Project learning method and device, equipment and storage medium

Abstract

The invention discloses a project learning method and device, equipment and a storage medium. When the method provided by the embodiment of the invention is executed, a large language model firstly generates a teaching process containing N tutorial blocks by acquiring a project target of a learning user and a related knowledge background, each tutorial block corresponds to a project function to realize teaching, and the plurality of tutorial blocks jointly cover the overall function of the project target. And the big language model sends the tutorial blocks to the user in sequence until all the tutorial blocks are completed. And after the learning of each tutorial block is finished, checking whether the user achieves a learning target, and if so, sending the next tutorial block to the user for continuous learning until the user completes the learning of all tutorial blocks. N is a positive integer greater than 2. According to the method, the teaching process is automatically generated by using the large language model, and the dependence on manual annotation is reduced, so that the labor cost is reduced and the efficiency is improved.

Description

[0001] The present application claims priority to the Chinese patent application No. 2025104657167, filed on April 14, 2025, and entitled "A project learning method, device, equipment and storage medium", the whole content of which is incorporated herein by reference. TECHNICAL FIELD

[0002] The present application relates to the field of artificial intelligence, in particular to a project learning method, device, equipment and storage medium. BACKGROUND

[0003] The traditional intelligent tutoring system structures the existing teaching resources (such as questions, videos, etc.), clearly marks the knowledge points and cognitive levels, and then makes personalized recommendations according to the student's interaction data (such as the correctness of the answers). This method is suitable for standardized learning tasks such as exercises and tests.

[0004] However, project-based learning involves more complex and dynamic tasks, and the performance and cognitive state of students cannot be simply marked by traditional structured labels, because project-based learning focuses on students' exploration, creativity and comprehensive ability in completing actual tasks, and these processes cannot be fully measured by fixed labels. In addition, the needs of students in project-based learning are not limited to resource recommendations, but more flexible and personalized guidance according to actual progress is needed.

[0005] Therefore, relying solely on traditional structured labeling in project-based learning is not only inefficient, but also has high labor costs, because more time and resources are needed for detailed labeling and updating. SUMMARY

[0006] Based on the above problems, the present application provides a project learning method, device, equipment and storage medium, which can improve the timeliness of project learning, and avoid attackers bypassing attack detection by deleting or tampering with logs.

[0007] The embodiments of the present application disclose the following technical solutions:

[0008] A project learning method applied to a large language model, the method comprising:

[0009] obtaining a project goal of a learning user and a user knowledge background; the user knowledge background is a knowledge background related to the project goal;

[0010] generating a teaching flow based on the project goal and the user knowledge background; the teaching flow comprises N teaching blocks with a learning order; each of the teaching blocks comprises an implementation teaching of a piece of project function in the project goal, and a plurality of the teaching blocks comprise an implementation teaching of the whole project function in the project goal; N is a positive integer greater than 2;

[0011] starting from a first teaching block in the teaching flow, sending each of the teaching blocks to the learning user in the learning order one by one until a learning end of an Nth teaching block;

[0012] wherein, after the learning user completes an ith teaching block, checking whether the learning user completes a teaching goal of the ith teaching block, and if so, continuing to send an (i+1)th teaching block until the learning user learns the N teaching blocks; i is a positive integer.

[0013] In a possible implementation, the generating the teaching flow based on the project goal and the user knowledge background comprises:

[0014] generating a project learning step according to the project goal and the user knowledge background;

[0015] splitting the project learning step into a plurality of sub-steps according to the learning order;

[0016] splitting each of the sub-steps into a plurality of teaching blocks based on a plurality of step goals of the sub-step;

[0017] integrating the teaching blocks of all the sub-steps together according to the learning order to obtain the teaching flow.

[0018] In a possible implementation, the checking whether the learning user completes the teaching goal of the ith teaching block comprises:

[0019] summarizing a teaching knowledge background in the first to ith teaching blocks;

[0020] generating an ith checkpoint question and an ith answer requirement based on the teaching knowledge background in the first to ith teaching blocks and the user knowledge background;

[0021] obtaining an ith user answer of the learning user to the ith checkpoint question;

[0022] if the ith user answer meets the ith answer requirement, determining that the learning user has completed the teaching goal of the ith teaching block; otherwise, determining that the learning user has not completed the teaching goal of the ith teaching block.

[0023] In a possible implementation, the method further comprises:

[0024] If it is checked that the learning user does not complete the teaching goal of the ith course block, the uncompleted reason is found based on the ith user answer and the ith answer requirement;

[0025] The teaching explanation is generated based on the uncompleted reason;

[0026] The teaching explanation is sent to the learning user in the form of a course block.

[0027] In a possible implementation, after the teaching explanation is sent to the learning user in the form of a course block, the method further includes:

[0028] An ith updated user answer of the learning user for the ith course block is obtained; the ith updated user answer is a user answer rewritten by the learning user for the ith course block after learning the teaching explanation;

[0029] It is checked whether the ith updated user answer meets the ith answer requirement, if yes, it is determined that the learning user completes the teaching goal of the ith course block; otherwise, a new uncompleted reason is found based on the ith updated user answer and the ith answer requirement, and the teaching explanation is generated based on the new uncompleted reason, and the subsequent steps are executed to generate a new ith updated user answer until the new ith updated user answer meets the ith answer requirement.

[0030] In a possible implementation, the method further includes:

[0031] In response to receiving the learning goal, the learning course block and the learning mode of the learning user, a current learning progress of the learning user is summarized; the learning course block is any one of N course blocks included in the teaching process; the learning mode includes any one of explaining a concept, debugging an error and generating a quiz; the current learning progress includes a number of the course blocks that have been completed;

[0032] A learning reply is generated based on the current learning progress, the learning goal, the learning course block and the learning mode of the learning user.

