GraphRAG-based teacher teaching courseware writing method and device

Abstract

The invention provides a GraphRAG-based teacher teaching courseware writing method and apparatus. The method comprises the steps of obtaining basic information of teaching courseware; based on the constructed teaching knowledge graph, retrieving multi-dimensional knowledge points associated with the basic information through GraphRAG; and inputting the multi-dimensional knowledge points into a pre-trained writing prompt depth model to obtain courseware writing prompt information. According to the method and the device, the multi-dimensional knowledge points associated with the basic information of the teaching courseware are obtained by constructing the teaching knowledge graph, and then the courseware writing prompt information is generated through the deep convolution model, so that a user can complete the writing work of the courseware based on the prompt information, omission is avoided, and the writing speed is increased.

Description

Technical Field

[0001] This application relates to the field of deep model processing technology, and more specifically, to a method and apparatus for writing teaching courseware based on GraphRAG. Background Technology

[0002] Currently, when teachers compile teaching materials, they mainly rely on manual design and experience accumulation. When faced with the structuring, logical organization, and interdisciplinary integration of knowledge content, it is often time-consuming and prone to omissions. Summary of the Invention

[0003] To address the aforementioned issues, the first aspect of this application provides a method for developing teaching materials for teachers based on GraphRAG, which includes:

[0004] Obtain basic information about the teaching materials;

[0005] Based on the constructed teaching knowledge graph, GraphRAG is used to retrieve multi-dimensional knowledge points associated with basic information;

[0006] By inputting multi-dimensional knowledge points into a pre-trained deep learning model for writing prompts, we can obtain prompts for writing courseware.

[0007] The second aspect of this application provides a system for creating teacher teaching courseware based on GraphRAG, which includes:

[0008] The information acquisition module is used to acquire basic information about the teaching materials.

[0009] The graph retrieval module is used to retrieve multi-dimensional knowledge points associated with basic information based on the constructed teaching knowledge graph through GraphRAG.

[0010] A generation module is developed, which is used to input multi-dimensional knowledge points into a pre-trained deep model for writing prompts and obtain courseware writing prompt information.

[0011] A third aspect of this application provides an electronic device, including: a memory and a processor; the memory being configurable to store a program, and the processor being coupled to the memory for executing the program in the memory for:

[0012] Obtain basic information about the teaching materials;

[0013] Based on the constructed teaching knowledge graph, GraphRAG is used to retrieve multi-dimensional knowledge points associated with basic information;

[0014] By inputting multi-dimensional knowledge points into a pre-trained deep learning model for writing prompts, we can obtain prompts for writing courseware.

[0015] A fourth aspect of this application provides a computer-readable storage medium having a computer program stored thereon, the program being executed by a processor to implement the aforementioned GraphRAG-based method for writing teacher teaching courseware.

[0016] In this application, a teaching knowledge graph is constructed to obtain multi-dimensional knowledge points associated with the basic information of teaching courseware. Then, a deep convolutional model is used to generate courseware writing prompts. In this way, users can complete the courseware writing work based on the prompts, thereby avoiding omissions and improving the writing speed. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating a GraphRAG-based method for creating teaching courseware for teachers according to an embodiment of this application.

[0018] Figure 2 This is an architecture diagram of the code generation model for the teacher teaching courseware writing method based on GraphRAG according to an embodiment of this application;

[0019] Figure 3 This is an architectural diagram of a GraphRAG-based teacher teaching courseware writing device according to an embodiment of this application;

[0020] Figure 4 This is an architectural diagram of an electronic device according to an embodiment of this application. Detailed Implementation

[0021] To make the above-mentioned objects, features, and advantages of this application more apparent and understandable, specific embodiments of this application will be described in detail below with reference to the accompanying drawings. Although exemplary embodiments of this application are shown in the drawings, it should be understood that this application can be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of this application and to fully convey the scope of this application to those skilled in the art.

[0022] It should be noted that, unless otherwise stated, the technical or scientific terms used in this application shall have the ordinary meaning as understood by one of ordinary skill in the art.

[0023] This application provides the above-described method for writing teacher teaching courseware based on GraphRAG. The specific solution of this method is as follows: Figures 1-2 As shown, this method can be executed by a GraphRAG-based teacher instructional material writing device, which can be integrated into electronic devices such as computers, servers, computer clusters, and data centers. Combined with... Figure 1As shown, the method for creating teacher teaching courseware based on GraphRAG includes:

[0024] S101, Obtain basic information about the teaching materials;

[0025] Among them, the basic information of the teaching courseware allows teachers to input preliminary teaching topics or courseware outlines. Teachers can input basic information such as teaching topics, objectives, student characteristics, and existing teaching outlines through text or voice.

