Virtual teaching training method and system based on large language model, and medium

Abstract

The application discloses a virtual teaching training method and system based on a large language model and a medium. The virtual teaching training method based on the large language model comprises the following steps: determining preset characteristic parameters of virtual students according to teaching levels and subject contents; constructing a large language model, generating virtual students based on the large language model according to the preset characteristic parameters of the virtual students, and inputting multi-modal teaching information; the virtual students perform model training and learning on the multi-modal teaching information, and generate learning state information and feedback information; a virtual teaching expert proposes a suggestion of dynamically adjusting and optimizing a teaching strategy based on an evaluation result; dynamic learning and interactive feedback are performed on virtual students with different prior knowledge levels and cognitive abilities, a virtual expert system is constructed to perform real-time evaluation and individualized improvement suggestions on a teaching process, teaching strategies are continuously optimized for a user, and the teaching quality of a specific subject is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent education, in particular to a virtual teaching training method and system based on a large language model and a medium. BACKGROUND

[0002] Existing intelligent education technology mainly uses computer simulation and virtual reality for teacher training. In the field of virtual teaching training, systems such as TeachLivE and Mursion mainly use mixed reality technology to create a simulated classroom environment, providing teachers with risk-free teaching and behavior management training, simulating real classroom interactions by controlling virtual student roles, thereby helping teachers improve their teaching and management skills.

[0003] The existing technology has the following problems:

[0004] 1) Limitation of professional knowledge: Current intelligent education technology systems mainly focus on classroom management and general teaching skill training, and do not support the evaluation of teaching effectiveness for specific subject professional knowledge. Therefore, users may have difficulty evaluating and adjusting teaching methods and strategies for specific subjects when using such systems.

[0005] 2) Non-resettable: In existing intelligent education platforms, once virtual students begin education training, their learning history and accumulated knowledge cannot be simply cleared or reset to the initial state. This memory allows virtual students to adjust their behavior and responses based on previous interactions and learning outcomes.

[0006] 3) Lack of interaction: Existing systems use actors to simulate virtual student responses to increase the naturalness of interaction, but this method still relies on pre-set scripts, actor improvisation, or backstage operator settings. This may affect the personalization and targeting of teaching, especially in situations that require customized teaching for different student needs.

[0007] 4) Limitations of expert evaluation: Existing technology lacks automated expert evaluation functions, making it difficult to systematically analyze and provide real-time feedback on teaching effectiveness and to provide objective improvement suggestions based on the entire teaching process and student learning performance. SUMMARY

[0008] The purpose of the present application is to provide a virtual teaching training method based on a large language model to solve the problems mentioned in the background.

[0009] To achieve the above-mentioned application purpose, the first technical solution adopted by the present application is: a virtual teaching training method based on a large language model, comprising the following steps:

[0010] S1. Based on the teaching level and subject content, determine the preset characteristic parameters of the virtual student. The preset characteristic parameters of the virtual student include prior knowledge level, cognitive level and learning style.

[0011] S2, construct a large language model, generate virtual students based on the pre-defined feature parameters of virtual students, and input multimodal teaching information;

[0012] S3, based on multimodal teaching information input, allows virtual students to learn and update their learning status, and generates feedback information based on their learning status;

[0013] S4. The feedback information is analyzed and evaluated using a teaching expert system based on a large language model to obtain the evaluation results.

[0014] S5, based on the evaluation results, outputs adjustment and optimization strategies, and transmits the adjustment and optimization strategies to the terminal in real time.

[0015] Furthermore, in step S1, the prior knowledge level includes high school and university levels.

[0016] The prior knowledge levels at different stages, such as undergraduate or master's degree students, are used to simulate students' basic knowledge reserves.

[0017] The cognitive abilities mentioned include different cognitive levels such as basic, intermediate and advanced, to reflect students' ability to process information and learn;

[0018] The learning styles include three different learning styles: visual learners, auditory learners, and kinesthetic learners, making the behavior of virtual students closer to real personalized learning methods.