[0033] A learning device of a project, the device includes:

[0034] A first obtaining unit is configured to obtain a project goal of a learning user and a user knowledge background; the user knowledge background is a knowledge background related to the project goal;

[0035] The first generating unit is configured to generate a teaching flow based on the project target and the user knowledge background, wherein the teaching flow comprises N teaching blocks with a learning sequence, each of the teaching blocks comprises implementation teaching of a block of project functions in the project target, a plurality of the teaching blocks comprise implementation teaching of the overall project functions in the project target, and N is a positive integer greater than 2.

[0036] The first integrating unit is configured to send each of the teaching blocks to the learning user in sequence according to the learning sequence, starting from the first teaching block in the teaching flow, until the learning of the Nth teaching block is completed.

[0037] When the learning user completes the ith teaching block, the first integrating unit checks whether the learning user completes the teaching target of the ith teaching block, and if the target is reached, the (i+1)th teaching block is sent until the learning user completes the N teaching blocks, wherein i is a positive integer.

[0038] In a possible implementation, the first generating unit specifically comprises:

[0039] A project learning step generating unit is configured to generate a project learning step according to the project target and the user knowledge background.

[0040] A sub-step splitting unit is configured to split the project learning step into a plurality of sub-steps according to the learning sequence.

[0041] A teaching block generating unit is configured to split each of the sub-steps into a plurality of teaching blocks based on a plurality of step targets of each of the sub-steps.

[0042] An integrating unit is configured to integrate the teaching blocks of all the sub-steps together according to the learning sequence to obtain the teaching flow.

[0043] A project learning device comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the project learning method as described above when executing the computer program.

[0044] A computer readable storage medium stores instructions, and when the instructions are executed on a terminal device, the terminal device performs the project learning method as described above.

[0045] Compared with the prior art, the present application has the following beneficial effects:

[0046] The application provides a project learning method, device, equipment and storage medium. Specifically, in the execution of the project learning method provided by the application, based on a large language model, the project target and related knowledge background of a learning user are first obtained. The knowledge background of the user is related to the project target, which lays the foundation for the generation of the subsequent teaching process. On this basis, the teaching process containing N tutorial blocks is generated by combining the project target and the user knowledge background information. Each tutorial block covers the implementation teaching of a specific function in the project target, and multiple tutorial blocks collectively cover the implementation teaching of the overall project function. N is a positive integer greater than 2, which determines the complexity and depth of the teaching process. In the teaching process, learning starts from the first tutorial block of the teaching process and gradually expands. Each tutorial block is sent to the learning user in a preset learning order until the Nth tutorial block is completed. When the learning user completes the ith tutorial block, the large language model checks whether they have achieved the teaching goal set for the tutorial block. If the goal has been achieved, the large language model will continue to push the i+1th tutorial block, and the cycle continues until the learning user completes all N tutorial blocks.

[0047] The project learning based on the large language model of the application can automatically generate a learning process, reducing manual intervention and improving efficiency. Through intelligent learning path recommendation, the large language model can provide more rapid and accurate personalized guidance to students, reducing labor costs. In addition, the application can customize the teaching process according to the specific needs of students by obtaining their project targets and knowledge backgrounds. The design of each tutorial block is based on the project target and current knowledge background of the student, making the learning process more personalized, flexible and targeted. BRIEF DESCRIPTION OF DRAWINGS

[0048] To make the technical solutions in the embodiments or the prior art clearer, the following will briefly introduce the drawings needed in the embodiments or the prior art description. Obviously, the drawings in the following description are only some embodiments of the application, and other drawings can be obtained by those skilled in the art without creative effort.

[0049] Figure 1 A method flowchart of a project learning method provided by the application embodiment;

[0050] Figure 2 A method flowchart of a project teaching process generation method provided by the application embodiment;

[0051] Figure 3 A method flowchart of a tutorial block teaching goal completion checking method provided by the application embodiment;

[0052] Figure 4 A method flowchart of a personalized learning feedback and tutoring method provided for an embodiment of the present application;

[0053] Figure 5 A structural schematic diagram of a project learning device provided for an embodiment of the present application. DETAILED DESCRIPTION

[0054] In order to facilitate the understanding of the technical solutions provided by the embodiments of the present application, the background art related to the embodiments of the present application will be described first.

[0055] The traditional intelligent tutoring system provides personalized recommendations based on the structured annotation of teaching resources and the interactive data of students. This approach is suitable for standardized learning tasks, such as exercises and tests.

[0056] However, the tasks involved in project-based learning are more complex and dynamic, and the performance of students cannot be simply described by traditional structured labels. Project-based learning emphasizes the exploration, creativity and comprehensive ability of students in actual tasks, which cannot be fully measured by fixed labels. In addition, the needs of students in project-based learning pay more attention to flexible and personalized guidance, rather than relying solely on resource recommendations. Therefore, in project-based learning, simply relying on traditional structured annotation methods is not only inefficient, but also requires a large amount of manual cost for detailed annotation and updating.