[0026] S102, based on the constructed teaching knowledge graph, retrieves multi-dimensional knowledge points associated with basic information through GraphRAG;

[0027] S103: Input multi-dimensional knowledge points into the pre-trained deep model for writing prompts to obtain courseware writing prompts.

[0028] In this application, a teaching knowledge graph is constructed to obtain multi-dimensional knowledge points associated with the basic information of teaching courseware. Then, a deep convolutional model is used to generate courseware writing prompts. In this way, users can complete the courseware writing work based on the prompts, thereby avoiding omissions and improving the writing speed.

[0029] The main solution of this application is as follows: First, teachers input preliminary teaching topics or courseware outlines; then, the system uses the GraphRAG mechanism to retrieve relevant knowledge and cases from massive educational and subject content using knowledge graphs, and inputs them into the generative model; after pre-training and fine-tuning, the generative model outputs text with knowledge associations and teaching idea prompts; finally, the prompts are optimized by the feedback and optimization module and displayed to teachers in an interactive manner, supporting teachers to adjust and modify their needs in real time.

[0030] Among them, GraphRAG (Graph-based Retrieval-Augmented Generation) is an enhanced retrieval generation framework that combines knowledge graphs.

[0031] In one implementation, the step of retrieving multi-dimensional knowledge points associated with basic information based on the constructed teaching knowledge graph via GraphRAG includes:

[0032] The basic information is parsed to obtain the parsed data;

[0033] Obtain the constructed instructional knowledge graph;

[0034] Based on the teaching knowledge graph, the parsed data is mapped into multi-dimensional knowledge points.

[0035] This involves searching for nodes and related paths in the knowledge graph that are relevant to the input topic, thereby obtaining multi-dimensional knowledge point information.

[0036] This involves transforming teachers' unstructured, natural input into structured, semantic representations that machines can understand and process.

[0037] In this process, the user's query intent (data parsing) is "anchored" to the knowledge graph and its surrounding connections are explored, thereby uncovering crucial contextual knowledge that the user has not explicitly stated.

[0038] In this context, multi-dimensional knowledge information is no longer a scattered list of knowledge points, but a networked knowledge structure.

[0039] In one implementation, mapping parsed data into multi-dimensional knowledge points based on a teaching knowledge graph includes:

[0040] Build a knowledge point database;

[0041] Constructing a teaching knowledge graph;

[0042] Based on semantic similarity matching, data is analyzed to identify knowledge point nodes and associated paths;

[0043] Based on the matched knowledge point nodes, associated paths, and knowledge point database, multi-dimensional knowledge points are determined.

[0044] Specifically, specialized teaching knowledge graphs are constructed for major subject areas and interdisciplinary knowledge points. These graphs include course objectives, teaching cases, knowledge connections, and logical structures.

[0045] In one implementation, parsing the basic information to obtain parsed data includes:

[0046] The basic information is transformed to generate a high-dimensional fusion feature vector;

[0047] Multilayer perceptron processing is performed on the high-dimensional fused feature vector to obtain a combined vector;

[0048] The combined vectors are enhanced with several features to obtain analytical data.

[0049] In this application, the basic information includes: classification features, numerical features, and text features.

[0050] The specific process of converting the basic information package in this application is as follows:

[0051] Classification features: Discrete values ​​are mapped to low-dimensional dense vectors through an embedding layer (nn.Embedding).

[0052] Numerical characteristics: Numerical differences are handled through normalization (such as BatchNorm1d).

[0053] Text features: Text vectors are generated using a pre-trained model (such as BERT) and then dimensionality is reduced using a linear layer.

[0054] Then, the feature vectors from the three parts are concatenated into a high-dimensional fused feature vector. This high-dimensional fused feature vector is a fixed-length vector representing a comprehensive representation of the basic information.

[0055] In this application, a multilayer perceptron (MLP) is used to perform nonlinear transformation on the high-dimensional fused feature vector to extract higher-level features; the MLP consists of multiple fully connected layers and activation functions (such as ReLU).

[0056] In this application, the combined vector is a feature vector processed by MLP, which has a lower dimension.