[0019] Furthermore, in step S2, the steps for constructing the large language model are as follows:

[0020] Select and train a large language model, such as the GPT-4 model or the Kimi model;

[0021] Based on prior knowledge level, cognitive level and learning style, construct corresponding prior knowledge level dataset, cognitive level dataset and learning style dataset;

[0022] Create prompt words, and select matching data combinations from the prior knowledge level dataset, cognitive level dataset, and learning style dataset based on the prompt words;

[0023] After setting the parameters, input the teaching information to train the model (simulate the teaching process), allowing the virtual student to train and learn the teaching materials (simulate the learning process), and output feedback records in real time. The parameters include knowledge level, cognitive level, and learning style.

[0024] According to the feedback record, the model parameters are dynamically adjusted, and a final large language model is generated.

[0025] Further, according to the teaching level and the subject content, characteristic subject teaching materials are established, and virtual students dynamically learn and interactively feedback according to the provided specific subject teaching materials, and the specific subject teaching materials include texts, images or videos.

[0026] Further, based on the evaluation result, an adjustment and optimization strategy is adjusted and optimized, and the adjustment and optimization strategy is transmitted to the terminal in real time, specifically including:

[0027] The evaluation result is obtained, and the evaluation result is compared with the set condition information to obtain a teaching effectiveness evaluation value;

[0028] It is judged whether the teaching effectiveness evaluation value is less than or equal to the set teaching effectiveness threshold value;

[0029] If it is less than or equal to the teaching effectiveness threshold value, adjustment information for optimizing the teaching strategy is generated, and the adjustment information is transmitted to the terminal in real time;

[0030] If it is greater than the teaching effectiveness threshold value, the advantages and merits of the existing teaching strategy are summarized and transmitted to the terminal for reference by the user.

[0031] To achieve the above-mentioned purposes, the second technical solution adopted by the present application is: a virtual teaching training system based on a large language model, applied to a virtual teaching training method of a large language model, including:

[0032] A virtual student module, the virtual student module constructs a virtual student model, and determines preset characteristic parameters of the virtual student according to the teaching level and the subject content;

[0033] A virtual expert module, the virtual expert module is electrically connected with the virtual student module, the virtual expert module performs real-time evaluation on the teaching effect based on the teaching materials and the student feedback information, and proposes corresponding suggestions;

[0034] A hardware device, the hardware device is electrically connected with the virtual student module and the virtual expert module, and the hardware device is used for collecting texts, voices and images, and performing multi-modal interaction with the virtual student model.

[0035] Further, the hardware device includes a computer, a microphone and a camera.

[0036] The present application also provides a computer readable storage medium, which stores a computer program, and the computer program enables a computer to execute to realize the virtual teaching training method based on the large language model according to any one of the above.

[0037] Compared with the prior art, the application has the following advantages: the application uses virtual students with different prior knowledge levels and cognitive abilities to simulate dynamic learning and interactive feedback, and uses a virtual expert system to evaluate the teaching process in real time and provide personalized improvement suggestions, helping users continuously optimize teaching strategies and improve the teaching quality of specific disciplines. BRIEF DESCRIPTION OF DRAWINGS

[0038] Figure 1 A flowchart of a virtual teaching training method based on a large language model is shown;

[0039] Figure 2 A flowchart of a large language model construction of a virtual teaching training method based on a large language model is shown;

[0040] Figure 3 A block diagram of a virtual teaching training system based on a large language model is shown. DETAILED DESCRIPTION

[0041] In order 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 described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should be within the scope of protection of the present application.

[0042] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily describe a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or system including a series of steps or units does not necessarily limit to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products or systems.

[0043] It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict. The present application will be described in detail below with reference to the drawings and in combination with the embodiments.

