical event identification process in the AIgi le system has been further optimized, improving the intelligence and efficiency of requirements tracking and project management. These steps not only ensure the capture and recording of important R&D activities but also provide strong data support for team collaboration and project management decisions. Through this process, the system can intelligently identify and focus on operational behaviors that have a significant impact on the target R&D tasks, providing crucial information for subsequent requirements tracking, project management, and intelligent behavior analysis.

[0167] As an optional implementation, the operation events in the operation event set are aggregated to obtain the event aggregation result, including: aggregating the operation events in the operation event set onto the time axis according to the operation time corresponding to the operation event to obtain the event aggregation result.

[0168] In this embodiment, during the aggregation process of operation events in the operation event set to obtain the event aggregation result, the operation events in the operation event set are aggregated onto a timeline according to the operation time corresponding to the operation event, thus obtaining the event aggregation result. The timeline can be a user event timeline, or a user behavior timeline, or simply a user behavior timeline.

[0169] Optionally, aggregating the operation events in the operation event set onto a timeline according to the operation time to obtain the event aggregation result is an important step for the AIgi le system to systematically record and intelligently analyze user operation behavior. The above process not only provides a clear time clue for subsequent key event identification, but also makes it possible to analyze the time distribution and behavioral patterns of R&D activities.

[0170] Optionally, during the aggregation of operation events on a timeline, a timeline can be established. This timeline serves as a timeframe for recording user actions during the execution of R&D tasks. The starting point of the timeline can be the start time of a project or the start of a workday, while the ending point can be set according to actual needs, such as the end of a project or the end of a day. The AIgile system marks each operation event in the operation event set on the timeline according to its specific occurrence time. The operation time is provided by the system timestamp at the time the event occurred, ensuring the accurate position of the event on the timeline.

[0171] Optionally, the AIgi le system categorizes operation events on the timeline. For example, it labels different types of events such as click events, keyboard events, and notification events separately, which helps in subsequent analysis of the distribution of event types over time. When recording operation events on the timeline, contextual information about the event can also be added, such as the application environment in which the event occurred (e.g., IDE, browser, IM software), file or page information, and changes in code or text content before and after the event. This makes the description of each event on the timeline richer and more specific.

[0172] Optionally, the AIgi le system analyzes the types and content of adjacent events on the timeline, aggregating events with similar properties or continuous characteristics to form event aggregation results. For example, a series of consecutive code writing events can be aggregated into a "code development" event, or multiple searches for the same topic can be aggregated into a "deep search" event.

[0173] Optionally, event aggregation results can provide more structured and meaningful input for subsequent identification of key operational events using a large target model. By classifying and aggregating events on the timeline, the model can more easily identify events that have a significant impact on R&D tasks. Event aggregation results enable the analysis of the temporal distribution of user behavior. The system can generate statistics on the temporal distribution of R&D activities based on event records on the timeline, such as the types and amounts of user activities at different times of the day. This is very useful for understanding the work habits of R&D personnel and optimizing work schedules. By aggregating operational events on the timeline, the AIgi le system can gain insights into the behavioral patterns of R&D personnel. For example, it can identify that R&D personnel tend to search for information or interact with AI assistants when encountering technical difficulties, or that there are a large number of IDE operation events during the code writing phase. This information helps optimize the R&D process and improve team collaboration efficiency. Based on the event aggregation results on the timeline, project managers can track changes in project progress and requirement status, understand the progress of each work item at different times, and thus better manage projects and allocate resources.

[0174] In this embodiment, the AIgi le system employs a timeline algorithm to accurately aggregate operational events onto a timeline in chronological order, providing a temporal clue for event analysis. When aggregating events, the AIgi le system can perceive the contextual information of the event, such as applications, files, and code snippets, which helps improve the accuracy of event descriptions and the intelligence of event aggregation. The AIgi le system uses event classification and aggregation technology to automatically identify and aggregate similar events, reducing the number of events and improving the efficiency of subsequent analysis. Through the above methods, timeline aggregation of events reduces the workload of analyzing individual events and improves the efficiency of identifying key operational events. Through contextual supplementation and event aggregation, the event descriptions generated by the system are more accurate and specific, reflecting the true meaning of the events and the details of R&D activities. The time distribution analysis of event aggregation results provides project managers with a time-based insight into R&D behavior, helping to optimize project progress and resource allocation. Event aggregation results on the timeline can automatically trigger updates to requirement status and progress, reducing the need for manual operations and improving the efficiency and accuracy of requirement management. Timeline aggregation of events promotes information sharing and collaboration among team members, improving the efficiency and effectiveness of communication.

[0175] As an optional implementation, the key operation events include at least one of the following: a search operation event, wherein the search operation event is used to represent a search operation behavior performed on the target task information; a code operation event, wherein the code operation event is used to represent a code operation behavior performed on the target task information; and an interaction operation event, wherein the interaction operation event is used to represent an interaction operation behavior performed on the target task information using the target model.

[0176] In this embodiment, the key operation events may include one of the following: search operation events, code operation events, and interaction operation events. Search operation events can represent search operations performed on target task information and are also referred to as search events. Code operation events can represent code operations performed on target task information and are also referred to as code events. Interaction operation events can represent interactive operations performed on target task information using the target big model and are also referred to as AI events.

[0177] Optionally, the key operational events identified by the AIgi le system include search operation events, code operation events, and interaction operation events, which respectively represent behavioral characteristics and activities in different dimensions during the research and development process.

[0178] Optionally, search action events capture the information query behavior of R&D personnel during the development process, including but not limited to using browsers to search for technical documents, solutions, case studies, or searching for relevant materials in the enterprise's internal document library. Search action events reflect the information needs of R&D personnel when solving problems, learning new skills, or making decisions. Key attributes of search events include search keywords, search time, visited web pages, search results, and possible click behaviors. This information not only records the information acquisition process but also helps the system understand and predict the next actions of R&D personnel.

[0179] Optionally, code operation events record the activities of developers in the code editor, including writing, modifying, deleting, copying, pasting code, and interacting with code completion tools (such as Copi Lot). Code operation events specifically describe the technical details of the development process and can reflect code quality, development efficiency, and complexity. Key attributes of code operation events include operation time, operation type (e.g., writing, modifying), involved code files, number of lines of code, and prefixes and suffixes of the code snippets. This information provides a foundation for code review, progress tracking, and code statistics.

[0180] Optionally, interactive events track the interactions between developers and the AI ​​assistant, including asking questions, querying, requesting code suggestions, and providing feedback. Interactive events demonstrate the assistive role of AI in the development process, enabling the assessment of the AI ​​assistant's response speed, accuracy, and contribution to improving work efficiency. Key attributes of interactive events include event time, question content, AI response content, interaction duration, and the developer's adoption of AI suggestions. This information helps optimize the AI ​​assistant's performance and improve its usability in development scenarios.

[0181] In this embodiment, by identifying and recording these key operational events, the AIgi le system can more accurately track R&D progress and status, reducing errors and omissions from manual updates. Recording and analyzing search events, code events, and AI events provides in-depth insights into developers' work habits, skill usage, and problem-solving strategies, helping to optimize team collaboration and resource allocation. Statistics on code operation events help assess code quality, identify efficiency bottlenecks, and evaluate the effectiveness of code completion tools, thereby guiding code review and improvements in coding standards. Recording interactive operation events can evaluate the performance of the AI ​​assistant, promoting its more effective integration in the R&D process, while providing data-driven optimization directions for the AI ​​assistant's intelligent enhancement. Based on these events, the AIgi le system can automatically generate multi-dimensional statistical reports, including time distribution, lines of code, AI interaction history, and search records, providing project managers and team members with a comprehensive overview of work progress.

[0182] Optionally, a pre-trained large model can be used to analyze the set of operational events to identify key events in search, code, and interaction operations. These steps rely on the model's understanding of operational behaviors and the accuracy of event recognition. When identifying events, the AIgi le system can perceive the context of the event, including the operating system environment, application environment, filename, and code snippets, to ensure the accuracy and coherence of the event description. Through event attribute analysis, the system can classify events by type, facilitating subsequent statistics and report generation. Based on the event's time, type, and context, the system can generate multi-dimensional statistical information, supporting analysis from multiple perspectives such as code, time, and AI interaction. Through this analysis, by identifying and recording key operational events in the R&D process, the AIgi le system can provide more accurate requirement tracking, a deeper understanding of R&D behavior, and optimized code quality and efficiency evaluation. Furthermore, the integration and use of the AI ​​assistant have been deepened, improving the overall work efficiency and collaboration quality of the R&D team.

[0183] As an optional implementation, the product's R&D information is determined based on key operation events, including: when the key operation event is a code operation event, determining the R&D information of the code operation event in different dimensions; when the key operation event is an interaction operation event, determining the R&D information associated with the target model; and when the key operation event is a search operation event, determining the product's search metadata, wherein the search metadata (Record searchmetadata) is used to represent the basic data required to perform the search operation event.

[0184] In this embodiment, during the process of determining product development information based on key operation events, when the key operation event is a code operation event, the development information of the code operation event under different dimensions can be determined. When the key operation event is an interaction operation event, the development information associated with the target large model can be determined. When the key operation event is a search operation event, the product's search metadata can be determined. The dimensions of the development information for code operation events can include time distribution, code distribution, and file distribution. The development information critical to the target large model can include information such as switching from other applications to the large model application, switching from the large model application to other applications, model version, prompt word content, and large model output content. Search metadata can be used to represent the basic data required to execute search operation events, such as the engine, timestamps, and keywords.

[0185] Optionally, based on the identified key operational events, the AIgi le system can further identify and analyze product development information, providing more specific and in-depth data support for project management and requirements tracking.

[0186] Optionally, when the key operation event is a code operation event, the AIgi le system can determine the R&D information from different dimensions, including time distribution, code distribution, and file distribution.

[0187] Optionally, based on time distribution, the system can statistically analyze the time distribution of code operation events throughout a day or the entire project cycle, providing a visual analysis of development efficiency and work habits. For example, the AIgi le system can analyze that developers are most efficient at writing code in the morning, while spending more time on code review and debugging in the afternoon.

[0188] Optionally, for code distribution, the number of lines of code and code type (functional code, test code, refactoring code, etc.) involved in code operation events can be recorded to assess code complexity and the code contribution of developers. For example, it can be calculated that 5,000 lines of new code were added in this development, of which 3,000 lines are functional code, 1,500 lines are test code, and 500 lines are refactoring code.

[0189] Optionally, based on file distribution, analyze the list of files involved in code operation events and count the number of lines modified for each file to provide a basis for code review and version control. For example, if 12 files were identified as being modified in this development, and three of them had more than 100 lines modified, these files may be the implementation code for critical functions.

[0190] Optionally, during the process of determining R&D information under interactive operation events, when the key operation event is an interactive operation event, the AIgi le system can determine the R&D information associated with the target large model, including: Application Switch (APP Switch) records, recording the behavior of R&D personnel switching from other applications to the large model application and from the large model application to other applications, to analyze the frequency and scenarios of AI tool usage; Model version information, recording the large model version used during the interaction, which helps to retrospectively analyze the model's iteration effects and performance improvements; Prompt content, saving the prompts sent by R&D personnel to the large model to understand the R&D personnel's query intent and question type; Large model output content, recording the response content of the large model, which can be used to evaluate the model's accuracy and helpfulness, as well as the R&D personnel's adoption of AI suggestions; Interaction duration and response time, statistically analyzing the duration of each AI interaction and the model's response time, providing data basis for model performance optimization and the R&D personnel's waiting experience.

[0191] Optionally, during the process of determining R&D information under a search operation event, when the key operation event is a search operation event, the AIgi le system can determine the product's search metadata. This search metadata is the basic data required to execute the search operation event, including: search engine (recording the search engines used by R&D personnel, which helps analyze information sources and reliability); timestamps (saving the exact time of each search operation, providing time clues for time series analysis and behavioral pattern recognition); keywords (extracting the search keywords entered by R&D personnel, reflecting their information needs and key areas of focus); Uniform Resource Locator (URL) for the search results page (recording the URL of the search results page, which can be used for subsequent web page content crawling and in-depth analysis); and clicked search result links (recording the specific search result links clicked by R&D personnel, which helps understand the subsequent impact of search behavior and the depth of information acquisition).

[0192] In this embodiment, based on code operation events, interaction operation events, and search operation events, the AIgile system can automatically update requirement status and progress, reducing manual operations and improving the efficiency and accuracy of requirement tracking. The AIgile system can generate multi-dimensional data reports, including time distribution, code distribution, file distribution, AI interaction history, and search records, providing project managers with comprehensive data analysis and decision support. By analyzing the specific operations and behavioral patterns of developers during code writing, AI interaction, and information search, the AIgile system can provide in-depth insights into development behavior, helping to optimize development processes and improve team collaboration efficiency. Recording and analyzing interaction operation events can evaluate the performance of the AI ​​assistant, including response time and suggestion adoption rate, providing data support for further optimization and integration of the AI ​​assistant. The time and file distribution statistics of code operation events can help analyze code quality and development efficiency, identify coding bottlenecks, and optimize coding standards and code review processes.

[0193] Based on the above analysis, by determining the implementation method of R&D information based on key operational events, the AIgi le system can gain a deep understanding of the R&D process from multiple dimensions. It provides support in areas such as automated requirements management, R&D behavior insights, and AI assistant performance evaluation, thereby improving the intelligence and efficiency of project management, optimizing R&D processes, and enhancing team collaboration and code quality. This process not only reduces the burden of project management but also promotes the work efficiency and innovation capabilities of the R&D team.

[0194] As an optional implementation, the R&D information of code operation events in different dimensions includes at least one of the following: time information required for code operation events in the time dimension; code writing results of code operation events in the code dimension; and code files required for code operation events in the file dimension.

[0195] In this embodiment, the R&D information for code operation events across different dimensions includes at least one of the following: time information required for the code operation event in the time dimension, code coding results for the code operation event in the code dimension, and code files required for the code operation event in the file dimension. The time information can be the time distribution of various categories within the code, as well as the time consumption percentage of interactions. The code writing results can be statistical information on different activities of the encoded code in the code distribution dimension.

[0196] Optionally, the AIgi le system collects and analyzes R&D information across different dimensions of code operation events to provide more comprehensive and in-depth insights into code development behavior.

[0197] Optionally, in the time dimension, the AIgile system records and analyzes the time information required for code operation events, specifically including: time distribution, analyzing the time distribution of code operation events throughout a day or the entire project cycle. For example, the AIgile system can identify that developers are most efficient in the morning, while in the afternoon they tend to focus on code review and debugging. This analysis helps project managers understand the team's work habits and optimize work arrangements to improve efficiency. Time consumption percentage, statistically analyzing the time consumption percentage for each category of code operation events (e.g., functional code writing, unit testing, code refactoring), as well as the time consumption percentage for interactions with larger models (e.g., code completion, problem solving). For example, it was found that adopting code completion suggestions significantly reduced code editing time, while code refactoring consumed relatively more time resources. This information is crucial for evaluating time allocation in the development process and optimizing workflows.