[0057] In this application, a feature enhancement module (such as Mamba Block or self-attention mechanism) is used to enhance the combined vector multiple times to capture more complex feature interactions. Each enhancement includes: channel attention: calculating the global mean and generating weight coefficients to highlight important features; multi-scale transformation: generating new feature representations through fully connected layers; and weighted fusion: combining the original input with the transformation results to enhance the feature representation. The final generated personalized feature vector serves as the user profile vector.

[0058] In one implementation, the step of inputting multi-dimensional knowledge points into a pre-trained deep model for writing prompts to obtain courseware writing prompt information includes:

[0059] By integrating basic information and multi-dimensional knowledge points, a fused descriptive information is obtained;

[0060] Obtain pre-trained deep model writing prompts;

[0061] The fused description information is input into the pre-trained deep model for writing prompts to obtain prompts for writing courseware.

[0062] In this process, the retrieved knowledge information is integrated with the original input to form an enhanced input.

[0063] In this process, the pre-trained model is fine-tuned and combined with enhanced input to generate corresponding courseware prompts.

[0064] In this process, semantic vectors are used to retrieve auxiliary information related to the teaching topic within the knowledge graph, and this information is then fused with the original input.

[0065] Among them, a deep Transformer pre-trained model (after fine-tuning) is used to process the enhanced input and generate multi-round, hierarchical prompts for courseware writing.

[0066] In one implementation, combined with Figure 2As shown, the step of inputting the fused description information into the pre-trained deep model for writing prompts to obtain courseware writing prompts includes:

[0067] The fused descriptive information is embedded to obtain embedded features;

[0068] Multi-head self-attention processing is applied to the embedded features to obtain the first attention features;

[0069] The primary attention features are subjected to multi-layer perceptual processing to obtain perceptual features;

[0070] Continuous multi-head self-attention processing and activation processing are performed on the perceptual features to obtain prompts for courseware creation.

[0071] This involves converting human-readable natural language (words, sentences) into machine-understandable numerical forms (i.e., vectors / tensors) while preserving semantic information.

[0072] Among them, the first attention feature allows the model to simultaneously focus on the relationships between different parts of the input sequence, thereby gaining a deep understanding of the context.

[0073] Among them, after the attention mechanism integrates global information, the perceptual features are further processed and transformed by a powerful nonlinear function to extract higher-order and more abstract features.

[0074] In one implementation, it further includes:

[0075] Display the courseware creation prompts to the public;

[0076] Acquire viewer's eye-tracking data;

[0077] Based on eye-tracking data, the prompts for creating the courseware were optimized.

[0078] The initial prompts generated were optimized, and the suggestions were further refined to ensure that the prompts were highly instructive and practical.

[0079] In this application, the quality of the "prompt information" generated by the courseware is indirectly evaluated and optimized by observing how students (viewers) consume the courseware content.

[0080] Among them, the courseware writing prompts displayed to the public will show the courseware content finally generated from the prompts.

[0081] In this context, under controlled experimental conditions or with the availability of widely available hardware in the future, eye-tracking devices (such as bar-shaped devices installed under computer screens or built-in eye trackers in VR / AR devices) can be used to non-invasively collect students' eye movement data while learning courseware.

[0082] The system can generate two versions of courseware, A and B, using two different sets of prompts for the same teaching objective. Eye-tracking data analysis is used to determine which version of the courseware shows "healthier" eye-tracking data, thus identifying which prompt is superior.

[0083] In one implementation, the courseware writing prompts are optimized based on eye-tracking data, including:

[0084] Based on eye trackers to acquire users' eye movement data;

[0085] Based on the user's eye-tracking data, multiple focal areas are determined on the screen;

[0086] Optional prompts for creating courseware are displayed in each focused area;

[0087] Monitor users' real-time eye movement data;

[0088] Determine the true focusing area based on the user's real-time eye movement data;

[0089] The optimal prompts for creating courseware should be determined based on the actual focus area.

[0090] In this application, an eye tracker is used to track the user's gaze, and multiple "option areas" are preset on the screen. Each area displays a prompt message. The prompt message of the area that the user looks at for the longest time is automatically selected as the "optimal".

[0091] In this way, user preferences can be obtained through implicit feedback rather than explicit ratings (such as clicking a button), resulting in better evaluation results.

[0092] In this application, the user's eye movement data also includes: fixation point: the point where the line of sight is relatively stationary; fixation duration: the time spent on a certain point or area; pupil diameter: pupil dilation is sometimes positively correlated with cognitive load, interest or difficulty; saccade path: the trajectory of the line of sight moving between different points.