[0044] As Figure 1 shown, the present application provides a virtual teaching training method based on a large language model, including the following steps:

[0045] S1, determining preset characteristic parameters of virtual students according to teaching levels and subject contents, the preset characteristic parameters of virtual students including prior knowledge levels, cognitive levels and learning styles;

[0046] S2, constructing a large language model, generating virtual students based on the large language model according to the preset characteristic parameters of virtual students, and inputting multi-modal teaching information;

[0047] S3, the virtual students performing model training and learning on the multi-modal teaching information to generate learning state information and feedback information;

[0048] S4, analyzing and evaluating the feedback information by using a teaching expert system based on the large language model to obtain an evaluation result;

[0049] S5, outputting adjustment and optimization strategy suggestions based on the evaluation result, and transmitting the adjustment and optimization strategy suggestions to a terminal in real time.

[0050] It should be noted that, by simulating virtual students with different prior knowledge levels and cognitive abilities to dynamically learn and interact, and by using a virtual expert system to real-time evaluate and provide personalized improvement suggestions for the teaching process, the user can continuously optimize the teaching strategy and improve the teaching quality of a specific subject. By using the latest large language model and natural language processing technology, this method can create virtual students that can simulate different prior knowledge levels and cognitive abilities. These virtual students can dynamically learn and interact based on the specific subject teaching materials (such as text, images, videos, etc.) provided by the user, thereby supporting the user to continuously test and improve their teaching materials and teaching strategies in a risk-free and resettable simulation environment. The virtual teaching expert system can real-time evaluate based on the teaching process and student learning feedback, and provide personalized improvement suggestions to help the user better understand the teaching effect and optimize the teaching method.

[0051] Further, the present application is suitable for teachers and education researchers to develop virtual student models to improve the teaching effect of specific subjects, and is suitable for educational institutions (such as universities, primary and secondary schools) and training centers, especially in the aspects of teaching material optimization, teaching strategy testing, teaching effect feedback, etc.

[0052] According to the embodiment of the present application, in step S1, the prior knowledge level includes prior knowledge levels at different stages of high school, undergraduate or master's graduate, to simulate the basic knowledge reserve of students;

[0053] The cognitive ability includes different cognitive levels of basic, intermediate and advanced to reflect the ability of students to process information and learn;

[0054] The learning styles include three different learning styles of visual learners, auditory learners and kinesthetic learners, so that the behaviors of the virtual students are closer to the real personalized learning mode.

[0055] As shown in Figure 2 According to the embodiment of the present application, in step S2, the steps of constructing a large language model are as follows:

[0056] S201, selecting and training a large language model, the large language model including GPT-4 model or Kimi model, etc.;

[0057] S202, constructing a data set of prior knowledge level, cognitive level and learning style based on prior knowledge level, cognitive level and learning style;

[0058] S203, establishing a prompt word, selecting a data combination conforming to the preset parameters from the data set of prior knowledge level, cognitive level and learning style to construct the prompt word;

[0059] S204, training the large language model using the prompt word;

[0060] S205, testing the model using a test data set of the corresponding level (for example, test questions of the corresponding subject), and adjusting the prompt word according to the feedback of the model and the test results, so that the virtual student model meets the preset level.

[0061] According to the embodiment of the present application, the characteristic subject teaching materials are established according to the teaching level and the subject content, and the virtual student dynamically learns and interacts according to the provided specific subject teaching materials, and the specific subject teaching materials include text, image or video.

[0062] It should be noted that when constructing a virtual expert, the large model can be trained in parallel using the prompt word and the teaching history. The content is "you are a teaching expert, please give the evaluation and suggestion of the last teaching according to the above teaching materials and student feedback record", which makes the large language model understand that the next task is to evaluate the previous teaching process and make improvement suggestions. The prompt word is sent to the model together with the message record of the previous teaching process, so that the model can have a global understanding of the teaching process, and the last teaching evaluation and suggestion are returned.