[0198] Optionally, at the code level, the AIgile system collects and analyzes the coding results of code operation events, primarily focusing on: the distribution of lines of code, counting the number of lines of code added, modified, and deleted in code operation events, and the distribution of code snippets. For example, the AIgile system can identify that 1000 lines of code were added in this development, of which 800 lines are for main functionality, 150 lines are for unit tests, and 50 lines are for documentation comments and auxiliary code. This helps in assessing the complexity and workload of the code. For code snippets that adopt suggestions, if the code operation event includes the adoption of coding suggestions for a large model, the AIgile system records the adopted code snippets. This can be used for subsequent analysis of the developers' reliance on the AI ​​coding assistant and the efficiency of adoption.

[0199] Optionally, at the file level, the AIgile system focuses on the code files involved in code operation events, specifically including: a file list, recording the files modified during code operation events, including filenames and modification dates. For example, the AIgile system can identify modifications to 10 files involved in this development, with core functional files being modified most frequently. Code change statistics are also provided, counting the number of lines of code added, modified, and deleted during code operation events for each file. For example, one code file might have 500 lines of code added, 200 lines modified, and 100 lines deleted; while another code file might have 300 lines of code added and 150 lines modified.

[0200] In this embodiment, through the analysis of different dimensions of code operation events, the AIgi le system can optimize requirement management. Specifically, based on time information and code file statistics, the system can automatically update requirement status and progress, reducing the need for manual input of requirement information and improving the efficiency and accuracy of requirement management. The AIgi le system can also improve code instructions; code-level analysis helps identify problematic areas in the code, such as overly complex or frequently modified code segments, guiding code review and refactoring, thereby improving the overall quality of the code. The AIgi le system can also optimize workflows; statistical analysis of time and file information can reveal efficiency bottlenecks in the workflow, such as spending too much time on code refactoring or repeatedly modifying specific files. This helps optimize workflows, reduce ineffective work, and improve team productivity. The AIgi le system can evaluate AI assistant performance; through code-level data, the AIgi le system can evaluate the performance of the AI ​​assistant in code completion, problem answering, etc., such as suggestion adoption rate and time saved. This provides data support for the optimization and further integration of the AI ​​assistant. The AIgi le system can provide data-driven decision support. That is, based on comprehensive analysis of time, code, and file dimensions, project managers can obtain data-driven decision support, such as reasonably allocating team members' work tasks, adjusting project schedules, or optimizing resource allocation strategies.

[0201] In summary, the AIgi le system provides in-depth insights into code development behavior by collecting and analyzing R&D information across three dimensions: time, code, and file. This not only optimizes requirements management and code review processes but also improves the efficiency and productivity of the R&D team. The implementation methods described above fully leverage the capabilities of large-scale models, bringing automation and intelligence to project management and requirements tracking, representing an innovative application of the AIgi le system in the field of R&D management.

[0202] As an optional implementation, the R&D information associated with the target model includes at least one of the following: application information associated with the target model, wherein the application information is used to indicate information on switching from running the target model to running the application, or the application information is used to indicate information on switching from running the application to running the target model; the output results of the target model; and prompts input to the target model.

[0203] In this embodiment, the R&D information associated with the target large model may include at least one of the following: application information associated with the target large model, the output results of the target large model, and prompt information output to the target large model. The application information may represent information about switching from running the target large model to running an application, or information about switching from running an application to running the target large model.

[0204] Optionally, the AIgi le system can gain a deeper understanding of how developers use AI assistance during the development process and the impact of AI assistants on R&D activities by collecting R&D information associated with the target large model.

[0205] Optionally, the AIgi le system collects application information associated with the target large model. This information mainly includes: switch records, which record the behavior of developers switching to or from other applications (IDEs, browsers, document editors, etc.) before and after using the target large model (e.g., an AI model). This information reflects the usage scenarios of the AI ​​model in the development process. For example, developers may switch from the IDE to the AI ​​model application for help when encountering coding difficulties, or switch back to the IDE after obtaining AI suggestions. Contextual information, which collects contextual information when developers switch from other applications to the AI ​​model application, or vice versa, such as code files edited in the IDE, topics searched in the browser, and document content processed in the document editor. This helps to understand the specific needs and scenarios of developers when using the AI ​​model, providing a basis for subsequent AI model optimization and feature enhancement.

[0206] Optionally, regarding the output of the target large model, which refers to the response content generated by the target large model after the developer sends a prompt message, the AIgi le system records the following information: Output content, saving the code snippets, solutions, suggestions, etc. generated by the model, which can be used to evaluate the model's accuracy and helpfulness, as well as the developer's adoption of the model's suggestions; Output time, recording the exact time when the model generates the output content, which helps to statistically analyze the model's response speed and evaluate its performance under high load.

[0207] Optionally, the system collects the following information regarding prompts input to the target large model: Prompts input to the target large model refer to queries or requests made by developers when using AI tools. The AIgi le system collects: prompt content, recording the queries or requests sent by developers to the model, which helps in understanding the types of questions and information needs of developers, providing data support for further model optimization; and prompt time, saving the specific time when the prompts were sent by developers, providing temporal clues for time series analysis and behavioral pattern recognition.

[0208] In this embodiment, the analysis of application information reveals the usage scenarios and frequency of AI models in the R&D process, which helps optimize the integration of AI models with development tools and improve the usability and user experience of AI assistants. Records of output results and input prompts provide fundamental data for model performance evaluation, helping to identify the model's strengths and weaknesses and providing direction for model iteration and performance optimization. By analyzing changes in the behavior of R&D personnel before and after using the AI ​​model, the AIgile system can gain insights into the role of AI in the R&D process, providing data-driven suggestions for process optimization and team collaboration. Statistical analysis of model response time and the efficiency of R&D personnel in adopting suggestions can assess the AI ​​tool's contribution to improving R&D efficiency, providing quantitative evidence for adjusting or improving R&D strategies. Based on the collected information, the AIgile system can generate statistical analysis reports on R&D personnel interactions with the AI ​​model, including interaction frequency, question type, adoption rate, etc., providing in-depth insights for project management and decision support.

[0209] In summary, the AIgi le system, through the collection and analysis of R&D information associated with the target large model, has gained a deeper understanding of the integration and effectiveness of AI tools in the R&D process. This not only optimizes the performance and integration strategy of the AI ​​assistant but also improves the work efficiency and collaboration quality of the R&D team. The above implementation fully leverages the capabilities of the target large model, bringing intelligent improvements to project management and requirements tracking, representing an innovative application of the AIgi le system in the field of AI-assisted development. Through this information, project managers and R&D teams can gain a comprehensive understanding of AI usage, enabling them to make more informed decisions and improve R&D efficiency and project quality.

[0210] As an optional implementation, step S204, determining the target task information of the product from the task information set, includes: selecting the target task information from the task information set in response to a selection operation on the task information set; and determining the target task information from the task information set in response to a system event and / or an operation event associated with the target task information.

[0211] In this embodiment, during the process of determining the target task information of the product from the task information set, if a selection operation is performed on the task information set, the target task information can be selected from the task information set. If a system event and / or an operation event associated with the target task information is triggered, the target task information can be determined from the task information set. The selection operation can be an operation performed by the user actively selecting the target task information from the task information set. The system event can be a system notification event. The operation event associated with the target task information can be a user action.

[0212] Optionally, the AIgi le system can determine the target task information of the product from the task information set, and can adopt two operating modes.

[0213] Optionally, one operational model involves user-initiated selection. When developers need to focus on or handle specific work items (tasks), they can perform a selection operation through the task information set (work item list). For example, they can directly click on a work item, filter work items by entering keywords in the search box, or use a shortcut key to highlight the current work item. The AIgi le system will respond to these selection operations, extracting and identifying the selected work item as the target task information from the task information set. The AIgi le system will update the current work item context to ensure that subsequent behavior analysis and data statistics are targeted at the work item, thereby providing accurate requirement tracking and analysis.

[0214] Alternatively, another operational model involves the automatic determination of system events and operational events. AIgile system events include system notifications, push notifications, email alerts, etc. When these events occur, the AIgile system automatically analyzes the event content to determine whether it is related to a specific work item. For example, when a developer receives a code review notification for a specific work item, the AIgile system will automatically identify and determine that work item as the target task information. Operational events mainly refer to the specific actions of developers within applications such as IDEs, browsers, and document editors, such as search operations, code editing, and meeting minutes. When these operational events are highly relevant to a specific work item, the AIgile system will automatically determine this work item as the current target task information. When a system event or operational event is triggered, the AIgile system will automatically determine the most relevant work item from the task information set as the target task information based on the event content and context information, thereby ensuring that subsequent behavior analysis and data statistics can focus on the currently processed work item.

[0215] In this embodiment, by combining user-selected data with automatic identification by the AIgi le system, the AIgi le system ensures that the tracking of each work item focuses on the currently being processed task, avoiding confusion and omissions in requirement information and improving the accuracy and real-time performance of requirement tracking. Based on automatically determined target task information, the AIgi le system can update work item progress in real time, reducing manual operations by developers and allowing them to focus more on development work, thus improving overall R&D efficiency. Using the target task information determined in the above manner, the AIgi le system can generate multi-dimensional statistical reports including time distribution, code volume, and interaction history, providing project managers with in-depth data insights and decision support. Automatically determined target task information ensures information synchronization among team members, especially in agile development environments, helping team members better understand their respective work priorities and strengthening collaboration and communication. Based on intelligent analysis of target task information, the AIgi le system can identify the needs and challenges of developers, providing customized solutions and resource recommendations, further improving R&D efficiency and problem-solving capabilities.

[0216] In summary, by determining target task information through user-initiated selection and automatic determination triggered by system events / operation events, the aforementioned process not only ensures the accuracy of requirements tracking but also optimizes the R&D process, improves R&D efficiency and team collaboration. This represents a significant innovation of the AIgi le system in intelligent requirements management and R&D process tracking. Through this implementation method, the AIgi le system can provide more accurate and efficient requirements tracking and analysis services, helping R&D teams adjust work plans in a timely manner, optimize resource allocation, and improve project success rates and product quality.

[0217] As an optional implementation, in response to a system event and / or an operation event associated with the target task information, determining the target task information from the task information set includes: in response to a system event and / or an operation event associated with the target task information, creating the target task information in the task information set; or, in response to a system event and / or an operation event associated with the target task information, switching the initial task information in the task information set to the target task information.

[0218] In this embodiment, the process of determining the target task information from the task information set upon triggering a system event and / or an operation event associated with the target task information can involve creating the target task information in the task information set, i.e., creating a corresponding work item. Alternatively, the initial task information in the task information set can be switched to the target task information, i.e., automatically switched to the corresponding work item.

[0219] Optionally, when the AIgi le system triggers system events and / or operation events associated with the target task information, the process of determining the target task information from the task information set includes two operation modes: creating a new work item and automatically switching to an existing work item.

[0220] Optionally, when the AIgi le system detects a system notification related to the project, such as task assignment or issue feedback, it automatically analyzes the event content to determine whether a new work item needs to be created to correspond to the event. If the action event (e.g., searching for a new feature implementation method in a browser or starting to write new feature code in an IDE) is not associated with a work item in the current task information set, the AIgi le system can identify that it may be a new code development task or requirement, thereby triggering the creation of a new work item.

[0221] Optionally, during the creation of a new work item, the AIgi le system can analyze the context of the triggering event, including the event type, time of occurrence, and application environment, to determine whether the event is relevant to the current project or product requirements. Based on the event context and pre-defined templates, the AIgi le system automatically generates a preliminary description of the new work item, including its title, background, objectives, and scope. The new work item is associated with the application, file, and search keywords that triggered the event, ensuring the focus and accuracy of subsequent behavioral analysis and data statistics. The AIgi le system can prompt developers via a toast message to confirm the creation of the new work item; developers can choose to confirm or decline. If developers confirm the creation of the new work item, the AIgi le system will add the new work item to the task information set and update the current work item list for subsequent operations and management.

[0222] Optionally, when switching to an existing work item, if a developer receives a code review notification for a specific work item, the AIgi le system will automatically identify and determine whether the current work item needs to be switched to the work item associated with the notification. If an action event (such as opening a specific code file for editing in an IDE) is highly relevant to an existing work item, the AIgi le system will automatically identify this association and trigger the work item switch.

[0223] Optionally, during the switch to an existing work item, the AIgi le system analyzes the context of the operation event to determine if any work items closely related to the event exist in the task information set. If a work item associated with the operation event is found, the system automatically switches the current work item to this target task information, updates the work item context, and ensures that subsequent behavior analysis and data statistics can focus on the correct task. The AIgi le system prompts developers via a Toast message to confirm the switch to the new work item; developers can choose to confirm or decline. If the developers confirm the switch, the AIgi le system updates the progress status of the original work item, for example, changing the status from "Under Development" to "Pending Development," and sets the target task information to "Under Development."

[0224] In this embodiment, the AIgi le system enables real-time tracking and automatic updating of work items through the automatic identification and response to system events and operational events, ensuring the real-time nature and accuracy of project management. Automatic creation and switching of work items reduces manual operations for developers, allowing them to focus more on development work, avoiding repetitive labor during task switching, and improving overall R&D efficiency. Based on the process of automatically creating and switching work items, the AIgi le system can generate multi-dimensional statistical reports including time distribution, code volume, and interaction history, providing project managers and developers with in-depth data insights to help them better understand requirements and workflows and make more informed decisions. Through automatic identification and switching of work items, team members can promptly understand each other's work status and task priorities, strengthening internal collaboration and reducing communication costs. Real-time creation and updating of work items helps improve the speed of requirement response, ensuring a highly efficient and smooth process from requirement proposal to implementation. Based on automatically created and switched work items, the AIgi le system can intelligently analyze the needs and challenges of developers, providing customized solutions and resource recommendations, further improving R&D efficiency and problem-solving capabilities.

[0225] Through the steps described above, the AIgi le system can respond promptly and accurately to the actions of R&D personnel, achieving automation and intelligence in requirements management. In the process of automatically creating and switching work items, the AIgi le system fully leverages the analytical and judgmental capabilities of large-scale models, providing the R&D team with efficient and accurate requirements management and task tracking services, which helps improve project success rates and product quality. At the same time, the above process also simplifies the requirements management workflow, reduces the burden on R&D personnel, and allows them to focus more on core development tasks, thereby improving overall development efficiency.

[0226] As an optional implementation, the method further includes: generating operation prompt information from the task information set and key operation events, wherein the operation prompt information is used to indicate the creation of target task information, or the operation prompt information is used to indicate switching the initial task information to the target task information; creating the target task information in the task information set, including: when the operation prompt information indicates the creation of target task information, guiding the target model to create the target task information in the task information set according to the operation prompt information; and switching the initial task information in the task information set to the target task information, including: when the operation prompt information indicates the switching of the initial task information to the target task information, guiding the target model to switch the initial task information in the task information set to the target task information according to the operation prompt information.

[0227] In this embodiment, task information sets and key operation events can be used to generate operation prompts. During the creation of target task information within the task information set, if the operation prompt indicates the creation of target task information, the target model can be guided to create the target task information within the task information set according to the operation prompt. Similarly, during the switching from the initial task information in the task information set to the target task information, if the operation prompt indicates the switch, the target model can be guided to switch from the initial task information in the task information set to the target task information according to the operation prompt. The operation prompt can be used to indicate the creation of target task information or to indicate the switch from the initial task information to the target task information. The operation prompt can be a Prompt, used to determine whether to switch or create a new work item. The task information set and key operation events can be a Context, meaning the Context can include a list of work items and key user events within a time window.