[0057] To address this problem, embodiments of the present application provide a project learning method, apparatus, device, and storage medium. First, the learning user's project goals and the user's knowledge background related to the project goals are obtained. A teaching process is then generated based on the user's project goals and knowledge background. This teaching process consists of N tutorial blocks, each of which represents a learning task for a specific functional module within the project goal. The tutorial blocks are arranged according to a specific learning order to form a coherent learning path. Each tutorial block specifically explains how to implement a specific function within the project goal, ensuring that the learning user can gradually master the various functions required for the project. The combination of multiple tutorial blocks can help users achieve the overall functionality of the project goal. During the teaching process, starting with the first tutorial block in the teaching process, the tutorial blocks are gradually pushed to the learning user in the learning order until the learning task of the Nth tutorial block is completed. Each time a learning user completes a tutorial block, the large language model checks whether the user has achieved the learning goal for that tutorial block. If the user has achieved the learning goal for the current tutorial block, the next tutorial block is pushed until the user has completed the learning tasks for all tutorial blocks. This application's project-based learning approach, based on a large language model, automatically generates personalized learning processes, significantly reducing manual intervention and improving efficiency. Through intelligent learning path recommendations, the large language model can provide students with personalized guidance more quickly and accurately, significantly reducing labor costs.

[0058] The following will be combined with the accompanying drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0059] See also Figure 1 , this figure is a method flow chart of a project learning method provided in an embodiment of the present application. The method is implemented through a large language model. The large language model supports prompting and chain of thought technologies. These two technologies greatly enhance the model's understanding and generation capabilities.

[0060] Thought Chaining Technology: This technology allows the model to display its reasoning process when generating answers, explaining how the final answer was reached through intermediate steps and logical chains. This not only improves the interpretability and transparency of the generated content, but also enables the model to better understand and solve complex multi-step problems. For example, when generating a teaching process, the model can ensure that the generated teaching content is both coherent and reasonable by displaying the detailed reasoning for each step.

[0061] Prompting technique: The prompting technique refers to guiding the model to generate the expected output by providing specific input text or context information. This technique can help the model better adapt to specific task requirements and personal preferences. For example, when generating personalized teaching processes, prompts can be used to specify the user's specific project goals and knowledge background, so that the generated teaching content is more tailored to the user's needs.

[0062] In combination with these two techniques, large language models can efficiently implement the project learning method of the present application.

[0063] As shown in Figure 1 , the project learning method can include steps S101-S103:

[0064] S101: Obtain the project goal of the learning user and the user's knowledge background.

[0065] In the S101 stage, the large language model can understand the existing knowledge structure and goals of the learning user by analyzing the project goal of the learning user and the user's knowledge background, so as to determine the learning path. The thinking chain technology plays an important role in this stage, which combines the project goal with the user's knowledge background information to generate a hierarchical teaching process. This process not only considers the user's current knowledge level, but also presets the content of each tutorial block to gradually promote the progress of learning and ensure clear logical connection and goal achievement between each tutorial block.

[0066] S102: Generate a teaching process based on the project goal and the user's knowledge background.

[0067] In the S102 stage, based on the project goal and the user's knowledge background, the generated teaching process contains N tutorial blocks with a certain learning order, and the content of each tutorial block is around a certain aspect of the project function, gradually realizing the overall project goal. This reflects the reasoning ability of the thinking chain technology: deriving each sub-goal from the overall project goal and arranging it in the tutorial block in the appropriate order.

[0068] Wherein, N is a positive integer greater than 2.

[0069] Exemplarily, suppose a learning user wants to learn how to build a simple website. The user's project goal is "to build a website with basic functions", and the user's background knowledge includes some basic HTML and CSS knowledge, but is not familiar with JavaScript and front-end frameworks such as React or Vue. The large language model will use the prompting technique to understand the user's project goal and knowledge background, and use the thinking chain technology to develop a detailed, step-by-step teaching process. The teaching process will include multiple tutorial blocks, each of which focuses on achieving a specific function in the project goal.

[0070] N course block examples:

[0071] Course block 1:

[0072] Title: Learn HTML common tags;

[0073] Content: Learn HTML common tags, such as , <h1>, , and understand their roles and usage;

[0074] Objective: Master commonly used HTML tags and be able to correctly use these tags to build web page content.

[0075] Tutorial Block 2:

[0076] Title: Create a basic HTML document structure;

[0077] Content: Learn how to create a basic structure in an HTML document, including, <title>,< / title> the use of tags such as,

[0078] Objective: Be able to independently create an HTML document containing basic structure, and understand the role and necessity of each part.

[0079] Tutorial Block 3:

[0080] Title: Understand HTML element nesting;

[0081] Content: Learn how HTML elements are nested, including the difference between block-level elements and inline elements, and practice nesting elements reasonably to build complex page structures;

[0082] Objective: Correctly and clearly understand the nesting relationship of HTML elements, and have the ability to build complex web page structures.

[0083] Learning process:

[0084] From the first tutorial block, send it to the user, send tutorial block 1: "Learn HTML commonly used tags". Then check if the user has completed the goal of tutorial block 1, if the user has completed, continue to send tutorial block 2: "Create a basic HTML document structure".

[0085] Send subsequent tutorial blocks in turn: after completing tutorial block 2, send tutorial block 3: "Understand HTML element nesting".

[0086] In this way, the large language model can generate a detailed, step-by-step teaching process according to the user's project goals and knowledge background, and ensure that the user can achieve specific learning goals at each step.

[0087] See Figure 2 . Figure 2 A method flowchart of a project teaching process generation method is provided for the embodiments of the present application. Accordingly, step S102 generates a teaching process based on the project target and the user knowledge background, which can be implemented through steps A1-A4:

[0088] A1: generating project learning steps according to the project target and the user knowledge background.

[0089] In step A1, the key is to understand the project target and the user background to ensure that the learning steps are highly consistent with these targets and backgrounds.