[0063] According to the embodiment of the present application, the evaluation is carried out based on the teaching history, and the optimization suggestion of the teaching strategy is given, and the real-time is transmitted to the terminal, specifically including:

[0064] Obtaining an evaluation result, comparing the evaluation result with the set condition information, and obtaining a teaching effectiveness evaluation value;

[0065] Judging whether the teaching effectiveness evaluation value is less than or equal to the set teaching effectiveness threshold value;

[0066] If less than or equal to the teaching effectiveness threshold, generate adjustment information for teaching strategy optimization, and transmit the adjustment information to the terminal in real time;

[0067] If greater than the teaching effectiveness threshold, summarize the advantages and merits of the existing teaching strategy, and transmit to the terminal for user reference in real time.

[0068] As Figure 3 To achieve the above purposes, the second technical solution adopted by the present application is: a virtual teaching training system based on a large language model, applied to a virtual teaching training method of a large language model, comprising:

[0069] A virtual student module, the virtual student module constructs a virtual student model, and determines the preset feature parameters of the virtual student according to the teaching level and the subject content;

[0070] A virtual expert module, the virtual expert module is electrically connected with the virtual student module, the virtual expert module performs real-time evaluation on the teaching effect based on the teaching materials and the student feedback information, and proposes corresponding suggestions;

[0071] A hardware device, the hardware device is electrically connected with the virtual student module and the virtual expert module, and the hardware device is used for collecting text, voice and image, and performing multi-modal interaction with the virtual student model.

[0072] Application examples:

[0073] Case one: A junior high school mathematics teacher wants to test the effect of different teaching schemes. Through the virtual student system, the teacher simulates virtual students with corresponding prior knowledge level, uploads teaching materials (such as PPT or teaching materials), obtains the learning reaction of virtual students, and evaluates the effectiveness of each method. At the same time, the system provides real-time feedback to help the teacher adjust the teaching scheme in real time. By combining the suggestions given by the system, the teacher can improve the overall quality of the teaching scheme after repeated experiments.

[0074] Case two: The teaching development department of a university wants to implement a new teacher training system to improve the teaching ability of teachers. Through the virtual student system, the teacher simulates college students (high school graduates, freshmen, sophomores, etc.) with corresponding prior knowledge level, uploads teaching materials (such as PPT or voice or video), and obtains feedback. After the interaction is completed, the system also provides evaluation and suggestions to help the teacher improve the quality of professional knowledge teaching.

[0075] According to the embodiment of the present application, the hardware device includes a computer, a microphone and a camera.

[0076] The application further provides a computer-readable storage medium, which stores a computer program, and the computer program enables a computer to execute to implement the large language model-based virtual teaching training method of any one of the above.

[0077] In summary, the application continuously optimizes the teaching strategy and improves the teaching quality of a specific subject by simulating virtual students with different prior knowledge levels and cognitive abilities to perform dynamic learning and interactive feedback, and simultaneously using a constructed virtual expert system to perform real-time evaluation and personalized improvement suggestions on the teaching process.

[0078] Those skilled in the art can understand that, for the convenience of description, one is set as an example for the number of memories and processors. In actual terminals or servers, there can be multiple processors and memories. The memory can also be referred to as a storage medium or a storage device, and the embodiments of the application do not limit this.

[0079] It should be understood that, in the embodiments of the application, the processor can be a central processing unit (CPU), and the processor can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. The processor can also be a general-purpose microprocessor, a graphics processing unit (GPU) or one or more integrated circuits for executing relevant programs to perform the functions required by the embodiments of the application.

[0080] The processor can also be an integrated circuit chip having a processing capability. In implementation, various steps of the present application can be completed by integrated logic circuits or by instructions in the form of software in the processor. The processor can implement or execute the disclosed methods, steps and logic block diagrams in the embodiments of the present application. The steps of the methods disclosed in the embodiments of the present application can be directly embodied as hardware coding processing in the processor, or a combination of hardware and software modules in the processor. The software modules can be located in the storage medium, the flash memory, the read-only memory (ROM), the programmable read-only memory (PROM), the electrically programmable read-only memory (EPROM), the electrically erasable programmable read-only memory (EEPROM), the registers, or other mature storage mediums in the art. The storage medium is located in the storage of the memory, and the processor reads information in the memory to complete the functions of the units included in the methods, systems and storage mediums in the embodiments of the present application by means of hardware.