[0228] Optionally, the AIgi le system guides the creation or switching of work items in the target large model by generating operation prompts and utilizing task information sets and key operation events (Context). The above process ensures the automation and intelligence of R&D activities.

[0229] Optionally, during the generation of operation prompts, the analysis of task information sets and key operation events can be used to guide the target large model to perform specific operations.

[0230] Optionally, the AIgi le system first collects and analyzes the current task information set and key operation events within the time window to construct a context containing information such as user behavior and work item status. Based on the constructed context, the AIgi le system generates an operation prompt message. This prompt describes the purpose and requirement of the operation in the current context, such as the explicit intent and relevant information to create a new work item or switch to another work item. The prompt may include specific conditions and parameters to refine the operation instructions for the target large model, ensuring that the model can accurately understand and execute the required operation.

[0231] Optionally, during the creation of target task information in the task information set, when a Prompt indicates the creation of target task information, the AIgi le system passes the Prompt to the target large model. The target large model determines whether a new work item needs to be created based on the Prompt content. If the target large model determines that a new work item needs to be created, it can automatically generate a preliminary description of the new work item based on the information provided in the Prompt (e.g., background description, objective, scope, etc.). The AIgi le system collects the description of the new work item generated by the model, adds it to the task information set, and creates a new work item entry. The AIgi le system updates the user interface, such as the user interface (UI), to ensure the visibility and selectability of the new work item in the work item list.

[0232] Optionally, during the process of switching initial task information to target task information from the task information set, when a Prompt indicates a switch to the target task information, the AIgi le system passes the Prompt to the target large model. The model determines whether to switch the current work item to another work item based on the Prompt content, typically based on the correlation between work items and the context information of the current operation. If the target large model determines that a work item switch is needed, it identifies the target work item based on the information in the Prompt and generates a corresponding switch operation instruction. The AIgi le system executes the switch operation instruction generated by the model, updates the status of the current work item (e.g., marking it as "paused" or "pending"), and switches to the target work item, setting its status to "active" or "in progress." The AIgi le system updates the UI to display the newly switched work item, ensuring that developers can clearly understand the changes to the current work item.

[0233] In this embodiment, by automatically creating and switching work items, the AIgi le system can reduce manual operations by developers in requirements management and improve the automation level of requirements tracking. The generation of Prompts, based on intelligent analysis of developer behavior and work item status, provides clear operational guidance for the model, ensuring the accuracy and timeliness of decisions. Automatic work item creation and switching helps optimize the development workflow, ensuring smooth transitions between work items and reducing interruptions and waiting times. By reducing manual operations and optimizing workflows, the AIgi le system can improve development efficiency, allowing developers to focus on code development and problem solving, rather than tedious requirements management and status updates. Based on the automatically created and switched work item information, the AIgi le system can generate more accurate time distribution, code volume statistics, and AI interaction records, providing deep data insights for project management and decision support. By analyzing Context and Prompts, the AIgi le system can better understand the behavioral patterns and needs of developers, laying the foundation for providing personalized services and suggestions.

[0234] Through the steps described above, the AIgi le system automates and intelligently manages the creation and switching of work items, significantly improving the efficiency and accuracy of requirements management and task tracking. This process fully leverages the decision-making and generation capabilities of the target large model, ensuring smooth work item management and a positive work experience for developers.

[0235] As an optional implementation, step S206, obtaining the set of operation events associated with the target task information, includes: capturing at least one type of operation event associated with the target task information; obtaining context information affected by the type of operation event; and using the context information and the type of operation event to record the operation events in the set of operation events.

[0236] In this embodiment, during the process of acquiring the set of operation events associated with the target task information, the type operation events associated with the target task information can be captured. Context information resulting from the action of the type operation events can be obtained. Using the context information and the type operation events, the operation events in the operation event set are recorded. The type operation events can be click events, keyboard events, notification events, etc., and are not specifically limited here. Context information can include coordinates, window, application, page, and element information, etc., and is not specifically limited here.

[0237] Optionally, when acquiring the set of operational events associated with the target task information, the AIgi le system takes the following steps to ensure a comprehensive and accurate record of the activities of R&D personnel.

[0238] Optionally, in the AIgi le system, the types of operation events include, but are not limited to, click events, keyboard events, and notification events. The AIgi le system can monitor and capture these events in real time to identify activities associated with target task information.

[0239] Optionally, the AIgile system deploys event listeners on the user's computer to monitor click, keyboard, and notification events in any application (including IDEs, browsers, IM software, etc.). For captured events, the AIgile system needs to identify whether these events are associated with current target task information. For example, if a developer opens a code file related to a work item in an IDE, the AIgile system can identify it as a keyboard event associated with target task information. The AIgile system categorizes the identified events by type, such as click events, keyboard events, etc., for subsequent processing and analysis.

[0240] Optionally, contextual information is crucial for understanding the meaning of operational events and can provide the environment and background in which the operational event occurs.

[0241] Optionally, after capturing an operation event, the AIgile system immediately obtains the contextual information associated with the event, such as the coordinates of the event occurrence, the currently active window, the application name, and the page title. The AIgile system can analyze this contextual information to determine the semantic group of the event. For example, if the event occurs in the code editing area of ​​an IDE, the AIgile system can further obtain information such as the code file name, cursor position, and code snippet. For subsequent event processing and data analysis, the system will associate and store the captured and analyzed contextual information with the corresponding operation event.

[0242] Optionally, recording operation events in the operation event set is one of the core functions of the AIgi le system, ensuring the integrity and traceability of R&D personnel activities.

[0243] Optionally, the AIgi le system combines contextual information and type of operation events to generate detailed descriptions of events, such as which element was clicked, which file was coded, or which notification was received. Based on the nature and context of events, the AIgi le system clusters similar events together to form operation event sets, which facilitates subsequent generation of key events and updates to work item status. Under the management of operation event sets, the AIgi le system can record each captured and described operation event in the ODS (Operational Data Set) to store raw, unprocessed user events. The operation event sets recorded in the ODS will serve as the data foundation for subsequent DWD (Data Decision Making) and ADS (Application Data Set) analysis and processing.

[0244] In the embodiments of this application, the above embodiments ensure that the AIgi le system can comprehensively and accurately record the activities of R&D personnel, providing a solid data foundation for subsequent automated requirements management, multi-dimensional statistics, and in-depth data analysis. By capturing operation events, obtaining contextual information, and recording operation event sets, the AIgi le system can intelligently understand the behavior of R&D personnel, automatically update the status of work items, generate key events, and provide interactive insight analysis. The above-mentioned automated and intelligent event recording and analysis methods greatly improve the efficiency of requirements tracking and project management, ensuring the integrity and accuracy of information, which is difficult to match with traditional project management tools. At the same time, the above process also provides real-time and reliable data support for team management, resource allocation, and project schedule control, which helps to improve overall R&D efficiency and project success rate.

[0245] As an optional implementation, step S210, based on key operational events, determines the product's R&D information, including: updating the target task information based on key operational events; and converting the updated target task information into a report file that includes R&D information.

[0246] In this embodiment, during the process of determining product R&D information based on key operational events, the target task information can be updated based on the key operational events, and the updated target task information can be converted into a report file that includes R&D information.

[0247] Optionally, the AIgi le system takes the above steps to ensure the accuracy and timeliness of R&D information when determining product R&D information based on key operational events.

[0248] Optionally, in the process of updating target task information based on key operational events, the AIgi le system first identifies key operational events using large-scale modeling technology based on the recorded set of operational events. Key events may include code commits, search operations, AI interactions, etc., which have a direct impact on product development information. After identifying key operational events, the AIgi le system automatically updates the target task information according to the nature and context of the event. For example, if the event is a code commit, the AIgi le system may update the code volume statistics of the work item; if it is a search operation, it may update the search records of the work item. The AIgi le system can infer the status changes of the work item based on key operational events, such as from "not started" to "in progress," or from "under development" to "testing phase." The updated target task information is synchronized to the project management tool in real time, ensuring that team members can see the latest work item status and progress.

[0249] Optionally, in the process of converting the updated target task information into a report file that includes R&D information, the AIgi le system aggregates and organizes the updated target task information, including time distribution, code statistics, search records, and AI interaction history. Based on the aggregated information, the AIgi le system automatically generates a detailed R&D information report using large model technology. The report may include the following: the current status and progress of work items, the workload and efficiency of R&D personnel within a specific time period, the main problems encountered and solutions, code statistics and optimization suggestions under AI assistance, and analysis and trend prediction of search behavior. The generated report file can be formatted into an easy-to-read and understand document, facilitating access and analysis by team members, project managers, or stakeholders. The AIgi le system supports sharing the converted report file with relevant team members via email, instant messaging tools, or project management platforms, ensuring information transparency and team collaboration.

[0250] In this embodiment of the application, by automating the identification of key operational events and the updating of target task information, the manual recording work of R&D personnel is reduced, improving the efficiency of requirements tracking and project management. Simultaneously, the R&D information reports automatically generated by the AIgi le system provide project managers and team members with real-time, accurate, and comprehensive insights into R&D progress, aiding project decision-making and resource allocation. By converting updated target task information into R&D information reports, the AIgi le system digitizes and structures R&D information, facilitating subsequent data analysis and mining. For example, time distribution statistics in the report can be used to identify bottlenecks in R&D efficiency, while code statistics help assess the complexity and workload of work items. Furthermore, the AIgi le system can intelligently generate multi-dimensional BI analysis reports based on the data in the reports, providing deeper data support for project management. The above process also enhances project traceability and transparency. By recording and reporting key operational events, it ensures that the progress and status of each work item can be accurately tracked, facilitating communication and collaboration among team members and simplifying project auditing and post-mortem analysis. In summary, the above implementation of the AIgi le system significantly improves the efficiency and quality of product development and project management through automation and intelligent means.

[0251] As an optional implementation, the method further includes: generating update prompt information from the task information set, target task information, operation event set, context information of the target task information, and key operation events, wherein the update prompt information is used to indicate whether updating the target task information is allowed or prohibited; updating the target task information based on the key operation events includes: in response to the update prompt information indicating that updating the target task information is allowed, guiding the target model to update the target task information according to the update prompt information.

[0252] In this embodiment, update prompt information can be generated from the task information set, target task information, operation event set, context information of the target task information, and key operation events. During the process of updating the target task information based on key operation events, if the update prompt information indicates that updating the target task information is permitted, the target large model can be guided to update the target task information according to the update prompt information. The update prompt information can be used to indicate whether updating the target task information is permitted or prohibited, and can be a Prompt message.

[0253] Optionally, when updating target task information based on key operational events, the AIgi le system introduces the concept of update prompts to ensure the intelligence and accuracy of update decisions.

[0254] Optionally, during the generation of update prompts, the AIgi le system first constructs a Context that includes a set of task information, current target task information, a set of operation events, context information, and key operation events. These steps provide a comprehensive view of the large model, encompassing developer activities, work item status, the environment in which events occur, and the importance of the events. Based on the constructed Context, the AIgi le system automatically generates a Prompt from the large model. The Prompt can be used to describe whether the current target task information should be updated based on key operation events, and how to update it. For example, the Prompt might suggest updating the work item status, adding new code statistics, or recording search behavior.

[0255] Optionally, a Prompt may include specific conditions and parameters to refine the operational instructions for the target large model, ensuring that the model can accurately understand and perform the required operations, such as updating the progress percentage of work items or adding specific notes.

[0256] Optionally, during the process of updating target task information, when the AIgi le system generates a Prompt indicating that updating the target task information is permitted, the AIgi le system will respond to the aforementioned instruction and guide the target big model to perform the update operation. Based on the guidance of the Prompt, the target big model analyzes the nature and contextual information of key operational events and intelligently updates the target task information. For example, the target big model may identify a significant increase in code volume, thereby updating the progress status of the work item; or it may identify frequent search behavior, prompting an increase in the complexity assessment of the requirements. The AIgi le system records the updates of the target big model to the work item information and synchronizes the above updates to the project management tool in real time, ensuring that team members can see the latest work item status and progress. If the update prompt indicates that updating the target task information is prohibited, the system will intelligently determine, based on the description of the Prompt, that the current key operational event has little impact on the work item status or progress, and will not update it, avoiding meaningless changes and maintaining the accuracy and stability of the work item information.

[0257] In this embodiment, by using Prompt, the AIgi le system can intelligently determine whether and how to update target task information, avoiding meaningless updates and improving the accuracy and efficiency of requirements tracking. The AIgi le system automatically updates work item information based on key operational events, reducing manual operations by developers, making requirements tracking more automated, saving time and effort, and improving project management efficiency. The generation of Prompt is based on a comprehensive analysis of user operation events, helping the AIgi le system to better understand developers' behavioral patterns and providing a basis for subsequent personalized services and suggestions. Through updated prompts and automatic synchronization, team members can promptly understand changes in the status of work items, helping to improve communication efficiency and collaboration within the team, ensuring smooth project progress. The updated target task information can be used for subsequent BI analysis and data mining, such as analyzing development efficiency, assessing the complexity of work items, and predicting project progress, providing data support for project management.

[0258] As an optional implementation, the updated target task information is converted into a report file including R&D information, including: determining at least one of the following target information of the updated target task information: update progress information, update process information, and update remarks information, wherein the update progress information is used to indicate the progress of updating the updated target task information, the update process information is used to indicate the process of updating the updated target task information, and the update remarks information is used to indicate the information noted during the updating of the target task information; and the target information is converted into a report file.

[0259] In this embodiment, during the process of converting the updated target task information into a report file including R&D information, at least one of the following target information can be determined: update progress information, update process information, and update remarks information. The aforementioned target information can be converted into a report file. The update progress information can be used to indicate the progress of the updated target task information, and can be the current progress percentage of the work item. The update process information can be used to indicate the process of updating the updated target task information, and can be a detailed process record. The update remarks information can be used to indicate information noted during the target task information update process, and can be additional remarks information.

[0260] Optionally, when converting the updated target task information into a research and development information report file, the AIgi le system focuses on converting and presenting three key aspects of information: update progress information, update process information, and update notes information.

[0261] Optionally, the AIgi le system updates progress information, assessing and recording the current completion progress of work items, typically expressed as a percentage. For example, if the coding task of a work item is 70% complete, the progress information update might record the current progress as 70%. This helps team members and project managers understand the real-time completion status of work items.

[0262] Optionally, update the process information to record in detail the evolution of the work item from its start to its current state, including key operations performed, problems encountered, solutions, code change records, search history, etc. Recording process information helps to trace the history of the work item and facilitates knowledge sharing and collaboration among team members.

[0263] Optionally, update the notes information, recording any additional notes that developers or the team may add during the work item update process, such as special requirements for the task, obstacles encountered, and important points communicated with the team. Collecting notes information helps team members understand and make decisions subsequently.