[0090] Example 1, assuming that the project target is to build a website with basic functions. The user's knowledge background is familiar with basic HTML and CSS, but not very familiar with JavaScript and front-end frameworks. The corresponding learning steps are: Step 1: learn HTML basics. Step 2: learn CSS style design. Step 3: learn JavaScript basics. Step 4: learn advanced JavaScript. Step 5: learn front-end frameworks (such as React or Vue). Step 6: integrate all content and deploy the website.

[0091] A2: split the project learning steps into multiple sub-steps according to the learning order.

[0092] Splitting the project learning steps into sub-steps according to the learning order can facilitate more detailed division of complex learning content, reduce the difficulty of learning, and ensure that the learning process is more organized and clear.

[0093] Example 2, continue to illustrate with example 1, the learning step "learn HTML basics" can be split into sub-steps: Step 1.1 understand the basic syntax of HTML; Step 1.2 create a basic HTML document structure; Step 1.3 master common HTML tags; Step 1.4 learn how to use HTML tables.

[0094] A3: split the sub-steps into multiple tutorial blocks based on multiple step targets of each sub-step.

[0095] Split each sub-step into specific tutorial blocks, and each tutorial block corresponds to an independent learning target, which can ensure that each learning target is effectively covered.

[0096] Example 3, continue to illustrate with example 2, split the sub-step "Step 1.3 master common HTML tags" into multiple tutorial blocks: tutorial block 1: understand common HTML tags; tutorial block 2: create a basic HTML document structure; tutorial block 3: understand HTML element nesting.

[0097] A4: Integrate the tutorial blocks of all sub-steps according to the learning order to obtain the teaching process.

[0098] Integrating all the split tutorial blocks into a complete teaching process in a reasonable learning order can ensure that the learning process is orderly and gradual.

[0099] For example, the integrated teaching process is:

[0100] Tutorial block 1: Understand common HTML tags;

[0101] Tutorial Block 2: Creating a basic HTML document structure;

[0102] Tutorial Block 3: Understanding HTML element nesting;

[0103] Tutorial Block 4: Simple CSS Selectors;

[0104] Tutorial Block 5: Complex CSS Selectors;

[0105] Tutorial Block 6: Selector Priority;

[0106] Tutorial Block 7: Mastering Common HTML Tags;

[0107] Tutorial Block 8: Learn basic JavaScript syntax;

[0108] Tutorial Block 9: Learn DOM operations;

[0109] Tutorial Block 10: Learning event handling;

[0110] Tutorial Block 11: Learn the basic concepts of front-end frameworks (such as React);

[0111] Tutorial Block 12: Practical Project Integration and Deployment.

[0112] S103: Starting from the first tutorial block in the teaching process, each tutorial block is sent to the learning user in sequence according to the learning order until the learning of the Nth tutorial block is completed.

[0113] The application of prompt word technology is demonstrated in stage S103. Within each tutorial block, the large language model provides targeted prompts to help learners understand and master the tutorial content. After the user completes the learning objectives for each tutorial block, the large language model uses prompt words and relevant feedback to check whether the learner has achieved the predetermined objectives. If the objectives are achieved, the large language model uses prompt words to guide the learner to continue to the next tutorial block. This method uses prompt words to help users clearly identify learning tasks and progress.

[0114] The prompt word technology also helps the model accurately judge whether the user has achieved the goal after completing the study of each tutorial block, and decide whether to advance to the next tutorial block. This gradual advancement mechanism is based on the real-time feedback of the learning status and the reasoning judgment of the thinking chain of the large language model. Specifically, during the learning process, each time the user completes a tutorial block (numbered i), the large language model will check whether they have achieved the learning goal of the tutorial block. If the goal is achieved, the large language model will continue to send the next tutorial block (i.e., the i+1th) until the user completes all tutorial blocks (N in total). Among them, i represents the number of the current tutorial block, which is a positive integer starting from 1.

[0115] See also Figure 3 , Figure 3 A flowchart of a method for checking the completion of a tutorial block teaching objective provided in an embodiment of the present application. The method may specifically include steps B1-B4:

[0116] B1: Summarize the tutorial knowledge background in tutorial blocks 1 to i.

[0117] By reviewing and sorting out all the knowledge points from the first to the i-th tutorial block, we can confirm the knowledge that the learning user should master, providing the necessary background information for evaluating the completion of the current teaching objectives. This can more accurately determine whether the user truly understands and masters the relevant knowledge points.

[0118] B2: Based on the tutorial knowledge background in the 1st to ith tutorial blocks and the user knowledge background, generate the i-th checkpoint question and the i-th answer requirement.

[0119] Design relevant checkpoint questions and detailed answer requirements based on the knowledge that users have already learned and should master (i.e., the tutorial knowledge background) and their current knowledge level. These questions are intended to verify whether users have truly mastered the previously learned knowledge and ensure that they can correctly apply this knowledge in practice.

[0120] B3: Obtain the i-th user's answer to the i-th checkpoint question provided by the learning user.

[0121] Collect users' answers to checkpoint questions. This way, you can assess their understanding of the knowledge points and understand their performance in actual operations.

[0122] B4: If the answer given by the i-th user meets the i-th answer requirement, it is determined that the learning user has completed the teaching objective of the i-th tutorial block; otherwise, it is determined that the learning user has not completed the teaching objective of the i-th tutorial block.

[0123] The user is determined to have successfully completed the instructional objective of the i-th tutorial block by assessing whether the user's answer to the checkpoint question meets the pre-set requirement. If the answer meets the requirement, it means that the user has achieved the objective of the stage; if not, the user needs to continue learning or be provided with additional help to achieve the objective.