[0081] It should also be understood that the memory mentioned in the embodiments of the present application can be a volatile memory or a non-volatile memory, or can include both volatile and non-volatile memories. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM) or a flash memory. The volatile memory can be a random access memory (RAM) used as an external cache.

[0082] By way of example, and not limitation, many forms of RAM can be used, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM) and direct rambus random access memory (DR RAM).

[0083] The memory can also be a Compact Disc Read-Only Memory (CD-ROM), or other optical storage, optical disc storage (including compact disc, laser disc, optical disc, digital versatile disc, Blu-ray disc, etc.), magnetic storage media, or any other medium that can be used to carry or store desired program codes in the form of instructions or data structures and that can be accessed by a computer, but is not limited to this. The memory can exist independently and be connected to the processor through the bus. The memory can also be integrated with the processor, and the memory can store programs, and when the programs stored in the memory are executed by the processor, the processor is used to execute the steps of the determination method in the above embodiments of the present application.

[0084] It should be noted that when the processor is a general processor, DSP, ASIC, FPGA or other programmable logic device, discrete gate or transistor logic device, discrete hardware component, the memory (storage module) is integrated in the processor. It should be noted that the memory described herein is intended to include but not limited to these and any other suitable type of memory.

[0085] It should be understood that the term "and / or" herein is only to describe the association relationship of the associated objects, which means that there can be three relationships, for example, A and / or B can represent the three cases of A alone, A and B together, and B alone. In addition, the character " / " herein generally represents that the front and rear associated objects are in an "or" relationship.

[0086] In the implementation process, each step of the above method can be completed by the integrated logic circuit of the hardware in the processor or the instruction in the form of software. The steps of the method disclosed in combination with the embodiments of the present application can be directly embodied as hardware processor execution completion, or executed by hardware and software modules in the processor. The software module can be located in a mature storage medium in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, register, etc. The storage medium is located in the memory, and the processor reads the information in the memory, and combines the hardware to complete the steps of the above method. To avoid repetition, it will not be described in detail here.

[0087] Those of ordinary skill in the art can realize that the various illustrative logical blocks (ILB) and steps described in connection with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. The implementation choice depends on particular application and design constraints imposed on the technical solution. Those of ordinary skill can use different methods to implement the described functions for each particular application, but such implementation should not be considered beyond the scope of the present application.

[0088] In the above embodiments, all or part can be realized by software, hardware, firmware or any combination thereof. When realized by software, all or part can be realized in the form of a computer program product. The computer program product includes one or more computer instructions. When loaded and executed on a processor, the computer program instructions produce all or part of the processes or functions according to the embodiments of the present application. The computer can be a general-purpose computer, a computer network, or other programmable device.

[0089] The embodiments also provide a computer readable storage medium storing a computer program, which causes a computer to execute to realize the above-mentioned virtual teaching training method based on a large language model.

[0090] It should be noted that the computer instructions can be stored in a computer readable storage medium, or transmitted from one computer readable storage medium to another, for example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center through wired (such as coaxial cable, optical fiber) or wireless (such as infrared, wireless, microwave, etc.) mode, or transmitted from one website, computer, server or data center to a mobile phone processor through a wired mode. The computer readable storage medium can be any available medium that can be accessed by a computer or a data storage system such as a server, data center, etc. integrated with one or more available media. The available media can be magnetic media (such as floppy disk, hard disk), optical media (such as DVD), or semiconductor media (such as solid state disk) and the like.