[0093] In this application, an algorithmic model is established to analyze raw eye-tracking data. This algorithmic model may include:

[0094] Interest / Preference Recognition: The user's gaze is always long on a certain prompt option, and the pupils dilate slightly before the gaze moves away and becomes calm.

[0095] Confusion level identification: The user repeatedly scans back and forth on a certain option or keyword, with a scattered gaze, accompanied by frequent blinking or erratic pupil changes;

[0096] Reading pattern recognition: Did the user read most of the prompt information in a smooth and regular order (indicating good readability) or did they give up after only reading the beginning (indicating insufficient appeal)?

[0097] In this application, eye-tracking data is an extremely valuable input source, used as contextual information for understanding user intent and state.

[0098] In one implementation, it further includes:

[0099] Teachers evaluate and provide feedback on the prompts, the system collects the feedback data, and updates and optimizes the prompt depth model online, forming a closed loop of continuous learning and iterative prompt generation.

[0100] In this way, the system's output (user feedback) is used as training data to continuously optimize the system's performance. This transforms a static, one-off tool into a dynamic, evolving intelligent system.

[0101] It should be noted that the existing technology has the following shortcomings:

[0102] Insufficient personalization: Traditional template or rule-based tools cannot generate personalized courseware content suggestions based on teachers' subject characteristics, teaching styles, and students' needs;

[0103] Weak knowledge connections: Single keyword matching methods are insufficient to cover cross-disciplinary and cross-knowledge domain related content, making it difficult to form a systematic teaching approach;

[0104] Lack of real-time interaction: Existing tools lack real-time feedback and dynamic adjustment mechanisms, and cannot optimize content based on teachers' real-time input.

[0105] Insufficient deep semantic understanding: When modeling complex teaching logic and knowledge levels, traditional generative models often struggle to take into account multi-source information and contextual semantics.

[0106] In this application, knowledge graphs and graph structure retrieval are used to achieve deep association and intelligent retrieval of teaching knowledge points, curriculum system and interdisciplinary content;

[0107] In this application, a pre-trained and fine-tuned generative model is used in conjunction with real-time retrieval results to output courseware writing prompts with high semantic relevance and personalized customization.

[0108] In this application, a real-time interactive and continuously optimized system platform was constructed to realize closed-loop intelligent feedback from the input of the teacher's initial ideas to the generation of complete prompt information;

[0109] This application reduces the time cost, content omissions, and logical inconsistencies for teachers in the courseware preparation process.

[0110] This application provides a GraphRAG-based teaching courseware writing device for executing the GraphRAG-based teaching courseware writing method described above. The following is a detailed description of the GraphRAG-based teaching courseware writing device.

[0111] like Figure 3 As shown, the GraphRAG-based teacher teaching material writing device includes:

[0112] Information acquisition module 101 is used to acquire basic information of teaching courseware;

[0113] The graph retrieval module 102 is used to retrieve multi-dimensional knowledge points associated with basic information based on the constructed teaching knowledge graph through GraphRAG.

[0114] A generation module 103 is written, which is used to input multi-dimensional knowledge points into a pre-trained deep model for writing prompts to obtain courseware writing prompt information.

[0115] In one implementation, the map retrieval module 102 is further configured to:

[0116] The basic information is parsed to obtain parsed data; the constructed teaching knowledge graph is obtained; and based on the teaching knowledge graph, the parsed data is mapped into multi-dimensional knowledge points.

[0117] In one implementation, the map retrieval module 102 is further configured to:

[0118] Construct a knowledge point database; construct a teaching knowledge graph; analyze data and knowledge point nodes and associated paths based on semantic similarity matching; determine multi-dimensional knowledge points based on the matched knowledge point nodes, associated paths, and knowledge point database.

[0119] In one implementation, the map retrieval module 102 is further configured to:

[0120] The basic information is transformed to generate a high-dimensional fusion feature vector; the high-dimensional fusion feature vector is processed by multi-layer perception to obtain a combined vector; the combined vector is enhanced several times to obtain analytical data.

[0121] In one implementation, the generation module 103 is further configured to:

[0122] The basic information and multi-dimensional knowledge points are integrated to obtain integrated description information; a pre-trained deep model for writing prompts is obtained; the integrated description information is input into the pre-trained deep model for writing prompts to obtain courseware writing prompt information.