[0124] It should be noted that the form of the user's answer can be selecting an option or a sentence, depending on the specific form of the checkpoint question. For example, for a multiple-choice question, the user only needs to make a selection from the given options; for an open-ended question, the user needs to provide a description or explanation. Specifically, if the checkpoint question is a multiple-choice question, the user needs to select one or more correct answers from the given options; if the checkpoint question is an open-ended question, the user needs to write a paragraph to answer the question.

[0125] There are also two cases for the requirement of the i-th answer: one is that the user's selection must correspond to a pre-set option, such as the answer to a multiple-choice question needs to accurately correspond to the correct option; the other is that the content of the user's answer needs to contain key words in the standard answer, such as the answer to an open-ended question needs to contain key terms or concepts. This flexible way of evaluating user answers can not only ensure that users understand and master knowledge points, but also adapt to different forms of questions, thereby more comprehensively testing users' learning achievements.

[0126] Exemplarily, a checkpoint question of the multiple-choice type:

[0127] Checkpoint question: In HTML, <title>Where should the label go?< / title>

[0128] A. within the

[0129] B. within the

[0130] C. within the

[0131] D. <meta> within the

[0132] i-th answer requirement: the user's answer corresponds to option A.

[0133] User's answer: select option B (i.e., within the

[0134] Evaluation result: the user selected the wrong option (option B), which does not meet the answer requirement, and it is determined that the user has not completed the instructional objective of the tutorial block.

[0135] Checkpoint question of open-ended type:

[0136] Checkpoint question: Please briefly describe the role of the float attribute in CSS.

[0137] The i-th answer requirement: the user's answer includes keywords such as "float", "left", "right", "wrap" in the standard answer.

[0138] Standard answer: The float attribute is used to float an element to the left or right of its container, allowing text and other floating elements to flow around it.

[0139] User answer: The float attribute can make an element float to the left or right, and let other content wrap around it.

[0140] Evaluation result: The user's answer contains keywords such as "float", "left", "right", "wrap", which meets the answer requirements, and determines that the user has completed the teaching objectives of this tutorial block.

[0141] Reference Figure 4 , Figure 4 A personalized learning feedback and tutoring method is provided for the embodiments of the present application, which specifically includes steps C1-C3:

[0142] C1: If it is checked that the learning user has not completed the teaching objectives of the i-th tutorial block, find the reason for not completing based on the i-th user answer and the i-th answer requirement.

[0143] In the teaching process, it is inevitable that some learning users will not be able to complete the teaching objectives of a certain tutorial block. In order to ensure the quality of teaching and the progress of learning, if a user does not complete a certain teaching objective, the large language model will take corresponding measures for analysis and improvement.

[0144] Firstly, when it is checked that a certain learning user has not completed the teaching objectives of the i-th tutorial block, the large language model will further analyze the user's answer. Specifically, the large language model will analyze the relationship between the user's i-th answer (i.e., the user's specific answer in the tutorial) and the requirement of this answer to find the reason for not completing the teaching objectives. For example, if the user's answer does not meet the requirements of the question, the large language model will automatically identify the differences and determine whether it is due to understanding bias, weak knowledge grasp, or other factors.

[0145] C2: Generate teaching lectures based on the reason for not completing.

[0146] By identifying the reasons for not completing, the large language model can generate personalized teaching lectures based on the analysis results. These lecture contents will focus on strengthening the user's specific weaknesses. For example, if the user answers incorrectly on a certain knowledge point, the large language model will provide more targeted lectures according to the type of error to help the user understand and master this part of the content.

[0147] C3: Send the teaching lectures in the form of tutorial blocks to the learning user.

[0148] Finally, to help users fill in knowledge gaps in a timely manner, the large language model will generate teaching explanations and send them to the learning user in the form of tutorial blocks. The explanation content can include explanatory text, diagrams, videos, or other auxiliary learning materials. In this way, the learning user can obtain timely feedback and guidance, further deepening their understanding of the knowledge points, and thus laying a foundation for successfully completing subsequent learning tasks.

[0149] In summary, steps C1-C3 provide effective support for improving learning effectiveness and teaching quality by checking user answers, analyzing reasons for not completing, generating personalized explanation content, and sending it to the user in a timely manner.

[0150] In one possible implementation, after the teaching explanation is sent to the learning user in the form of a tutorial block in step C3, it further includes:

[0151] Obtain the i-th updated answer of the learning user for the i-th tutorial block, and check whether the i-th updated answer meets the i-th answer requirement. If so, it is determined that the learning user has completed the teaching goal of the i-th tutorial block; otherwise, based on the i-th updated answer and the i-th answer requirement, find a new reason for not completing, and based on the new reason for not completing, generate a teaching explanation and subsequent steps to generate a new i-th updated answer until the new i-th updated answer meets the i-th answer requirement. Wherein the i-th updated answer is the user answer rewritten by the learning user for the i-th tutorial block after learning the teaching explanation.

[0152] This process aims to help the learning user eventually achieve and complete the teaching goal through continuous teaching explanation and answer updating. The specific process is to provide personalized teaching guidance based on the learning user's answer performance in each case of not meeting the requirements, until the learning user can meet the answer requirements of the question, thereby successfully completing the teaching goal.