[0264] Optionally, during the process of converting target information into report files, the AIgi le system integrates and structures the collected target information (update progress, process information, and remarks) to form a comprehensive R&D information report. The AIgi le system converts the integrated information into a specific report file format, such as PDF, Word, or HTML, for easy reading and sharing. The report file may include the following sections: a work item summary, listing work items including name, status, and progress; progress analysis, showing the updated progress information for each work item, including percentage of completion and remaining time; process review, providing updated process information for each work item, including timelines and key event records; and a remarks summary, collecting and organizing updated remarks for work items to help team members understand areas of particular interest. The AIgi le system uses large model technology to automatically generate report files, ensuring the accuracy and timeliness of the information. The report files are not only easy to read but also include graphs and tables to visually represent R&D information. The generated report files can be stored in the AIgi le system as part of the historical record and can also be shared with team members via email, instant messaging tools, or project management platforms, ensuring transparency and accessibility.

[0265] In this embodiment, by converting the updated target task information into a report file containing R&D information, the progress, process, and notes of each work item can be clearly understood through the report file, which helps improve the transparency of project management. The process and notes information in the report file facilitates communication and information sharing among team members, promoting collaboration and problem-solving. Progress analysis and process review provide detailed data, helping project managers make data-driven decisions, such as resource reallocation and adjusting project plans. Stored report files can serve as a basis for project debriefing and auditing, helping the team summarize lessons learned and identify areas for improvement. The notes collected through the report files further reveal the specific needs or preferences of team members, laying the foundation for providing personalized services and recommendations.

[0266] In summary, the AIgi le system, by converting updated target task information into report files containing R&D information, achieves the structuring, visualization, and sharing of R&D information, improving the efficiency and quality of project management. This represents an innovative practice in the fields of project management and requirements tracking. The process fully leverages the analytical and transformation capabilities of large-scale models, transforming complex data into easily understandable information, providing powerful data support and tools for project management and team collaboration.

[0267] As an optional implementation, the method further includes: using the hierarchical data storage results to hierarchically store operation events and associated event information of operation events, wherein the associated event information is used to represent events and / or information associated with the operation events.

[0268] In this embodiment, a hierarchical data storage structure can be used to store operation events and their associated event information in a hierarchical manner.

[0269] Optionally, this embodiment is used to efficiently manage and utilize operational events and their associated information. The above steps involve hierarchical data storage technology and segmented storage strategies for event information.

[0270] Alternatively, tiered data storage is a data management strategy that stores data in different tiers or types of storage media based on its access frequency and importance. Frequently accessed "hot" data is stored in high-speed storage (such as SSDs and RAM), while less frequently accessed "cold" data is moved to lower-cost, higher-capacity storage. Tiered storage can optimize data access speed and storage costs, improving the overall performance and efficiency of the system.

[0271] Optionally, in software development and data analysis, operational events (such as code submissions, system operations, and user interactions) are often closely related to a series of associated event information (such as state changes before and after the event, other events triggered by the event, and contextual information of the event). This associated event information is crucial for understanding the background, causes, and results of the event, but it may contain a large amount of data and details. Storing it directly with operational events can lead to problems such as complex data management, high storage costs, and slow access speeds.

[0272] Optionally, the partitioned storage strategy stores operation events and their associated event information separately. Operation events, as frequently accessed "hot" data, are stored in high-speed storage, while associated event information, as less frequently accessed "cold" data, is stored in lower-cost, higher-capacity storage. This allows for rapid retrieval of operation events, while access to lower-cost, slower storage is only needed for analyzing or retrieving associated event information, achieving both high efficiency in data access and cost-effectiveness in storage.

[0273] Optionally, the storage method for related event information needs to enable rapid retrieval and recovery to facilitate event analysis, data mining, and troubleshooting. Possible implementation methods include creating indexes, using efficient database query techniques, and employing data compression and archiving strategies.

[0274] Optionally, in software development management, related event information can be used to deeply analyze the impact of operational events, optimize development processes, and improve the efficiency of requirements management and project management. For example, when a critical operational event (such as a code merge failure) is detected, all information related to the event (such as code review history, merge conflict details, team communication records, etc.) can be quickly accessed and analyzed to help the development team quickly locate the problem and take effective measures to resolve it.

[0275] In the embodiments of this application, the strategy of separating storage operation events and related event information, combined with hierarchical data storage technology, can effectively manage large-scale datasets, improve data access speed, reduce storage costs, and provide rich data analysis and event backtracking capabilities, which is of great significance for software development, data analysis and other scenarios.

[0276] As an optional implementation, the hierarchical data storage structure includes a first-level data storage result, a second-level data storage structure, a third-level data storage structure, and a fourth-level data storage structure. The associated event information includes: operation event sets, key operation events, aggregated key operation events, and R&D information. The hierarchical data storage structure stores the associated event information of operation events and operation times hierarchically, including: storing operation event sets using the first-level data storage structure; storing key operation events using the second-level data storage structure; storing aggregated key operation events using the third-level data storage structure; and storing R&D information using the fourth-level data storage structure.

[0277] In this embodiment, a first-layer data storage structure can be used to store operational events. A second-layer data storage structure can be used to store critical operational events. A third-layer data storage structure can be used to store aggregated critical operational events. A fourth-layer data storage structure can be used to store R&D information. Specifically, the first-layer data storage structure can be an ODS (Original Data Storage System) for storing raw user events. The second-layer data storage structure can be a DWD (Data-Driven Data Storage System) for storing critical events extracted from the raw user events. The third-layer data storage structure can be a DWS (Data-Driven Data Storage System) for storing aggregated data of critical events. The fourth-layer data storage structure can be an ADS (Advanced Data Storage System) for storing highly summarized analytical data at different time granularities.

[0278] Optionally, the AIgi le system employs a multi-layered data storage structure to manage operational events, critical events, and aggregated data, thereby ensuring data integrity and efficient analysis.

[0279] Optionally, within the ODS, the AIgi system captures user keyboard events, click events, and notification events through the operating system's accessibility API or a specific application's programming interface. These events contain raw records of user actions such as coding in an IDE, searching in a browser, and communicating in IM software. The captured events are first stored at the ODS layer, where the data is unprocessed, preserving the event's original state and contextual information, such as event type, timestamp, coordinates, window ID, application name, and page title.

[0280] Optionally, within the DWD, the AIgi le system analyzes the raw events stored in the ODS layer, using large model technology to identify key operational events that directly impact the product development process, such as code submissions, search behaviors, and AI interactions. These key operational events are structured and stored in the DWD layer, where the data has been categorized and extracted according to event type, retaining detailed information such as event type, time, and context for subsequent analysis and processing.

[0281] Optionally, within the DWS, the AIgi le system aggregates key operational events stored in the DWD layer to generate aggregated key operational event data, such as code volume statistics, search frequency, and AI interaction counts. At the DWS layer, the AIgi le system constructs a series of analytical metrics to analyze and display R&D information from different dimensions, such as time distribution, code distribution, and file distribution.

[0282] Optionally, within ADS, the AIgi le system stores highly aggregated data at the ADS layer, summarizing the analytical metrics of the DWS layer at a time granularity, such as daily, weekly, and monthly aggregated data, to support insightful analysis across different time windows. The ADS layer constructs application data, including statistical analysis information such as time distribution statistics, work item progress, work reports, AI interaction history, and search history, to facilitate interactive queries and multi-dimensional analysis.

[0283] In this embodiment, the introduction of a multi-layered data storage structure provides the AIgi le system with the following advantages: The ODS layer ensures a complete record of user operation events, providing a solid foundation for subsequent event analysis. The DWD and DWS layers enable rapid extraction and aggregation of key events, improving the efficiency and speed of data analysis. The statistical analysis information in the ADS layer supports interactive queries, allowing users to understand and analyze R&D information from multiple dimensions, providing real-time data support for decision-making. Through event recording and analysis in the DWD and DWS layers, the AIgi le system can accurately trace back the entire R&D process of work items, facilitating team review and auditing. The multi-layered data storage structure supports hierarchical queries, allowing users to obtain data at different granular levels as needed, such as raw events, key events, analytical metrics, or highly summarized report data. The hierarchical storage architecture is easily expandable, allowing the addition of new event types or analytical dimensions, such as code quality metrics or team collaboration data, to meet the diverse needs of project management and requirements tracking.

[0284] In summary, the AIgi le system, by employing a four-layer data storage structure of ODS, DWD, DWS, and ADS, achieves efficient management and in-depth analysis of user operation events, key operation events, and aggregated data. This provides accurate, real-time, and comprehensive data support for project management and requirements tracking, serving as a crucial technical foundation for automated and intelligent requirements management. The multi-layered data storage architecture not only improves data processing efficiency but also ensures the flexibility and depth of data analysis, forming the basis for large-model-driven requirements management and R&D event analysis.

[0285] As an optional implementation, the method further includes: acquiring inquiry information regarding R&D information; and analyzing the inquiry information using a target model to obtain response results that match the inquiry information.

[0286] In this embodiment, interactive insight analysis can be performed; specifically, queries regarding R&D information can be obtained. The query information can be analyzed using a target large model to obtain response results that match the query information.

[0287] Optionally, the AIgi le system achieves interactive insight analysis by acquiring query information related to R&D information and using a target large model for analysis to obtain response results that match the query information.

[0288] Optionally, during the process of obtaining queries regarding R&D information, team members or project managers can submit specific queries to the AIgi le system, such as, "Which work items saw the largest increase in code volume this week?" or "Who interacted with the large model the most last week?" The queries can be in natural language, allowing users to express their needs intuitively. The queries are input into the AIgi le system for the target large model to understand and analyze. The AIgi le system can preprocess the natural language queries to ensure the model accurately understands the user's intent.

[0289] Optionally, during the analysis of query information using the target big model, the target big model can understand the input query information, analyzing the query intent, keywords, and contextual information. The target big model can invoke a specific parser or use natural language processing techniques to understand the details of the query. Based on the understood query intent, the target big model retrieves data matching the query information from the ADS layer. For example, if the query is for code volume statistics, the model will look for relevant code events and statistics from the ADS layer. The retrieved data is further analyzed and processed to generate detailed response results. This may involve data filtering, summarization, sorting, or calculation of specific metrics. Based on the analyzed and processed data, the model generates response results matching the query information. The response results may be a statistical table, a chart, or a detailed text description, depending on the type and content of the query.

[0290] In this embodiment, by utilizing a target large model to analyze query information, the AIgi le system can quickly generate response results, improving the response speed of interactive queries. The response results, based on actual R&D information, provide data support for project managers and team members, helping them make more informed decisions. Interactive insight analysis allows users to gain a deeper understanding of R&D information from multiple dimensions, such as time distribution, code statistics, and AI interaction history, which helps identify potential problems and optimization points. The AIgi le system can provide personalized responses based on user queries, meeting the different needs of different users or roles for R&D information. Through interactive insight analysis, team members can clearly understand the R&D status of the project, promoting information sharing and team collaboration.

[0291] In summary, the AIgi le system achieves interactive insight analysis of R&D information by acquiring queries about R&D matters and analyzing them using a target large model to obtain responses that match the queries. This process fully leverages the natural language understanding and data analysis capabilities of the large model, providing real-time, accurate, and in-depth data insights for project management and team collaboration, significantly improving the transparency and decision-making efficiency of the R&D process. Through interactive insight analysis, the AIgi le system not only provides real-time feedback on R&D information but also supports in-depth data mining and predictive analysis, offering comprehensive data support for project management.

[0292] This application provides another method for product information processing from the perspective of a product development platform. Figure 3 This is a flowchart of another product information processing method according to an embodiment of this application, such as... Figure 3 As shown, the method may include the following steps:

[0293] Step S302: Obtain the task information set of the smart product, wherein the task information in the task information set is used to describe the R&D tasks that need to be performed in the R&D process of the smart product.

[0294] Step S304: Determine the target task information of the intelligent product from the task information set, wherein the target task information is used to describe the target R&D tasks that were not completed during the R&D process of the intelligent product.

[0295] Step S306: Obtain the set of operation events associated with the target task information, wherein the operation events in the set of operation events are used to represent the operation behavior performed on the target task information.

[0296] Step S308: Using the target model, identify at least one key operation event from the set of operation events. The key operation event is more important to the target R&D task than the operation events in the set of operation events other than the key operation event. The target model is obtained by training a large model.

[0297] Step S310: Based on key operational events, determine the R&D information of the smart product, wherein the R&D information is used to represent the different statistical information generated by the smart product in the R&D process.

[0298] The information processing method provided in this application aims to efficiently generate product development information from the perspective of a product development platform by automatically collecting, intelligently analyzing, and statistically analyzing key operational events, thereby improving the accuracy and efficiency of project management.

[0299] Optionally, in the product development process, projects often include multiple tasks, each with a specific execution status and importance. The product development platform needs to track the progress of these tasks in real time and analyze development activities for effective project management and decision-making. Product development personnel perform numerous operations during development, such as coding, searching for information, interacting with AI models, and testing code. These operational events are scattered across different tools and platforms; manually collecting and analyzing these events is not only time-consuming but also prone to omissions and errors. Automating the capture of these operational events and intelligently identifying and analyzing key operational events is crucial. Utilizing large-scale models for intelligent analysis of task information and operational events can identify key operational events that significantly impact the target development task. Through learning from massive amounts of data, large-scale models can understand the importance of events, thereby intelligently filtering out events that directly affect task progress and results.

[0300] In this embodiment, by determining the product's target task information and key operational events, the execution status and progress of tasks can be accurately tracked and analyzed, avoiding errors and omissions from manual recording. Identifying and distinguishing key operational events that significantly impact tasks facilitates refined task management and ensures the rational allocation of resources and time. R&D information generated based on key operational events provides real-time insights, helping project managers and team members understand the project's real-time status and adjust strategies or resolve issues promptly. Automated collection and analysis of operational events reduces the need for manual operations, improves the automation level of project management, and thus enhances overall management efficiency. The statistics and analysis of R&D information provide data support, allowing team members to make more scientific decisions based on data, such as resource allocation, schedule adjustments, and problem prioritization. Automatically generated R&D information, including task progress, code statistics, and AI interaction records, promotes information sharing among team members and enhances collaboration and communication. Analyzing key operational events and R&D information helps optimize work allocation, rationally schedule tasks, and improve team efficiency and project output quality. Intelligent analysis of operational events and refined management of target R&D tasks can significantly improve the user experience for R&D personnel, reduce meaningless operations, and increase job satisfaction. Automated requirements tracking and analysis reduce the need for manual operations, thereby saving labor costs and improving cost-effectiveness.

[0301] In steps S302 to S310 of this embodiment, key operation events and R&D information are generated by automatically acquiring and intelligently analyzing operation events, providing the product development platform with refined, real-time, and data-driven requirements tracking and analysis services. This process effectively solves problems in traditional project management such as information dispersion, low quality of requirements submission, and inaccurate statistical information, significantly improving the efficiency and quality of project management. It represents a major innovation in project management and requirements tracking. It not only enhances project transparency and team collaboration efficiency but also provides reliable data support for decision-making, making it a key technology for agile development and intelligent project management. Furthermore, it achieves the technical effect of improving product information processing efficiency, solving the technical problem of low product information processing efficiency.