[0123] In one implementation, the generation module 103 is further configured to:

[0124] The fused description information is embedded to obtain embedded features; the embedded features are processed by multi-head self-attention to obtain first attention features; the first attention features are processed by multi-layer perception to obtain perceptual features; and the perceptual features are processed by continuous multi-head self-attention and activation to obtain courseware writing prompts.

[0125] In one implementation, an optimization module is also included, which is used to:

[0126] The system displays prompts for creating the courseware; acquires eye-tracking data from viewers; and optimizes the prompts based on the eye-tracking data.

[0127] The GraphRAG-based teacher teaching courseware writing device provided in the above embodiments of this application corresponds to the GraphRAG-based teacher teaching courseware writing method provided in the embodiments of this application. Therefore, the specific content in this system corresponds to the GraphRAG-based teacher teaching courseware writing method. The specific content can be referred to the records in the GraphRAG-based teacher teaching courseware writing method, which will not be repeated in this application.

[0128] The GraphRAG-based teacher teaching courseware writing device and the GraphRAG-based teacher teaching courseware writing method provided in the above embodiments of this application are based on the same inventive concept and have the same beneficial effects as the methods adopted, run or implemented by the applications stored therein.

[0129] The above describes the internal functions and structure of the GraphRAG-based teacher instructional material creation device, such as... Figure 4 As shown, in practice, the GraphRAG-based teacher teaching material writing device can be implemented as an electronic device, including: a memory 301 and a processor 303.

[0130] Memory 301 can be configured to store a program.

[0131] Additionally, memory 301 can also be configured to store various other data to support operation on the electronic device. Examples of this data include instructions for any application or method used to operate on the electronic device, contact data, phonebook data, messages, pictures, videos, etc.

[0132] Memory 301 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. Processor 303, coupled to memory 301, is used to execute programs in memory 301 for:

[0133] Obtain basic information about the teaching materials;

[0134] Based on the constructed teaching knowledge graph, GraphRAG is used to retrieve multi-dimensional knowledge points associated with basic information;

[0135] By inputting multi-dimensional knowledge points into a pre-trained deep learning model for writing prompts, we can obtain prompts for writing courseware.

[0136] In one implementation, the processor 303 is further configured to:

[0137] The basic information is parsed to obtain parsed data; the constructed teaching knowledge graph is obtained; and based on the teaching knowledge graph, the parsed data is mapped into multi-dimensional knowledge points.

[0138] In one implementation, the processor 303 is further configured to:

[0139] Construct a knowledge point database; construct a teaching knowledge graph; analyze data and knowledge point nodes and associated paths based on semantic similarity matching; determine multi-dimensional knowledge points based on the matched knowledge point nodes, associated paths, and knowledge point database.

[0140] In one implementation, the processor 303 is further configured to:

[0141] The basic information is transformed to generate a high-dimensional fusion feature vector; the high-dimensional fusion feature vector is processed by multi-layer perception to obtain a combined vector; the combined vector is enhanced several times to obtain analytical data.

[0142] In one implementation, the processor 303 is further configured to:

[0143] The basic information and multi-dimensional knowledge points are integrated to obtain integrated description information; a pre-trained deep model for writing prompts is obtained; the integrated description information is input into the pre-trained deep model for writing prompts to obtain courseware writing prompt information.

[0144] In one implementation, the processor 303 is further configured to:

[0145] The fused description information is embedded to obtain embedded features; the embedded features are processed by multi-head self-attention to obtain first attention features; the first attention features are processed by multi-layer perception to obtain perceptual features; and the perceptual features are processed by continuous multi-head self-attention and activation to obtain courseware writing prompts.

[0146] In one implementation, the processor 303 is further configured to:

[0147] The system displays prompts for creating the courseware; acquires eye-tracking data from viewers; and optimizes the prompts based on the eye-tracking data.

[0148] In this application, Figure 4 The diagram only shows some components and does not mean that the electronic device includes only these components. Figure 4 The components shown.

[0149] The electronic device provided in this embodiment is based on the same inventive concept as the GraphRAG-based teacher teaching courseware writing method provided in this application embodiment, and has the same beneficial effects as the methods adopted, run or implemented by the applications stored therein.

[0150] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-readable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0151] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 The computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to operate in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0152] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0153] In a typical configuration, a computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory. Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0154] This application also provides a computer-readable storage medium corresponding to the GraphRAG-based teacher teaching courseware writing method provided in the foregoing embodiments, wherein a computer program (i.e., program product) is stored thereon. When the computer program is run by a processor, it executes the interactive image analysis assistance method for 3D aerial imaging provided in any of the foregoing embodiments.