[0153] Specifically, first, the large language model will obtain the learning user's rewritten answer based on new understanding, known as the "i-th updated answer". Once the learning user submits the updated answer, the large language model will again check it to assess whether it has met the i-th answer requirement. If the new updated answer meets the requirement, the large language model will confirm that the learning user has successfully completed the teaching goal of the i-th tutorial block, marking the progress of the learning process. Conversely, if the updated answer still does not meet the requirements, the large language model will again analyze the shortcomings of the new answer, find a new reason for not completing, and generate further teaching explanations or guidance content based on it, forming a new cycle until the learning user can finally submit an answer that meets the requirements.

[0154] Through such a feedback mechanism, large language models not only help learners constantly improve their answers, but also ensure that each learner receives sufficient guidance and support in the process of mastering knowledge. This approach ensures that learners do not stagnate due to a particular problem, but rather continuously improve their abilities through continuous learning and reflection, ultimately achieving the completion of learning goals. The core of this process lies in personalized feedback, timely error correction, and gradual learning reinforcement, through continuous explanation and answer updates to ensure that learners master all necessary knowledge points and ultimately achieve the teaching objectives.

[0155] In one possible implementation, the method further includes:

[0156] In response to receiving the learning goal, learning tutorial block, and learning method of the learning user, summarizing the current learning progress of the learning user, and generating a learning reply based on the current learning progress of the learning user, the learning goal, the learning tutorial block, and the learning method. Wherein, the current learning progress includes the number of completed learning tutorial blocks.

[0157] When the learning user receives the learning goal, learning tutorial block, and learning method, the large language model will first summarize the user's current learning progress based on this information. The learning goal is the specific learning content that the learning user wants to learn, the learning tutorial block is one of the N modules designed in the teaching process, and each tutorial block contains different knowledge points and skill requirements. The learning method refers to the form of teaching feedback that the user wants the large language model to provide during the learning process, including explaining concepts, debugging errors, and generating quizzes, etc.

[0158] In summarizing the learning progress, the large language model will first evaluate the number of completed tutorial blocks as an important indicator of the learning user's current learning progress. This indicator can help the large language model understand how much knowledge the user has mastered and which parts may need further learning or improvement. The number of completed tutorial blocks directly reflects the progress the learner has experienced in the learning process, which is crucial for determining the learner's learning status and developing the next learning plan.

[0159] Once the current learning progress of the user is obtained, the large language model will generate a learning reply by combining the learning goal, the content of the tutorial block, and the learning approach. This learning reply is not just a simple feedback, it should include an evaluation of progress towards the learning goal, as well as suggestions and guidance based on the user's current learning progress. For example, if the user has completed a certain number of tutorial blocks, but has not yet mastered some key concepts, the large language model can provide targeted support based on the learning approach adopted by the user (such as explaining concepts or debugging errors), helping the user to fill in knowledge gaps. The large language model can also recommend additional learning tasks or resources for the user based on the content of the tutorial blocks that have not been completed, ensuring that the learning goal is completed on time.

[0160] By combining the user's learning progress with the learning goal, the large language model can flexibly generate personalized learning replies, helping the user to clearly understand their strengths and weaknesses in learning, and providing more efficient and targeted suggestions for subsequent learning.

[0161] In one possible implementation, the learning approach further includes visualization. "Visualization" in the learning approach refers to the use of visual tools such as images, charts, and videos to help learners understand and remember information. This method can effectively improve learning effectiveness by converting abstract information into visual forms that are easy to understand. Visualization helps to present complex concepts or data more intuitively, allowing learners to quickly grasp key content and enhance memory and understanding.

[0162] In one possible implementation, the method further includes:

[0163] Checking whether the tutorial block, the teaching explanation, and the learning reply contain direct answers to the checkpoint questions, and if so, removing such content from the tutorial block, the teaching explanation, and the learning reply.

[0164] In order to improve the effectiveness and depth of learning, the tutorial block, the teaching explanation, and the learning reply should avoid directly providing answers to the checkpoint questions.

[0165] Based on the content of S101-S103, the project target of the learning user and the knowledge background related thereto are first obtained. On this basis, the large language model generates a teaching process including multiple tutorial blocks according to the project target and the knowledge background provided by the user. These tutorial blocks are arranged in a specific learning order, and each tutorial block teaches a specific function in the project target. Multiple tutorial blocks together constitute a teaching path for implementing the entire project function. The order number of each tutorial block is N, and N is greater than 2, ensuring the integrity and complexity of the teaching process. Then, starting from the first tutorial block in the teaching process, the large language model gradually sends each tutorial block to the learning user in the predetermined learning order until the last tutorial block is learned. During the learning process, when the learning user completes a specific tutorial block, the large language model checks whether the teaching goal of the tutorial block is achieved. If the goal is achieved, the large language model will continue to send the next tutorial block to the user for learning; if it is not achieved, the large language model will give further guidance or adjust the teaching strategy until the learning user completes the learning of all N tutorial blocks. The present application utilizes the project learning of the large language model to automatically generate a learning process, reduces manual intervention, and improves efficiency. Through intelligent recommendation of the learning path, the large language model can more quickly and accurately provide personalized guidance for students, thereby reducing labor costs.

[0166] Referring to Figure 5 , Figure 5 A structural schematic diagram of a project learning device provided by an embodiment of the present application is shown in FIG. 1. As shown in FIG. 1, the project learning device includes: Figure 5

[0167] A first obtaining unit 501 is configured to obtain a project target of a learning user and a user knowledge background; the user knowledge background is a knowledge background related to the project target;

[0168] A first generating unit 502 is configured to generate a teaching process based on the project target and the user knowledge background; the teaching process includes N tutorial blocks with a learning order; each tutorial block includes implementation teaching of a block of project functions in the project target, and multiple tutorial blocks include implementation teaching of the overall project function of the project target; N is a positive integer greater than 2;

[0169] A first comprehensive unit 503 is configured to gradually send each tutorial block to the learning user in the learning order, starting from the first tutorial block in the teaching process, until the learning of the Nth tutorial block is completed.