[0302] This application also provides a product information processing method from the human-computer interaction perspective. Figure 4 This is a flowchart of another product information processing method according to an embodiment of this application, such as... Figure 4 As shown, the method may include the following steps:

[0303] Step S402: Display the product's task information set on the operation interface. The task information set describes the R&D tasks that need to be performed in the product's R&D process.

[0304] In the technical solution provided by step S402 of this application, the task information set of the product is displayed intuitively on the operation interface, thereby providing users with a clear task overview and execution guidance.

[0305] Optionally, the task information set can be displayed in the form of a list or cards on the user interface, allowing developers to see the development tasks to be executed at a glance. Each task may include key details such as task name, status (e.g., pending development, under development, tested, completed), priority, estimated completion time, and associated personnel.

[0306] Optionally, the task information set on the user interface should be able to update in real time, reflecting the latest status of the tasks. This helps ensure that team members have an accurate understanding of the project's real-time progress, improving team collaboration and decision-making efficiency. The task information set should be highly interactive, allowing users to view and filter tasks based on different filtering criteria (such as status, priority, associated personnel, etc.), and supporting clicking on tasks to view detailed information or perform actions, such as assigning tasks, changing status, adding notes, etc.

[0307] In this embodiment, the task information set is intuitively displayed on the user interface, providing strong support for task management during product development. This not only enhances team members' visibility and understanding of tasks but also promotes efficient task allocation, tracking, and management, a key step in achieving agile development, optimizing project management, and tracking requirements. This display mechanism enables the product development platform to better adapt to fast-paced development environments, ensuring that each task is processed promptly and accurately, thereby improving the efficiency and output quality of the entire development process.

[0308] Step S404: In response to the information processing command applied to the operation interface, the product's R&D information is displayed on the operation interface. The R&D information represents different statistical information generated by the product during the R&D process. At least one key operation event is determined based on the operation event set associated with the target task information in the task information set. The target task information describes the target R&D task that was not completed during the product's R&D process. The operation events in the operation event set represent the operation behaviors performed on the target task information. The importance of the key operation event to the target R&D task is greater than the importance of the operation events in the operation event set other than the key operation event to the target R&D task. The key operation event is identified from the operation event set using the target large model, which is obtained by training a large model.

[0309] In the technical solution provided by step S404 of this application, if an information processing instruction on the operation interface is detected, the task information set can be processed to obtain research and development information, and the research and development information can be displayed on the operation interface.

[0310] In steps S402 to S404 of this embodiment, the product's task information set is displayed on the operation interface; in response to information processing instructions applied to the operation interface, the product's R&D information is displayed on the operation interface. This achieves the technical effect of improving the product's information processing efficiency and solves the technical problem of low product information processing efficiency.

[0311] This application also provides a method for storing product information. Figure 5 This is a flowchart of a product information storage method according to an embodiment of this application, such as... Figure 5 As shown, this method enables efficient storage and management of product information, particularly task information and operational events within the product development process. The method aims to improve the efficiency and accuracy of R&D information processing by identifying and storing key operational events through hierarchical data storage and machine learning models. The method may include the following steps:

[0312] Step S502: Obtain the product's task information set, wherein the task information in the task information set is used to describe the R&D tasks that need to be performed in the product's R&D process.

[0313] In the technical solution provided in step S502 of this application, product-related R&D task information is collected, including task description, status, priority, responsible person, and completion time. This information can come from multiple data sources such as project management tools, code repositories, requirement documents, and test reports. The purpose of obtaining this information is to provide foundational data for subsequent analysis steps, helping the system understand the current R&D status and task requirements of the product.

[0314] Step S504: Determine the target task information of the product from the task information set, wherein the target task information is used to describe the target R&D tasks that were not completed during the product R&D process.

[0315] In the technical solution provided in step S504 of this application, among all the acquired task information, it is possible to identify which tasks are currently incomplete, i.e., the target R&D tasks. This involves filtering task statuses, such as selecting tasks in the "pending" or "in progress" states. Identifying target R&D tasks helps to focus on key development activities, improving R&D efficiency and the rationality of resource allocation.

[0316] Step S506: Query the set of operation events associated with the target task information from the hierarchical data storage structure. The operation events in the set of operation events are used to represent the operation behaviors performed on the target task information.

[0317] In the technical solution provided in step S506 of this application, a set of operation events associated with the target R&D task is obtained by querying the hierarchical data storage structure. Operation events can be specific actions performed by R&D personnel during task execution, such as code submission, file modification, meeting minutes, and communication emails. By associating operation events, the execution status of the task can be tracked, and potential risks or bottlenecks can be identified.

[0318] Step S508: Using the target model, identify at least one key operation event from the set of operation events, and store the key operation event in the data storage structure of the corresponding level in the hierarchical data storage structure. The key operation event is more important to the target R&D task than the operation events in the set of operation events other than the key operation event. The target model is obtained by training a large model.

[0319] In the technical solution provided in step S508 of this application, a trained target model is used to identify key operational events that have a significant impact on the target R&D task from the set of operational events. Large-scale model training may involve deep learning techniques, using a large amount of historical data to learn and understand the importance of events. Identifying key operational events helps to quickly locate problems and improve the priority and efficiency of event processing. Key operational events are stored in the corresponding level of data storage in a hierarchical data storage structure. The above steps emphasize a hierarchical data storage strategy. Events of higher importance are stored in storage layers with faster access speeds and higher costs, while other events may be stored in storage layers with lower costs and slower speeds. This strategy ensures that critical information is quickly available when needed, while reducing overall storage costs.

[0320] Step S510: Based on key operation events, determine the product's R&D information and store the R&D information in the corresponding level of the hierarchical data storage structure. The R&D information is used to represent different statistical information generated by the product during the R&D process.

[0321] In the technical solution provided in step S510 of this application, by analyzing key operational events, the system can generate various statistical information about the product during the R&D process, such as task completion rate, developer efficiency, and code quality. Identifying R&D information helps project managers understand the R&D progress, evaluate team performance, and adjust R&D strategies in a timely manner. The R&D information is also stored in the corresponding level of the hierarchical data storage structure. These steps further emphasize hierarchical management of data storage, ensuring that important R&D information can be accessed at any time, while simultaneously reducing the storage and management costs of infrequently used data.

[0322] In this embodiment, by combining hierarchical data storage and machine learning models, efficient management of product information is achieved, which helps to optimize the R&D process, improve R&D efficiency, and reduce the total cost of data storage and management.

[0323] According to an embodiment of this application, an embodiment of a product information processing system is also provided. Figure 6 This is a schematic diagram of an information processing system for a product according to an embodiment of this application, such as... Figure 6 As shown, the product's information processing system 600 may include: a client 601 and a server 602.

[0324] Client 601 is used to transmit a set of task information for the product, wherein the task information in the set describes the R&D tasks that need to be performed in the product's R&D process.

[0325] Server 602 is used to: determine the target task information of the product from the task information set, wherein the target task information describes the target R&D tasks that have not been completed during the product's R&D process; obtain the set of operation events associated with the target task information, wherein the operation events in the operation event set represent the operation behaviors performed on the target task information; identify at least one key operation event from the operation event set using a target model, wherein the key operation event is more important to the target R&D task than the operation events in the operation event set other than the key operation event, and the target model is obtained by training a large model; determine the product's R&D information based on the key operation event, wherein the R&D information represents the different statistical information generated by the product during the R&D process; and transmit the R&D information to the client.

[0326] In this embodiment, an information processing system 600 for a product is provided. A client 602 transmits a set of task information for the product; a server 604 determines the target task information of the product from the task information set; obtains a set of operation events associated with the target task information; uses a target model to identify at least one key operation event from the operation event set; determines the product's R&D information based on the key operation event; and transmits the R&D information to the client. This system improves the efficiency of product information processing and solves the technical problem of low product information processing efficiency.

[0327] 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 this application, such as the data to be verified, are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of the relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0328] The embodiments of this application will be further explained below.

[0329] Currently, in agile product development, requirements tracking and analysis are crucial for on-time product delivery and subsequent iterations.

[0330] The project management process in related technologies has the following problems: information is scattered, with work item information scattered across various products and platforms, making it difficult to aggregate and analyze the various stages of the process; the quality of requirement entries is low, as the above-mentioned project management methods usually rely on manual recording, which is not only time-consuming and laborious but also prone to errors and omissions; and the statistical information is inaccurate, as development precedes recording or forgetting to update progress can lead to untimely requirement flow and inaccurate statistical information.

[0331] The methods described above often involve browser searches, code development, document organization, and meeting minutes during the completion of work items. Information is scattered across various products and platforms, making it difficult to aggregate and analyze these processes. Manual recording is typically necessary, which is time-consuming, labor-intensive, and prone to errors and omissions. Low-quality requirements increase communication costs between different roles. While standardizing requirement templates and requiring R&D teams to describe requirements in a fixed format can improve quality, this also increases pre-development manpower costs. Often, developers develop first and then record progress, or forget to update progress, leading to untimely requirement flow. In fact, developers are very likely to only update progress upon completion of the work item, such as changing statuses from undeveloped to under development to pending release to released, resulting in inaccurate statistics at each stage of the work item.

[0332] Therefore, the technical problem of low information processing efficiency of the product still exists.

[0333] Furthermore, this application provides an intelligent product requirements management and process tracking system (AIgi le). Through automation, it utilizes large-scale modeling technology to track and analyze requirements, supporting automatic creation, workflow, and multi-dimensional statistical analysis of requirements / work items. The AIgi le system is suitable for product development teams requiring project management, requirements tracking, and analysis. It can be integrated into project management tools such as Treloo, as well as IDEs and browser plugins. By recording, aggregating, and analyzing user events, it automatically updates, summarizes, and statistically analyzes work items. AIgi le records user keyboard and mouse events, aggregates active elements, semantic environments, and contextual content, and analyzes and statistically analyzes user behavior through large-scale modeling to automatically update, summarize, and statistically analyze work items. This achieves the technical effect of improving product information processing efficiency and solves the technical problem of low product information processing efficiency.

[0334] The method described in this embodiment will be further described below.

[0335] In the embodiments of this application, Figure 7 This is a flowchart of an AIgi le work item management and analysis process according to an embodiment of this application, such as... Figure 7 As shown, the method may include the following steps:

[0336] Step S701: Collect work items.

[0337] In this embodiment, a list of work items is obtained from a project management tool (such as Jira, Trello, etc.).

[0338] Step S702: Select the work item.

[0339] In this embodiment, the user selects the currently active work item through a list or fuzzy search, and the AIgi le system can also automatically create or switch work items based on user behavior.

[0340] Step S703: Record user events.

[0341] In this embodiment, the user's keyboard and mouse events, including clicks, keyboard inputs, notification events, etc., are recorded (tracked), and the corresponding semantic environment and contextual content are obtained.

[0342] Step S704: Aggregate user events.

[0343] In this embodiment, recorded user events are categorized and aggregated to generate time distribution statistics, timeline organization, and various analysis reports.

[0344] Step S705: Update the work item.

[0345] In this embodiment, the requirement flow automatically updates the status and progress of work items in the project management tool based on aggregated user events.

[0346] Step S706: Summarize the work items.

[0347] In this embodiment, a detailed summary report of the work items is generated, including time distribution, code statistics, AI interaction records, search records, etc.

[0348] Figure 8 This is a flowchart illustrating an AIgi le work item list retrieval and display process according to an embodiment of this application, such as... Figure 8 As shown, the method may include step S801, collecting work items. Work items are obtained from associated project management applications, such as Aone, JIRA, and Trello. The obtained work items can be displayed in the form of a work item list, such as AIGile work items 1-4, and information such as the status, priority, and project to which each AIGile work item belongs can be displayed.

[0349] Figure 9 This is a schematic diagram of an AIgi le user-initiated work item selection interface according to an embodiment of this application, such as... Figure 9 As shown, users can actively select the current task item through a list or fuzzy search. For example, users can enter various information about the task item they want to search for in the search box to filter for task items that meet the above information, and then select the desired task item from the list. For example, they can select the AIgi le task item 1, which is currently in the pending development stage, to process.

[0350] Figure 10 This is a schematic diagram illustrating the automatic creation and switching process of AIgi le system events and user behavior triggered by an embodiment of this application, as shown below. Figure 10 As shown, work items are automatically created or switched based on system events and user behavior. For example, if a user performs a click event in application A and a keyboard event in application B, the user's behavior switches between the two applications, and a work item needs to be switched. You can also use the "√" and "×" on the right to determine whether a new work item needs to be created.

[0351] For example, a user receives a message in the app about a task assignment or problem feedback for a certain project. AIgi le listens for system notification events in the software, identifies the message content and relevant project information, and automatically creates or switches to the corresponding work item.

[0352] For another example, a user opens a project-related webpage in their browser, such as project documentation, a code repository, or a customer feedback page. AIgile intelligently analyzes and automatically creates or switches to work items related to that webpage by recording the user's click events and browser page information.

[0353] As an optional example, users can view or add a meeting or task related to a project within their calendar app. AIgi le automatically creates or switches to the corresponding work item by listening to events in the calendar app, identifying the content and time of the meeting or task.

[0354] Figure 11 This is a flowchart illustrating an AIgi le-based work item creation and switching decision-making process according to an embodiment of this application, such as... Figure 11 As shown, the method may include the following steps:

[0355] Step S1101: Enter the list of work items and the key events for the user within the time window.

[0356] In this embodiment, a Context can be input, which may include a list of work items or key user events within a time window.

[0357] Step S1102: Determine whether to switch work items or create work items.

[0358] In this embodiment, the decision to switch or create a new work item is made based on the context.

[0359] Step S1103: Output the content of the newly created work item.

[0360] In this embodiment, if a new work item is created, the operation creation and work content can be output.

[0361] Step S1104: Output the content of the switching work item.

[0362] In this embodiment, if a work item is switched, the operation switch and work content can be output.

[0363] Step S1105: Display a Toast notification.

[0364] In this embodiment, the user is prompted via a Toast message, and after confirmation, the user can create / switch work items.

[0365] Step S1106: Create / switch work items.

[0366] Figure 12(a) is a flowchart of an AIgile user keyboard and mouse event capture and context information extraction process according to an embodiment of this application. As shown in Figure 12(a), the method may include the following steps:

[0367] Step S1201, User input.

[0368] In this embodiment, before any recording or analysis begins, the user performs actions on the computer, such as writing code in an IDE, searching for information in a browser, or interacting with an AI application. This triggers subsequent event capture and context information extraction processes.

[0369] Step S1202: Capture mouse / keyboard events.

[0370] In this embodiment, the AIgi le system listens for mouse clicks and keyboard input events from the user in different applications, providing raw data for subsequent contextual information extraction and event analysis.

[0371] Step S1203: Obtain keyboard event information.

[0372] In this embodiment, after capturing a keyboard event, the AIgi le system extracts detailed information related to the event, such as the key type, the window and application it was applied to, and the cursor position. This refines user behavior and provides more specific data for subsequent event processing and information logging.

[0373] Step S1204: Process keyboard event information using the AIgi le system.