[0091] Finally, it should be noted that the above is only the preferred embodiment of the present application and is not intended to limit the present application. Although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or equivalently replace some of the technical features, as long as they are within the spirit and principles of the present application. Any modification, equivalent replacement, improvement, etc. made within the scope of the present application should be included in the protection scope of the present application.

Claims

1. A method for virtual teaching training based on a large language model, characterized in that, Comprising the following steps: S1, determining preset virtual student characteristic parameters according to teaching levels and subject content, the preset virtual student characteristic parameters including prior knowledge level, cognitive level and learning style; S2, constructing a large language model, generating a virtual student based on the large language model according to the preset virtual student characteristic parameters, and inputting multi-modal teaching information; S3, the virtual student conducts model training and learning on the multi-modal teaching information, and generates learning state information and feedback information; S4, using a teaching expert system based on a large language model to analyze and evaluate the input teaching information and the feedback information of the virtual student, and obtaining an evaluation result; S5, outputting adjustment and optimization strategy suggestions based on the evaluation result, and transmitting the adjustment and optimization strategy suggestions to the terminal in real time; In step S2, the steps of constructing a large language model are as follows: Select and train a large language model, the large language model including a GPT-4 model or a Kimi model; Based on the prior knowledge level, cognitive level and learning style, the corresponding prior knowledge level dataset, cognitive level dataset and learning style dataset are constructed; Establish a prompt word, and select a matching data combination from the prior knowledge level dataset, cognitive level dataset and learning style dataset according to the prompt word; After setting parameters, input teaching information to train the model, let the virtual student conduct model training and learning on the teaching materials, and output feedback records in real time, the parameters including knowledge level, cognitive level and learning style; According to the feedback record, dynamically adjust the model parameters to generate the final large language model; According to the teaching level and the subject content, the characteristic subject teaching materials are established, and the virtual student conducts dynamic learning and interactive feedback according to the provided specific subject teaching materials, the specific subject teaching materials including text, image or video.

2. The virtual teaching training method based on a large language model of claim 1, wherein, In step S1, the prior knowledge level includes prior knowledge levels at different stages of high school, undergraduate or master's graduate, to simulate the basic knowledge reserve of students; The cognitive level includes different cognitive levels of basic, medium and high, to reflect the ability of students to process information and learn; The learning style includes three different learning styles of visual learners, auditory learners and kinesthetic learners, so that the behavior of the virtual student is closer to the real personalized learning way. 3.The virtual teaching training method based on a large language model of claim 1, wherein, Based on the evaluation result, adjustment and optimization strategy suggestions are outputted, and the adjustment and optimization strategy suggestions are transmitted to the terminal in real time, specifically including: Obtain the evaluation result, compare the evaluation result with the set condition information, and obtain the teaching effectiveness evaluation value; Determine whether the teaching effectiveness evaluation value is less than or equal to the set teaching effectiveness threshold value; If less than or equal to the teaching effectiveness threshold value, generate adjustment information for optimizing the teaching strategy, and transmit the adjustment information to the terminal in real time; If greater than the teaching effectiveness threshold value, summarize the advantages and merits of the existing teaching strategy, and transmit to the terminal for reference by the user in real time.

4. A large language model-based virtual teaching training system applied to the large language model-based virtual teaching training method of any one of claims 1-3, characterized in that, Comprising: A virtual student module for constructing a virtual student model and determining preset virtual student characteristic parameters according to teaching levels and subject content; A virtual expert module is electrically connected with the virtual student module, and the virtual expert module performs real-time evaluation on teaching effect based on teaching materials and student feedback information and proposes corresponding suggestions; A hardware device is electrically connected with the virtual student module and the virtual expert module, and the hardware device is used for collecting text, voice and image and interacting with the virtual student model in a multi-modal manner.

5. The large language model based virtual teaching training system of claim 4, wherein, The hardware device includes a computer, a microphone and a camera.

6. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a computer program, and the computer program enables the computer to execute to realize the virtual teaching training method based on the large language model in any one of claims 1-3.

Images