[0170] ​When the learning user finishes the ith course block, it is checked whether the learning user finishes the teaching target of the ith course block, and if the target is reached, the ith+1 course block is sent until the learning user finishes the N course blocks; i is a positive integer.

[0171] In a possible implementation, the first generating unit 502 specifically includes:

[0172] A project learning step generating unit, configured to generate a project learning step according to the project target and the user knowledge background;

[0173] A substep splitting unit, configured to split the project learning step into a plurality of substeps according to the learning sequence;

[0174] A course block generating unit, configured to split each substep into a plurality of course blocks based on a plurality of step targets of the substep;

[0175] An integrating unit, configured to integrate the course blocks of all substeps together according to the learning sequence to obtain the teaching flow.

[0176] In a possible implementation, the apparatus further includes:

[0177] A knowledge background summarizing unit, configured to summarize course knowledge backgrounds in the 1th to ith course blocks;

[0178] A second generating unit, configured to generate an ith checkpoint question and an ith answer requirement based on the course knowledge backgrounds in the 1th to ith course blocks and the user knowledge background;

[0179] A second acquiring unit, configured to acquire an ith user answer of the learning user to the ith checkpoint question;

[0180] A determining unit, configured to determine that the learning user has finished the teaching target of the ith course block if the ith user answer meets the ith answer requirement, and otherwise, determine that the learning user has not finished the teaching target of the ith course block.

[0181] In a possible implementation, the apparatus further includes:

[0182] An unfinished reason finding unit, configured to find an unfinished reason based on the ith user answer and the ith answer requirement if it is checked that the learning user has not finished the teaching target of the ith course block;

[0183] A third generating unit, configured to generate a teaching explanation based on the unfinished reason;

[0184] The explanation sending unit is configured to send the teaching explanation in the form of a course block to the learning user.

[0185] In a possible implementation, the apparatus further includes:

[0186] The third obtaining unit is configured to obtain an ith updated answer of the learning user for the ith course block; the ith updated answer is a user answer of the learning user re-written for the ith course block after learning the teaching explanation;

[0187] The second synthesizing unit is configured to check whether the ith updated answer meets the ith answer requirement, and if yes, determine that the learning user has completed the teaching target of the ith course block; otherwise, find a new uncompleted reason based on the ith updated answer and the ith answer requirement, and perform the generating teaching explanation based on the uncompleted reason and the subsequent steps to generate a new ith updated answer based on the new uncompleted reason, until the new ith updated answer meets the ith answer requirement.

[0188] In a possible implementation, the apparatus further includes:

[0189] The current learning progress summarizing unit is configured to summarize a current learning progress of the learning user in response to receiving a learning target, a learning course block and a learning manner of the learning user; the learning course block is any one of N course blocks included in the teaching flow; the learning manner includes any one of explaining a concept, debugging an error and generating a quiz; and the current learning progress includes a number of the course blocks that have been completed.

[0190] The fourth generating unit is configured to generate a learning reply based on the current learning progress, the learning target, the learning course block and the learning manner of the learning user.

[0191] In addition, the embodiments of the present application further provide a project learning device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the project learning method as described above when executing the computer program.

[0192] In addition, the embodiments of the present application further provide a computer readable storage medium, which stores instructions, and the instructions, when executed on a terminal device, cause the terminal device to perform the project learning method as described above.

[0193] The embodiment of the present application provides a project learning device, first, a first acquisition unit 501 is used to acquire a project target of a learning user and a user knowledge background, wherein the user knowledge background is a knowledge background related to the project target. A first generation unit 502 is used to generate a teaching process based on the project target and the user knowledge background, wherein the teaching process comprises N teaching blocks with a learning order; each teaching block comprises implementation teaching of a block of project functions in the project target, and the plurality of teaching blocks comprise implementation teaching of overall project functions of the project target; and N is a positive integer greater than 2. Then, a first comprehensive unit 503 is used to send each teaching block to the learning user in turn according to the learning order from the first teaching block in the teaching process, until the learning of the Nth teaching block is completed. When the learning user completes the ith teaching block, whether the learning user completes the teaching target of the ith teaching block is checked, if the target is reached, the (i+1)th teaching block is continuously sent until the learning user learns the N teaching blocks; i is a positive integer. The project learning based on the large language model can automatically generate a learning process, reduces manual intervention, and thus improves the efficiency. Through intelligent learning path recommendation, the large language model can provide personalized guidance for students more quickly and accurately, while reducing the labor cost.

[0194] The above describes in detail the project learning method, device, equipment and storage medium provided by the present application. The embodiments in the specification are described in a progressive manner, and each embodiment mainly describes the differences from other embodiments. The same or similar parts of each embodiment can be referred to each other. For the device disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple, and the related parts can be referred to the method part. It should be pointed out that for ordinary skilled in the art, without departing from the principle of the present application, some improvements and modifications can be made to the present application, and these improvements and modifications also fall within the protection scope of the claims of the present application.

[0195] It should be understood that in the present application, "at least one" means one or more, "multiple" means two or more. "And / or", used to describe the association relationship of the associated objects, means that there can be three relationships, for example, "A and / or B" can mean: only A, only B and A and B exist at the same time, wherein A and B can be singular or plural. The character " / " generally represents that the front and rear associated objects are in an "or" relationship. "At least one of the following" or similar expressions means any combination of these items, including any combination of single item or multiple items. For example, at least one of a, b or c, can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", wherein a, b, and c can be single or multiple.