[0374] In this embodiment, the AIgi le system uses Large Model (LLM) technology to analyze keyboard event information and identify the intent behind user behavior, such as coding, searching, and AI interaction. This transforms raw keyboard and mouse events into key event descriptions, facilitating subsequent analysis and application.

[0375] Step S1205: Analyze and record relevant information.

[0376] In this embodiment, based on the analysis results of the large model, key information about user behavior is recorded. This information may include timestamps, event types, contextual content, and associated work items. Storing user behavior data provides a data foundation for subsequent multi-dimensional statistical analysis and work item management.

[0377] In summary, the above process begins with user input, captures and analyzes keyboard and mouse events, utilizes AIgi le's large model technology to identify the user's intent, and then extracts and records contextual information to provide data for subsequent analysis and management. These steps together constitute AIgi le's ability to automatically track and analyze user activities during the R&D process, providing refined requirements management, efficient workflow, and multi-dimensional statistical analysis capabilities.

[0378] Figure 12(b) is a schematic diagram of the specific information content included in a click event information according to an embodiment of this application. As shown in Figure 12(b), the click event information may include coordinates, window, application, page, and element. The click event can be triggered by the user actively clicking the mouse.

[0379] Figure 12(c) is a schematic diagram of the specific information content included in keyboard event information according to an embodiment of this application. As shown in Figure 12(c), keyboard events can be triggered by the user actively clicking the keyboard. Keyboard event information may include keys, windows, applications, pages, and elements.

[0380] Notification events can include meeting notifications, text message notifications, and email notifications. Meeting notifications can record information such as the meeting time and participants. Text message notifications can record the message content and sender. Email notifications can record information such as the email subject, content, and sender.

[0381] AIgi implements the function of recording user keyboard and mouse events in the operating system. Event recording can be achieved through the accessibility interface or the window programming interface.

[0382] Optionally, for click events, the context information of the click event can be obtained through the accessibility interface, including coordinates, window, application, page, and element. Context information is obtained based on semantic grouping. For example, if a browser element is clicked, a list of elements at the same level as that element is retrieved. If an option in a menu is clicked, other options in the parent and sibling layouts of the menu are retrieved.

[0383] Optionally, keyboard events are handled through the accessibility interface, obtaining the keyboard event type, target key, affected window, application, page, and element. Context information is obtained through the currently active application, cursor position, and listening for specific key combinations. For special events, AIgile identifies specific shortcut keys and populates context information based on changes in content before and after the event. For example, if a user adopts code completion suggestions from a code completion tool (Copi lot) during coding, AIgile will determine whether the user adopted the completion based on changes in the content of the affected element after the event; if so, it will record the corresponding code snippet.

[0384] Figure 13 This is a schematic diagram illustrating an AIgi le MacOS notification center listening and notification event capture process according to an embodiment of this application, as shown below. Figure 13 As shown, the Accessibility Interface (AMI) is used to listen for layout changes in the macOS Notification Center application and retrieve application notifications.

[0385] Figure 14This is a flowchart illustrating an AIgi le user event timeline construction and key event generation process according to an embodiment of this application, such as... Figure 14 As shown, click events, keyboard events, and notification events are aggregated into a user event timeline in chronological order. User events can include actions such as clicking in application a, switching to application b for code development, and receiving system message pushes from application c during development.

[0386] Figure 15 This is a schematic diagram illustrating the process of identifying and recording key events in an AIgi le large model according to an embodiment of this application, such as... Figure 15 As shown, for the event details on the timeline, the corresponding key events are identified and recorded through a large model.

[0387] Figure 16 This is a flowchart of a process for generating key events from a large model according to an embodiment of this application, such as... Figure 16 As shown, the method may include:

[0388] Step S1601: Input the work item list, the current work item, the user event list and context of the user within the time window.

[0389] In this embodiment, the input includes a list of work items, the current work item, a list of user events recorded within a specific time window, and contextual descriptions. The AIgi le system receives key data inputs from a hierarchical storage architecture (ODS, DWD, DWS, and ADS), including the current work item list, the user's currently ongoing work item, a list of user events recorded within a specific time window, and contextual information related to the events. This provides complete and specific data input for generating key events for the large model, ensuring the accuracy and relevance of event generation.

[0390] Step S1602: Generate key event descriptions.

[0391] In this embodiment, the large model analyzes user events and context within a time window to generate descriptions and explanations of user behavior, such as search events, code events, and AI interaction events. This transforms raw user events into meaningful descriptions, facilitating subsequent statistical analysis and work item management.

[0392] Step S1603: Determine the operation type.

[0393] In this embodiment, based on the generated key event description, the AIgi le system determines whether the operation belongs to one of the key events: creation, update, or deletion. The processing flow for the key events is then determined to guide the subsequent output steps.

[0394] Step S1604: Output the newly created critical event.

[0395] In this embodiment, if the determination result is "new", the AIgi le system outputs detailed information about the newly created key event, including the event's time, type, context, and other relevant attributes. Creating a new key event record supplements the event flow of the work item, providing new data for subsequent statistical analysis and work item status updates.

[0396] Step S1605: Output the update key events.

[0397] In this embodiment, if the determination result is an update, the AIgi le system outputs detailed information about the updated critical events, updates existing critical event records, and reflects changes in user behavior or work item status. Maintaining the up-to-date state of critical event records ensures the real-time and accurate updating of work item progress and status.

[0398] Step S1606: Output the key event for deletion.

[0399] In this embodiment, if the determination result is deletion, the AIgi le system outputs the operation to delete the critical event, removing critical event records that are no longer relevant or outdated. This maintains the purity and relevance of the critical event list, preventing invalid events from interfering with subsequent statistical analysis and work item management.

[0400] Figure 17 This is a schematic diagram illustrating an AIgi le search event generation and index building process according to an embodiment of this application, as shown below. Figure 17 The diagram illustrates the user's search process in a browser and the relevant information recorded by the system. The user enters search keywords, such as "devops methodology," into the browser. The AIgile system records the time of the search event, such as "2024 / 10 / 01 20:00:00,666." The AIgile system identifies the search engine used by the user. After the search is completed, the user browses the search results page, and the AIgile system records the URL of that page, such as "https: / / XXXXXXX." The user clicks a link in the search results, and the AIgile system records the clicked search result link. The entire search event flow can be executed in the following steps: Step S1701, execute the search; Step S1702, record search metadata; Step S1703, click a search result item; Step S1704, record the metadata of the search result item. Through the above process, the AIgile system not only records the time, keywords, search engine used, and result page URL, but also specifically records the search result link clicked by the user. Figure 17The system presents key attributes of search events, including search time, keywords, search engine, search results page, and user click behavior on search results. This information collectively constitutes a detailed search event record, providing foundational data for subsequent multi-dimensional statistical analysis. The AIgi le system can utilize these search event records to generate time distributions of user search behavior, statistical analysis of search keywords, and analysis of user interests and preferences based on search result click behavior, providing valuable insights for task management.

[0401] Figure 18 This is a schematic diagram of an AIgi le code event generation and statistical analysis process according to an embodiment of this application, such as... Figure 18 As shown, this illustrates a typical flow of code events in the software development process and records key information. Functional code displays the code portion of a feature or function being developed. The user opens or activates the Integrated Development Environment (IDE). The user creates or opens a new file. The user writes code, including functional code, unit tests, code refactoring, and copy lot-assisted coding. Combining project structure and user events, a large model determines the type of code written. During user coding, AIgile can perform time statistics, code statistics, and file statistics. The user tests the code. The user submits code changes. Code review is performed. Pull requests are merged. The code is deployed to the staging environment. Testing is conducted in the staging environment. The code is deployed to the production environment.

[0402] Optionally, such as Figure 18 As shown in the code event flow example, the user starts coding from Feature Code. It records newly added, modified, or deleted code snippets. It records the execution status of unit tests. It records the details of code refactoring. It records the time of code commits.

[0403] Figure 19 This is a schematic diagram illustrating the distribution of AIgi le code development time and the proportion of Copi lot interaction time according to an embodiment of this application. Figure 19 As shown, the statistics for different user coding activities (functional code, unit testing, code refactoring, and Copi lot-assisted coding) include the time distribution for each category and the percentage of time spent interacting with Copi lot.

[0404] Figure 20 This is a schematic diagram illustrating the code distribution and copy lot adoption rate analysis of an AIgi le code development activity according to an embodiment of this application, as shown below. Figure 20 As shown, the statistics for different user coding activities (functional code, unit testing, code refactoring, and Copi lot-assisted coding) include the cumulative number of lines of code and the percentage of lines that adopted Copi lot coding suggestions.

[0405] Figure 21 This is a schematic diagram illustrating the distribution of AIgi le code submission files and the statistics of Copi lot acceptance rate according to an embodiment of this application, as shown below. Figure 21 As shown, this displays statistics on the code files included in a single user code commit, including the number of lines added, modified, and deleted, as well as the percentage of lines that adopted Copi Lot coding suggestions. The information is retrieved from commit events, providing commit-level statistics distributed by time and line count according to code files.

[0406] Figure 22 This is a schematic diagram illustrating AIgi le code event generation and multi-dimensional statistics according to an embodiment of this application, such as... Figure 22 As shown, the final code events are generated by combining the above information, including commit information, development timeline, file list, and statistical information (time distribution, code distribution, file distribution).

[0407] Optionally, AI events can be interactive operations performed by the user and the large model. Time can be the time when the AI ​​interaction occurs. Interaction types can include asking questions, answering questions, code completion, etc. Input content can be content entered by the user. Output content can be the output content of the large model. Interaction duration can be the duration of the interaction. Response events can be the response time of the large model.

[0408] Figure 23 This is a schematic diagram illustrating an AIgi le user interaction event analysis and statistical process according to an embodiment of this application, such as... Figure 23The diagram illustrates the process of user interaction with a large AI model and the key information recorded. User actions involve users needing AI assistance while coding or performing other tasks. This interaction is triggered by inputting questions or requests (such as code completion or problem-solving). Interaction logs include time, interaction type, input content, output content, interaction time, and response time. The flow of user-AI interaction is illustrated using icons and arrows: the user first switches to the AI ​​tool, inputs a question or request, the AI ​​receives and processes the request, and finally outputs the result. The AIgi le system records the time, interaction type, input and output content, interaction duration, and AI response time throughout this process. Specific interaction examples include the user's question, the AI's response, the relevant code file name, the interaction time, and its duration. The importance of recording the time, type, content, and performance metrics (interaction duration, response time) of AI interactions is highlighted. This information helps the AIgi le system analyze user frequency, efficiency, and preferences for AI tools, providing a basis for optimizing AI functions and improving user experience. It also provides a new dimension for work item management and statistical analysis: the efficiency and contribution of AI-assisted development.

[0409] Optionally, such as Figure 23 As shown, users can ask questions by copying and pasting content, or by writing their own. The large model can return answers either in a streaming or one-time manner, and users can copy the returned results.

[0410] Figure 24 This is a flowchart of an AIgi le user-large model interaction method according to an embodiment of this application, such as... Figure 24 As shown, the user's workflow for using the AI ​​application involves the following steps: When the user switches applications, from one application to the large model application, the application information is recorded. When the user sends a prompt, the model version and prompt content are recorded. When the large model outputs data, the output content is recorded. When the user switches applications again, from the large model application to another application, the application information is recorded.

[0411] Figure 25 This is a schematic diagram of an AIgi le user-large model interaction loop and event recording process according to an embodiment of this application, as shown below. Figure 25 As shown, through this process, users can switch to the large model while writing the code file h.java, raise code questions, and modify the code based on the results.

[0412] Figure 26 This is a schematic diagram illustrating an AIgi le code writing and intelligent consultation event generation process according to an embodiment of this application, such as... Figure 26As shown, AIgi le can generate corresponding AI events through the above process.

[0413] Figure 27 This is a flowchart of an AIgi le-based automated update process for work items based on key events, according to an embodiment of this application. Figure 27 As shown, the method may include the following steps:

[0414] Step S2701: Input the work item list, the current work item, the user event list within the time window, the context, and the key events.

[0415] In this embodiment, the system obtains a list of work items in the current project, information on the work item the user is currently processing, user events recorded over a period of time (including clicks, keyboard input, notification reception, etc.), the context of the events (such as application, page, element, etc.), and key events that have already been generated (search, code operation, AI interaction, etc.). These inputs constitute the basic data for the AIgi le system to determine and update work items.

[0416] Step S2702: Determine whether to update the work item.

[0417] In this embodiment, the current input, including key events and contextual information, is analyzed based on a large model to determine whether the user's behavior affects the progress, status, or description of the current work item and requires an update. If the determination result is "yes," the update operation begins; if the determination result is "no," the process ends, and no update is performed on the work item.

[0418] Step S2703: Output the update operation and content.

[0419] In this embodiment, once it is determined that a work item needs to be updated, the AIgi le system will output specific update operations and content, including but not limited to updating the status of the work item (e.g., from "pending development" to "under development"), modifying the description of the work item to reflect the latest progress, and adjusting the priority of the work item. The output update content provides specific instructions for subsequent update operations.

[0420] Step S2704: Determine the output content.

[0421] In this embodiment, it can be determined whether the output content is progress, process record, or remarks.

[0422] Step S2705: Update progress.

[0423] In this embodiment, the impact of critical events on the work item is assessed before updating the work item to ensure that the update reflects the actual progress and the user's actual operations. If there is a discrepancy between the analysis results of critical events and the current state of the work item, the AIgi le system can adjust the update schedule to more accurately reflect the actual state of the work item.

[0424] Step S2706: Update the process log.

[0425] In this embodiment, while updating a work item, the AIgi le system can record the update process, including the reason for the update, the status changes before and after the update, and the update time. These records are valuable for subsequent requirements tracking and project management, helping to analyze the progress and efficiency of work items.

[0426] Step S2707, update the remarks.

[0427] In this embodiment, when updating the status or description of a work item, notes can be added or modified. These notes can include additional explanations from the user during the operation, problems encountered, solutions, etc., providing richer contextual information for the work item and facilitating communication and collaboration among team members.

[0428] Figure 28 This is a schematic diagram of a user event analysis based on hierarchical storage according to an embodiment of this application, such as... Figure 28 As shown, AIgi uses a hierarchical storage structure for recording, aggregating, and analyzing user events. ODS stores raw user events, including click events, keyboard events, and notification events. DWD stores key events extracted from raw user events, including search events, code events, and AI events. DWS stores aggregated data of key events, along with summary metrics built for each key event, supporting analysis and querying from multiple dimensions. ADS stores highly summarized analytical data at different time granularities, including time distribution, work item progress, work reports, AI interaction history, search history, and other statistical analysis information, supporting interactive queries.

[0429] Figure 29 This is a flowchart of a model training method according to an embodiment of this application, such as... Figure 29 As shown, the model training method may include the following steps:

[0430] Step S2901, pre-train the model.

[0431] In this embodiment, we start with a large model that has already been pre-trained.

[0432] Optionally, pre-training refers to training a model on a large-scale, general dataset, enabling it to learn a wide range of data features and patterns. Pre-trained models typically use unsupervised learning. The purpose of pre-training is to give the model well-initialized weights, allowing it to converge faster in subsequent fine-tuning or specialized training while avoiding overfitting. In the software development field, pre-trained models can be trained on general-purpose data such as code repositories, documentation, or error logs, giving the model the basic ability to understand and process code or text.