[0155] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information by any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0156] The computer-readable storage medium provided in the above embodiments of this application and the interactive image analysis assistance method for 3D aerial imaging provided in the embodiments of this application are based on the same inventive concept and have the same beneficial effects as the methods adopted, run or implemented by the applications stored therein.

[0157] It should be noted that numerous specific details are set forth in the specification provided herein. However, it will be understood that embodiments of this application may be practiced without these specific details. In some instances, well-known structures and techniques have not been shown in detail so as not to obscure the understanding of this specification.

[0158] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0159] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for creating teacher teaching courseware based on GraphRAG, characterized in that, include: Obtain basic information about the teaching materials; Based on the constructed teaching knowledge graph, multi-dimensional knowledge points associated with basic information are retrieved through GraphRAG; By inputting multi-dimensional knowledge points into a pre-trained deep learning model for writing prompts, we can obtain prompts for writing courseware.

2. The method for writing teacher teaching courseware based on GraphRAG according to claim 1, characterized in that, The aforementioned method, based on the constructed teaching knowledge graph, uses GraphRAG to retrieve multi-dimensional knowledge points associated with basic information, including: The basic information is parsed to obtain the parsed data; Obtain the constructed instructional knowledge graph; Based on the teaching knowledge graph, the parsed data is mapped into multi-dimensional knowledge points.

3. The method for writing teacher teaching courseware based on GraphRAG according to claim 2, characterized in that, The aforementioned teaching knowledge graph-based approach maps parsed data into multi-dimensional knowledge points, including: Build a knowledge point database; Constructing a teaching knowledge graph; Based on semantic similarity matching, data is analyzed to identify knowledge point nodes and associated paths; Based on the matched knowledge point nodes, associated paths, and knowledge point database, multi-dimensional knowledge points are determined.

4. The method for writing teacher teaching courseware based on GraphRAG according to claim 2, characterized in that, The process of parsing the basic information to obtain parsed data includes: The basic information is transformed to generate a high-dimensional fusion feature vector; Multilayer perceptron processing is performed on the high-dimensional fused feature vector to obtain a combined vector; The combined vectors are enhanced with several features to obtain analytical data.

5. The method for developing teacher teaching courseware based on GraphRAG according to any one of claims 1-4, characterized in that, The process of inputting multi-dimensional knowledge points into a pre-trained deep learning model for writing prompts yields courseware writing prompts, including: By integrating basic information and multi-dimensional knowledge points, a fused descriptive information is obtained; Obtain pre-trained deep model writing prompts; By inputting the fused description information into the pre-trained deep model for writing prompts, we can obtain prompts for writing courseware.

6. The method for writing teacher teaching courseware based on GraphRAG according to claim 5, characterized in that, The process of inputting fused descriptive information into a pre-trained deep model for writing prompts yields courseware writing prompts, including: The fused descriptive information is embedded to obtain embedded features; Multi-head self-attention processing is applied to the embedded features to obtain the first attention features; The primary attention features are subjected to multi-layer perceptual processing to obtain perceptual features; Continuous multi-head self-attention processing and activation processing are performed on the perceptual features to obtain prompts for courseware creation.

7. The method for developing teacher teaching courseware based on GraphRAG according to any one of claims 1-4, characterized in that, Also includes: Display the courseware creation prompts to the public; Acquire viewer's eye-tracking data; Based on eye-tracking data, the prompts for creating the courseware were optimized.

8. A teacher's teaching courseware writing device based on GraphRAG, characterized in that, include: The information acquisition module is used to acquire basic information about the teaching materials. The graph retrieval module is used to retrieve multi-dimensional knowledge points associated with basic information based on the constructed teaching knowledge graph through GraphRAG. A generation module is developed, which is used to input multi-dimensional knowledge points into a pre-trained deep model for writing prompts and obtain courseware writing prompt information.

9. An electronic device, characterized in that, include: Memory and processor; The memory is used to store programs; The processor, coupled to the memory, is used to execute the program for: Obtain basic information about the teaching materials; Based on the constructed teaching knowledge graph, multi-dimensional knowledge points associated with basic information are retrieved through GraphRAG; By inputting multi-dimensional knowledge points into a pre-trained deep learning model for writing prompts, we can obtain prompts for writing courseware.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, The program is executed by the processor to implement the GraphRAG-based teacher teaching courseware writing method according to any one of claims 17.

Images