[0196] It is also to be noted that, as used in this specification and the appended claims, the singular forms "a," "an," and "the" include plural referents unless the context clearly dictates otherwise. Furthermore, the terms "comprising," "containing," or any other variation thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements is not necessarily limited to only those elements, but can include other elements not expressly listed or inherent to such process, method, article, or apparatus. < / h1>

Claims

1. A project learning method, characterized in that: Applied to a large language model, the method includes: Acquire the project goal and user knowledge background of the learning user; the user knowledge background is the knowledge background related to the project goal; A teaching process is generated based on the project goal and the user's knowledge background; the teaching process includes N tutorial blocks with a learning order; each tutorial block includes teaching on how to implement a project function within the project goal, and multiple tutorial blocks include teaching on how to implement the entire project function within the project goal; N is a positive integer greater than 2; Starting from the first tutorial block in the teaching process, each tutorial block is sent to the learning user in sequence according to the learning order until the Nth tutorial block is completed; Among them, when the learning user completes the i-th tutorial block, check whether the learning user has completed the teaching goal of the i-th tutorial block. If the goal is achieved, continue to send the i+1-th tutorial block until the learning user completes N tutorial blocks; i is a positive integer.

2. The method according to claim 1, characterized in that Generating a teaching process based on the project goal and the user's knowledge background includes: Generate project learning steps according to the project goal and the user knowledge background; Splitting the project learning steps into multiple sub-steps according to the learning sequence; Split the sub-steps into multiple tutorial chunks based on the multiple step goals of each sub-step; The tutorial blocks of all sub-steps are integrated together according to the learning sequence to obtain the teaching process.

3. The method according to claim 1, characterized in that The checking whether the learning user has completed the teaching objective of the i-th tutorial block includes: Summarize the tutorial knowledge background in the first to i tutorial blocks; Based on the tutorial knowledge background in the first to i tutorial blocks and the user knowledge background, generating an i-th checkpoint question and an i-th answer requirement; Obtaining an i-th user's answer to the i-th checkpoint question by the learning user; If the answer given by the i-th user meets the i-th answer requirement, it is determined that the learning user has completed the teaching objective of the i-th tutorial block; otherwise, it is determined that the learning user has not completed the teaching objective of the i-th tutorial block.

4. The method according to claim 3, characterized in that The method further comprises: If it is detected that the learning user has not completed the teaching objective of the i-th tutorial block, then finding the reason for the non-completion based on the answer of the i-th user and the i-th answer requirement; Generate a teaching explanation based on the reasons for non-completion; The teaching explanation is sent to the learning user in the form of tutorial blocks.

5. The method according to claim 4, characterized in that After sending the teaching explanation in the form of tutorial blocks to the learning user, the method further includes: Obtaining an i-th updated answer by the learning user for the i-th tutorial block; the i-th updated answer is a user answer rewritten by the learning user for the i-th tutorial block after learning the teaching explanation; Check whether the i-th updated answer meets the i-th answer requirement. If so, determine that the learning user has completed the teaching objective of the i-th tutorial block; otherwise, find a new reason for incompletion based on the i-th updated answer and the i-th answer requirement, and based on the new reason for incompletion, execute the generation of teaching explanations based on the incompletion reason and subsequent steps to generate a new i-th updated answer, until the new i-th updated answer meets the i-th answer requirement.

6. The method according to claim 1, characterized in that The method further comprises: In response to receiving the learning goal, learning tutorial block, and learning method of the learning user, summarizing the current learning progress of the learning user; the learning tutorial block is any one of the N tutorial blocks included in the teaching process; the learning method includes any one of explaining concepts, debugging errors, and generating quizzes; the current learning progress includes the number of the tutorial blocks that have been completed; A learning response is generated based on the current learning progress, the learning goal, the learning tutorial block, and the learning method of the learning user.

7. A project learning device, characterized in that: The device comprises: A first acquisition unit is configured to acquire a learning user's project goal and user knowledge background; the user knowledge background is knowledge background related to the project goal; A first generating unit is configured to generate a teaching process based on the project goal and the user's knowledge background; the teaching process includes N tutorial blocks having a learning order; each tutorial block includes teaching on how to implement a project function within the project goal, and multiple tutorial blocks include teaching on how to implement the entire project function within the project goal; N is a positive integer greater than 2; A first integration unit is configured to start from the first tutorial block in the teaching process and gradually send each tutorial block to the learning user in a learning order until the Nth tutorial block is learned; Among them, when the learning user completes the i-th tutorial block, check whether the learning user has completed the teaching goal of the i-th tutorial block. If the goal is achieved, continue to send the i+1-th tutorial block until the learning user completes N tutorial blocks; i is a positive integer.

8. The device according to claim 7, characterized in that The first generating unit specifically includes: A project learning step generating unit, configured to generate project learning steps according to the project goal and the user's knowledge background; A sub-step splitting unit, configured to split the project learning step into a plurality of sub-steps according to the learning sequence; A tutorial block generation unit, configured to split a sub-step into multiple tutorial blocks based on multiple step targets of each sub-step; The integration unit is used to integrate the tutorial blocks of all sub-steps according to the learning sequence to obtain the teaching process.

9. A project learning device, characterized in that, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the learning method for the project according to any one of claims 1 to 6 is implemented.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores instructions, and when the instructions are executed on a terminal device, the terminal device executes the project learning method according to any one of claims 1 to 6.

Images