[0433] Step S2902: Collect data in the software development field.

[0434] In this embodiment, domain-specific data related to software development is collected.

[0435] Optionally, data specific to the software development domain can be collected. This data may include code snippets, comments, documentation, error messages, user feedback, etc. The collected data will be used for model fine-tuning and performance optimization, enabling the model to better adapt to software development scenarios and solve specific problems such as code generation, error detection, and documentation understanding.

[0436] Step S2903, data preprocessing.

[0437] In this embodiment, the collected data undergoes preprocessing such as cleaning and labeling.

[0438] Optionally, data preprocessing is a crucial step in machine learning and deep learning, including data cleaning (removing noise and irrelevant information), data formatting (such as converting text into numeric vectors), and data augmentation (improving the model's generalization ability). For data in the software development field, preprocessing may also include keyword extraction, code structuring, and error annotation to ensure that the model can learn effective features from the data.

[0439] Step S2904, domain adaptation fine-tuning.

[0440] In this embodiment, the pre-trained model is fine-tuned using pre-processed data to adapt it to tasks in the software development domain.

[0441] Optionally, domain-adaptive fine-tuning refers to further training the model using collected domain-specific data, building upon the pre-trained model. The aim of this operation is to enable the model to better understand the specific needs and details within the domain, thereby improving its performance on software development tasks. During fine-tuning, some layers of the model may be frozen, with training only on the last few layers and any newly added layers, to reduce computational costs and maintain the model's generality.

[0442] Step S2905: Evaluate model performance.

[0443] In this embodiment, the performance of the fine-tuned model on a specific task is evaluated.

[0444] Optionally, model performance evaluation involves checking the model's performance using a reserved test dataset after fine-tuning. This may include calculating metrics such as accuracy, recall, and score, as well as analyzing the model's behavior in specific scenarios. Performance evaluation is a crucial step in ensuring that the model can effectively solve problems in the software development domain.

[0445] Step S2906: Determine whether the performance meets the requirements.

[0446] In this embodiment, based on the performance evaluation results, it is determined whether the model has met the expected performance standards. If not, step S2908 can be executed to adjust the model structure, parameters, or add more data for retraining. Conversely, step S2907 can be executed.

[0447] Step S2907: Deploy the model.

[0448] In this embodiment, the trained model is deployed to a real-world application.

[0449] Optionally, once the model's performance meets the requirements, it can be deployed to a real software development environment. This can include integration into development tools, code editor workflows, etc., to provide automated functions such as code completion, error detection, and design suggestions, thereby improving the efficiency and quality of software development.

[0450] Step S2908: Adjust hyperparameters / increase data.

[0451] In this embodiment, if the performance requirements are not met, the hyperparameters can be adjusted or more data can be added, and then the model can be fine-tuned.

[0452] Optionally, if the model performs poorly in software development applications, it may be necessary to adjust the model's hyperparameters (such as learning rate, batch size, etc.) or add more domain-specific data for retraining. Hyperparameter tuning and data augmentation are methods to improve model performance, enabling the model to better fit the data distribution in the software development domain.

[0453] Figure 30 This is a flowchart illustrating a method for incremental training and continuous learning based on user feedback, according to an embodiment of this application. Figure 30 As shown, the method may include the following steps:

[0454] Step S3001: The user uses the system.

[0455] In this embodiment, the user uses the system in actual work.

[0456] Alternatively, users may actually use the system in their daily work, such as as software development tools, code editors, or automated testing platforms. User actions and interactions form the basis for collecting feedback information, which can reveal the system's performance, strengths, and weaknesses in real-world scenarios.

[0457] Step S3002: Generate operation events and feedback.

[0458] In this embodiment, the system records user operation events and collects user feedback information.

[0459] Optionally, the system automatically records user action events, including but not limited to user interface clicks, code submissions, and search queries. Simultaneously, the system collects feedback information provided directly or indirectly by users, such as feedback provided through surveys, customer service, social media, or directly within the system. This feedback can be related to various aspects such as system functionality, response speed, and user interface experience.

[0460] Step S3003: Collect user feedback data.

[0461] In this embodiment, user feedback data is collected.

[0462] Optionally, the collected operation events and user feedback information are aggregated into a feedback dataset. This dataset typically contains rich information that can be used to analyze user behavior and needs, helping to identify areas for system improvement.

[0463] Step S3004, data preprocessing.

[0464] In this embodiment, the collected user feedback data is preprocessed.

[0465] Optionally, data preprocessing is a key step in the machine learning process. It includes operations such as cleaning data (removing noise and irrelevant information), data formatting (e.g., converting text data into numeric vectors), and data standardization or normalization to ensure data quality and consistency, providing clean and structured data for model training.

[0466] Step S3005: Merge with the existing dataset.

[0467] In this embodiment, the preprocessed user feedback data is merged with the existing training dataset.

[0468] Optionally, the preprocessed user feedback data can be merged with the existing training dataset to form a larger dataset. These steps help the model better understand user needs and behavioral patterns, enhancing its generalization ability.

[0469] Step S3006: Incrementally train the model.

[0470] In this embodiment, the merged dataset is used to incrementally train the existing model.

[0471] Optionally, the existing model can be incrementally trained using the merged dataset. This aims to preserve the model's existing knowledge while fine-tuning it with new data to adapt to new user feedback and needs. Incremental learning avoids training the model from scratch, saving time and computational resources.

[0472] Step S3007: Evaluate model performance.

[0473] In this embodiment, the performance of the model after incremental training is evaluated.

[0474] Optionally, after incremental training, it is necessary to evaluate the model's ability to handle new data and new requirements. Performance evaluation may include metrics such as accuracy, response speed, and user satisfaction to ensure that the model improvements are effective.

[0475] Step S3008: Determine if performance has improved.

[0476] In this embodiment, if the model performance is improved, step S3009 can be executed; otherwise, step S3010 can be executed.

[0477] Optionally, if the model performance is significantly improved, i.e., reaches or exceeds the performance evaluation threshold, it indicates that the use of user feedback is effective; if there is no improvement or the performance deteriorates, it may be necessary to re-examine the training process.

[0478] Step S3009: Update the deployment model.

[0479] In this embodiment, the performance-enhanced model update is deployed to the actual application.

[0480] Optionally, after the model performance is improved, the new model version can be deployed to the production environment to provide users with optimized services. This deployment process may include steps such as model version control and system integration testing to ensure that the model update does not negatively impact existing systems.

[0481] Step S3010: Adjust hyperparameters / improve data.

[0482] In this embodiment, if the performance improvement requirements are not met, the hyperparameters can be adjusted or the data improved, and then incremental training can be performed again.

[0483] Alternatively, if the model performance does not achieve the expected improvement, it is necessary to return to the earlier stages of model training, adjust the model's hyperparameters (such as learning rate, batch size, optimizer, etc.), or collect more and higher quality data to further train the model.

[0484] The target large model in this application embodiment supports two deployment methods: cloud deployment and terminal deployment.

[0485] For example, the advantages of deploying a large target model on the cloud (cloud services) are that computing resources can be flexibly adjusted according to demand, high concurrency requests can be easily handled, and cloud service providers are responsible for hardware and software maintenance, making maintenance simple. However, the disadvantages of the above deployment method are that performance is greatly affected by network conditions, there may be data security and privacy issues, and long-term use may result in high cloud service costs.

[0486] For another example, the advantages of deploying a large target model on a terminal include the elimination of network transmission, faster response times, local data processing, enhanced security, and the ability to operate without a network connection. However, the disadvantages of this deployment method include limited computing and storage resources on the terminal device, the need for manual or over-the-air (OTA) updates for the model, and the potential for increased power consumption due to continuous operation. If the large target model is deployed on a terminal, it can be referred to as a terminal-based large model.

[0487] It should be noted that the above deployment methods for the target large model are only illustrative examples and are not subject to specific restrictions. Each of the above deployment methods has its own advantages and disadvantages, and the choice can be made according to the actual situation and user needs.

[0488] In this embodiment, to determine the R&D information of a product, a task information set for the product can be obtained. This task information set allows analysis of the R&D tasks the product needs to perform in the R&D process, identifying unfinished target R&D tasks and their corresponding target task information. The operational behaviors required to complete these target R&D tasks can be determined, resulting in a set of operational events corresponding to the target task information. This set of operational events can be input into a target model, which analyzes each event, identifying those with high importance to the target R&D task as key operational events. The product's R&D information can then be determined based on these key operational events. In this embodiment, AIgi le solves the problem of low product information processing efficiency in existing technologies through automated work item management, operational behavior recording, and intelligent analysis of large models. It not only quickly and accurately identifies and manages R&D tasks but also deeply analyzes operational behaviors, automatically extracts key operational events, and generates rich statistical reports, thus providing more efficient and intelligent decision support for project management. This achieves the technical effect of improving product information processing efficiency and solves the technical problem of low product information processing efficiency.

[0489] According to embodiments of this application, a method for implementing the above is also provided. Figure 2 The information processing method for the product shown is a product information processing device.

[0490] Figure 31 This is a schematic diagram of an information processing device for a product according to an embodiment of this application, such as... Figure 31 As shown, the information processing device 3100 of the product may include: a first acquisition unit 3102, a first determination unit 3104, a second acquisition unit 3106, a first identification unit 3108, and a second determination unit 3110.

[0491] The first acquisition unit 3102 is used to acquire the task information set of the product.

[0492] The first determining unit 3104 is used to determine the target task information of the product from the task information set.

[0493] The second acquisition unit 3106 is used to acquire the set of operation events associated with the target task information.

[0494] The first identification unit 3108 is used to identify at least one key operation event from the set of operation events using the target model.

[0495] The second determining unit 3110 is used to determine the product's R&D information based on key operational events.

[0496] Here, the first acquisition unit 3102, the first determination unit 3104, the second acquisition unit 3106, the first identification unit 3108, and the second determination unit 3110 correspond to steps S202 to S210 in the embodiments. The five units and the corresponding steps implement the same instances and application scenarios, but are not limited to the content disclosed in the above embodiments. It should be noted that the above units can be hardware components or software components stored in memory (e.g., memory 3504) and processed by one or more processors (e.g., processor 3502). The above units can also be part of a device and run in the computer terminal provided in the following embodiments.

[0497] According to embodiments of this application, a method for implementing the above is also provided. Figure 3 The information processing method for the product shown is a product information processing device.

[0498] Figure 32 This is a schematic diagram of an information processing device for a product according to an embodiment of this application, such as... Figure 32 As shown, the information processing device 3200 of the product may include: a third acquisition unit 3202, a third determination unit 3204, a fourth acquisition unit 3206, a second identification unit 3208, and a fourth determination unit 3210.

[0499] The third acquisition unit 3202 is used to acquire the task information set of the smart product.

[0500] The third determining unit 3204 is used to determine the target task information of the intelligent product from the task information set.

[0501] The fourth acquisition unit 3206 is used to acquire the set of operation events associated with the target task information.

[0502] The second identification unit 3208 is used to identify at least one key operation event from the set of operation events using the target model.

[0503] The fourth determining unit 3210 is used to determine the R&D information of intelligent products based on key operational events.

[0504] It should be noted that the third acquisition unit 3202, the third determination unit 3204, the fourth acquisition unit 3206, the second identification unit 3208, and the fourth determination unit 3210 correspond to steps S302 to S310 in the above embodiments. The five units and their corresponding steps implement the same instances and application scenarios, but are not limited to the content disclosed in the above embodiments. It should also be noted that the above units can be hardware or software components stored in a memory (e.g., memory 3504) and processed by one or more processors (e.g., processor 3502). These units can also be part of a device and run in the computer terminal provided in the following embodiments.

[0505] According to embodiments of this application, a method for implementing the above is also provided. Figure 4 The information processing method for the product shown is a product information processing device.

[0506] Figure 33 This is a schematic diagram of an information processing device for another product according to an embodiment of this application, such as... Figure 33 As shown, the information processing device 3300 of the product may include: a first display unit 3302 and a third display unit 3304.

[0507] The first display unit 3302 is used to display the product's task information set on the operation interface.

[0508] The third display unit 3304 is used to respond to information processing commands applied to the operation interface and display product research and development information on the operation interface.

[0509] Here, the first display unit 3302 and the third display unit 3304 correspond to steps S402 to S404 in the above embodiments. The two units and the corresponding steps implement the same instances and application scenarios, but are not limited to the content disclosed in the above embodiments. It should be noted that the above units can be hardware components or software components stored in memory (e.g., memory 3504) and processed by one or more processors (e.g., processor 3502). The above units can also be part of a device and run in the computer terminal provided in the following embodiments.

[0510] According to embodiments of this application, a method for implementing the above is also provided. Figure 5 The product information storage method shown is a product information storage device.

[0511] Figure 34 This is a schematic diagram of an information storage device for a product according to an embodiment of this application, such as... Figure 34 As shown, the information storage device 3400 of the product may include: a fifth acquisition unit 3402, a fifth determination unit 3404, a query unit 3406, a first storage unit 3408, and a second storage unit 3410.

[0512] The fifth acquisition unit 3402 is used to acquire the product's task information set.

[0513] The fifth determining unit 3404 is used to determine the target task information of the product from the task information set.

[0514] The query unit 3406 is used to query the set of operation events associated with the target task information from the hierarchical data storage structure.

[0515] The first storage unit 3408 is used to identify at least one key operation event from the set of operation events using the target model, and to store the key operation event in the data storage structure of the corresponding level in the hierarchical data storage structure.

[0516] The second storage unit 3410 is used to determine the product's R&D information based on key operational events and store the R&D information in the corresponding level of the hierarchical data storage structure.

[0517] Here, the fifth acquisition unit 3402, the fifth determination unit 3404, the query unit 3406, the first storage unit 3408, and the second storage unit 3410 correspond to steps S502 to S510 in the above embodiments. The five units and the corresponding steps implement the same instances and application scenarios, but are not limited to the content disclosed in the above embodiments. It should be noted that the above units can be hardware components or software components stored in memory (e.g., memory 3504) and processed by one or more processors (e.g., processor 3502). The above units can also be part of a device and run in the computer terminal provided in the following embodiments.

[0518] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0519] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, they can also be implemented by hardware. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0520] In the information processing device of this product, if it is necessary to determine the R&D information of a certain product, the task information set of the product can be obtained. Through the task information set, the R&D tasks that the product needs to perform in the R&D process can be analyzed, and the target R&D tasks that the product has not yet completed in the R&D process can be identified, along with the target task information corresponding to those tasks. The operational behaviors required to complete the target R&D tasks can be determined, resulting in a set of operational events corresponding to the target task information. This set of operational events can be input into a target large model, which analyzes each operational event in the set, identifying those with high importance to the target R&D tasks as key operational events. The product's R&D information can then be determined based on these key operational events. In this embodiment, AIgi le solves the problem of low product information processing efficiency in the prior art through automated work item management, operational behavior recording, and intelligent analysis of a large model. It can not only quickly and accurately identify and manage R&D tasks but also deeply analyze operational behaviors, automatically extract key operational events, and generate rich statistical reports, thus providing more efficient and intelligent decision support for project management. This achieves the technical effect of improving the product's information processing efficiency and solves the technical problem of low product information processing efficiency.

[0521] It should be noted that the preferred embodiments involved in the above embodiments of this application are the same as the solutions, application scenarios and implementation processes provided in the above embodiments, but are not limited to the solutions provided in the above embodiments.

[0522] Embodiments of this application may provide a computing device. Figure 35 This is a structural block diagram of a computing device according to an embodiment of this application. Figure 35 As shown, the computing device 3500 may include: one or more (only one is shown in the figure) processors 3502, memory 3504, memory controller, and peripheral interfaces. The processor 3502 can be connected to the target large model for calling the target large model to process product information. It should be noted that the aforementioned target large model can be deployed in the cloud or on a terminal; no specific limitation is made here.

[0523] The aforementioned computing device can be understood as an integrated smart terminal, including but not limited to servers, desktop computers, PCs (Personal Computers), all-in-one model machines, etc., and the computing device may have the model described in the above embodiments of this application pre-installed.

[0524] Specifically, this computing device can pre-install various types of models, including but not limited to models in natural language processing, visual processing, speech processing, code processing, and multimodal task processing, thus providing diverse model selection. In different product forms, this computing device can support one or more model usage methods, including but not limited to model training, model invocation, model fine-tuning, model deployment, model inference, and application. In some product forms, this computing device also supports model management, including but not limited to multi-type model management (supporting the management of discriminative, generative, and other types of models), model version control (supporting the control of different model versions), and model evaluation (evaluating model performance and effectiveness based on model evaluation tools). In other product forms, this computing device can also create applications based on models, providing API calling capabilities, allowing models to be called into created applications through API interfaces, and providing application management tools for application management and monitoring.

[0525] Furthermore, the computing device can also include data management (supporting the creation and management of model tuning datasets), a training center (providing abundant training resources to help users learn and master AI technology), and basic control capabilities (providing enterprise-level basic control capabilities to ensure the security and efficient operation of the system). Through the above functions, it provides a comprehensive and integrated AI research and development, training, deployment, and application device.

[0526] The memory can be used to store software programs and modules, such as the program instructions / modules corresponding to the methods and apparatus in the embodiments of this application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, thereby implementing the methods in the above embodiments. The memory may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include memory remotely located relative to the processor, and these remote memories can be connected to terminal A via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0527] The processor can invoke an executable program stored in memory via a transmission device to execute the method described in any of the above embodiments.

[0528] Those skilled in the art will understand that, Figure 35The structure shown is for illustrative purposes only. The computing device can also be a smartphone (e.g., an Android phone, an iOS phone), a tablet computer, a PDA, a mobile internet device (MID), a tablet computer, or other terminal devices. This diagram does not limit the structure of the computing device described above. For example, computing device 3500 may include more or fewer components (such as network interfaces, display devices, etc.) than shown in this diagram, or may have a different configuration than shown in this diagram.

[0529] Embodiments of this application may provide an electronic device. Figure 36 This is a structural block diagram of an electronic device according to an embodiment of this application. Figure 36 As shown, the electronic device may include: an input / output device 3612; a memory 3614; and a processor 3616, wherein the processor 3616 is connected to the input / output device 3612 and the memory 3614 via a bus 3618. The processor 3616 can be connected to a target large model for calling the target large model to perform product information processing. It should be noted that the aforementioned target large model can be deployed in the cloud or on a terminal; no specific limitation is made here.

[0530] The memory can be used to store software programs and modules, such as the program instructions / modules corresponding to the methods and apparatus in the embodiments of this application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, thereby implementing the methods in the above embodiments. The memory may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include memory remotely located relative to the processor, and these remote memories can be connected to the terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0531] The processor can invoke an executable program stored in memory via a transmission device to execute the method described in any of the above embodiments.

[0532] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing the hardware related to the terminal device. The program can be stored in a computer-readable storage medium, which may include: flash drive, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0533] Optionally, in this embodiment, the storage medium may be located in a computing device.

[0534] Optionally, in this embodiment, the computer-readable storage medium is configured to store an executable program, which, when the executable program is running, controls the device where the computer-readable storage medium is located to execute the method described in any of the above embodiments.

[0535] Embodiments of this application also provide a computer program product. Optionally, in this embodiment, the computer program product may include a computer program that, when executed by a processor, implements the methods provided in the embodiments described above.

[0536] Embodiments of this application also provide a computer program product. Optionally, the computer program product may include a non-volatile computer-readable storage medium, which can be used to store a computer program that, when executed by a processor, implements the method provided in the above embodiments.

[0537] Embodiments of this application also provide a computer program. Optionally, in this embodiment, when the computer program is executed by a processor, it implements the method provided in the above embodiments.

[0538] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0539] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual couplings, direct couplings, or communication connections may be through some interfaces; indirect couplings or communication connections between units or modules may be electrical or other forms.

[0540] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0541] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0542] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, read-only memory, random access memory, portable hard drive, magnetic disk, or optical disk.

[0543] The above are merely preferred embodiments of this application. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for processing information about a product, characterized in that, include: Obtain a set of task information for the product, wherein the task information in the set of task information is used to describe the research and development tasks that need to be performed in the research and development process of the product. From the task information set, the target task information of the product is determined, wherein the target task information is used to describe the target R&D tasks that were not completed during the R&D process of the product; Obtain the set of operation events associated with the target task information, wherein the operation events in the set of operation events are used to represent the operation behavior performed on the target task information; Using a target model, at least one key operation event is identified from the set of operation events, wherein the importance of the key operation event to the target R&D task is greater than the importance of the operation events in the set of operation events other than the key operation event to the target R&D task, and the target model is obtained by training a large model; Based on the key operational events, the R&D information of the product is determined, wherein the R&D information is used to represent different statistical information generated by the product in the R&D process.

2. The method according to claim 1, characterized in that, Using the target model, at least one key operation event is identified from the set of operation events, including: The operation events in the operation event set are aggregated to obtain the event aggregation result; Using the target model, the key operational events are identified from the event aggregation results.

3. The method according to claim 2, characterized in that, Using the target model, the key operational events are identified from the event aggregation results, including: The task information set, the target task information, the operation event set, and the context information of the target task information are used to generate event prompt information, wherein the event prompt information is used to describe the key operation event; Based on the event prompts for the key operations, the target model is guided to determine the operation type, and the key operation events under the operation type are identified from the event aggregation results.

4. The method according to claim 2, characterized in that, The operation events in the operation event set are aggregated to obtain the event aggregation result, including: The operation events in the operation event set are aggregated onto a timeline according to the operation time corresponding to each operation event to obtain the event aggregation result.

5. The method according to claim 1, characterized in that, The critical operational events include at least one of the following: Search operation event, wherein the search operation event is used to represent the search operation behavior performed on the target task information; Code operation event, wherein the code operation event is used to represent the code operation behavior performed on the target task information; Interactive operation events, wherein the interactive operation events are used to represent interactive operation behaviors performed on the target task information using the target model.

6. The method according to claim 5, characterized in that, Based on the aforementioned key operational events, the product's R&D information is determined, including: When the key operation event is the code operation event, the R&D information of the code operation event is determined in different dimensions; If the key operation event is the interactive operation event, determine the R&D information associated with the target model; In the case where the key operation event is the search operation event, the search metadata of the product is determined, wherein the search metadata is used to represent the basic data required to perform the search operation event.

7. The method according to claim 6, characterized in that, The code operation events, under the different dimensions of the R&D information, include at least one of the following: The time information required for the code operation event in the time dimension; The code operation events are the code writing results in the code dimension; The code operation event requires the code file at the file level.

8. The method according to claim 6, characterized in that, The R&D information associated with the target model includes at least one of the following: Application information associated with the target model, wherein the application information is used to indicate information on switching from running the target model to running the application, or the application information is used to indicate information on switching from running the application to running the target model; The output of the target model; Prompt information for input into the target model.

9. The method according to claim 1, characterized in that, From the task information set, the target task information of the product is determined, including: In response to a selection operation on the task information set, the target task information is selected from the task information set; In response to triggering system events and / or operation events associated with the target task information, the target task information is determined from the task information set.

10. The method according to claim 9, characterized in that, In response to a system event and / or an operational event associated with the target task information, the target task information is determined from the task information set, including: In response to the system event and / or an operation event associated with the target task information, the target task information is created in the task information set; or... In response to the system event and / or the operation event associated with the target task information, the initial task information in the task information set is switched to the target task information.

11. The method according to claim 10, characterized in that, The method further includes: The task information set and the key operation events are used to generate operation prompt information, wherein the operation prompt information is used to indicate the creation of the target task information, or the operation prompt information is used to indicate the switching of the initial task information to the target task information; Creating the target task information in the task information set includes: when the operation prompt information indicates that the target task information should be created, guiding the target model to create the target task information in the task information set according to the operation prompt information; Switching the initial task information in the task information set to the target task information includes: when the operation prompt information indicates that the initial task information should be switched to the target task information, guiding the target model to switch the initial task information in the task information set to the target task information according to the operation prompt information.

12. The method according to claim 1, characterized in that, The set of operation events associated with the target task information includes: Capture at least one type of operation event associated with the target task information; Obtain the context information of the action event of the type described; Using the context information and the type of operation event, record the operation events in the operation event set.

13. The method according to claim 1, characterized in that, Based on the aforementioned key operational events, the product's R&D information is determined, including: Update the target task information based on the key operational events; The updated target task information is converted into a report file that includes the R&D information.

14. The method according to claim 13, characterized in that, The method further includes: The task information set, the target task information, the operation event set, the context information of the target task information, and the key operation events are used to generate update prompt information, wherein the update prompt information is used to indicate whether updating the target task information is allowed or prohibited. Based on the key operation event, updating the target task information includes: responding to the update prompt information indicating that updating the target task information is allowed, guiding the target model according to the update prompt information, and updating the target task information.

15. The method according to claim 14, characterized in that, The updated target task information is converted into a report file that includes the R&D information, including: The updated target task information is determined to include at least one of the following target information: update progress information, update process information, and update remarks information, wherein the update progress information is used to indicate the progress of updating the updated target task information, the update process information is used to indicate the process of updating the updated target task information, and the update remarks information is used to indicate the information noted during the updating process of the target task information. The target information is converted into the report file.

16. The method according to any one of claims 1 to 15, characterized in that, The method further includes: A hierarchical data storage structure is used to store the operation event and its associated event information in layers, wherein the associated event information is used to represent events and / or information associated with the operation event.

17. The method according to claim 16, characterized in that, The hierarchical data storage structure includes: a first-layer data storage structure, a second-layer data storage structure, a third-layer data storage structure, and a fourth-layer data storage structure. The associated event information includes: the operation event set, the key operation events, the aggregated key operation events, and the R&D information. The hierarchical data storage structure is used to store the operation events and their associated event information hierarchically, including: The operation event set is stored using the first-level data storage structure; The key operation events are stored using the second-layer data storage structure; The aggregated key operation events are stored using the third-layer data storage structure. The research and development information is stored using the fourth-layer data storage structure.

18. The method according to any one of claims 1 to 15, characterized in that, The method further includes: Obtain inquiry information regarding the aforementioned R&D information; The target model is used to analyze the query information to obtain a response that matches the query information.

19. An information processing method for a product, characterized in that, Applications in product development platforms, including: Obtain a task information set for intelligent products, wherein the task information in the task information set is used to describe the R&D tasks that need to be performed in the functional R&D process of the intelligent products; From the task information set, the target task information of the intelligent product is determined, wherein the target task information is used to describe the target R&D tasks that were not completed during the R&D process of the intelligent product; Obtain the set of operation events associated with the target task information, wherein the operation events in the set of operation events are used to represent the operation behavior performed on the target task information; Using a target model, at least one key operation event is identified from the set of operation events, wherein the importance of the key operation event to the target R&D task is greater than the importance of the operation events in the set of operation events other than the key operation event to the target R&D task, and the target model is obtained by training a large model; Based on the key operational events, the R&D information of the smart product is determined, wherein the R&D information is used to represent different statistical information generated by the smart product in the functional R&D process.

20. A method for processing information about a product, characterized in that, include: The task information set of the product is displayed on the operation interface, wherein the task information in the task information set is used to describe the research and development tasks that need to be performed in the research and development process of the product. In response to an information processing instruction applied to the operation interface, the R&D information of the product is displayed on the operation interface. The R&D information represents different statistical information generated by the product during the R&D process. At least one key operation event is determined based on the operation event set associated with the target task information in the task information set. The target task information describes a target R&D task that was not completed during the product's R&D process. The operation events in the operation event set represent operational behaviors performed on the target task information. The importance of the key operation event to the target R&D task is greater than the importance of other operation events in the operation event set (excluding the key operation event) to the target R&D task. The key operation event is identified from the operation event set using a target model, which is obtained by training a large model.

21. A method for storing information about a product, characterized in that, include: Obtain a set of task information for the product, wherein the task information in the set of task information is used to describe the research and development tasks that need to be performed in the research and development process of the product. From the task information set, the target task information of the product is determined, wherein the target task information is used to describe the target R&D tasks that were not completed during the R&D process of the product; From the hierarchical data storage structure, query the set of operation events associated with the target task information, wherein the operation events in the set of operation events are used to represent the operation behavior performed on the target task information; Using a target model, at least one key operation event is identified from the set of operation events, and the key operation event is stored in the data storage structure of the corresponding level in the hierarchical data storage structure. The key operation event is more important to the target R&D task than the operation events in the set of operation events other than the key operation event. The target model is obtained by training a large model. Based on the key operational events, the R&D information of the product is determined, and the R&D information is stored in the data storage structure of the corresponding level in the hierarchical data storage structure, wherein the R&D information is used to represent different statistical information generated by the product in the R&D process.

22. An information processing system for a product, characterized in that, include: A client is used to transmit a set of task information for a product, wherein the task information in the set of task information is used to describe the research and development tasks that need to be performed in the research and development process of the product. The server is configured to: determine target task information of the product from the task information set, wherein the target task information describes target R&D tasks that were not completed during the product's R&D process; obtain a set of operation events associated with the target task information, wherein the operation events in the operation event set represent operation behaviors performed on the target task information; identify at least one key operation event from the operation event set using a target model, wherein the importance of the key operation event to the target R&D task is greater than the importance of other operation events in the operation event set (excluding the key operation event) to the target R&D task, and the target model is obtained by training a large model; determine R&D information of the product based on the key operation event, wherein the R&D information represents different statistical information generated by the product during the R&D process; and transmit the R&D information to the client.

23. A computing device, characterized in that, include: Memory, which stores executable programs; A processor for running the program, wherein the program, when running, performs the method according to any one of claims 1 to 21.

24. An electronic device, characterized in that, include: Memory, which stores executable programs; A processor, connected to the memory via a bus, is used to run the program, wherein the program executes the method according to any one of claims 1 to 21.

25. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored executable program, wherein, when the executable program is executed, it controls the device on which the computer-readable storage medium resides to perform the method according to any one of claims 1 to 21.

26. A computer program product, characterized in that, It includes a computer program that, when executed by a processor, implements the method described in any one of claims 1 to 21.

Images