Interactive rich media textbook and courseware design method based on artificial intelligence

Abstract

The invention discloses an interactive rich media textbook and courseware design method based on artificial intelligence, and relates to the field of textbook system research and development, and the method comprises the steps that a resource center provides required resources for a classroom creating and sharing module and a teaching and research ecological platform; the classroom creating and sharing module introduces a virtual simulation scene, a 3D model, a model hotspot, an interactive scene, VR experience and a classroom record based on resources provided by the resource center, designs teaching courseware in combination with teacher experience, and forms a man-machine collaborative personalized teaching mode; the teaching and research ecological platform carries out emotion calculation and social analysis on responses of students by utilizing data mining and artificial intelligence analysis technologies based on a personalized teaching mode, analyzes the response degrees of the students to teaching courseware, and generates a visual knowledge point graph according to resources provided by the resource center; the visual knowledge point graph is used for covering the whole teaching link of pre-class preview, in-class interaction and after-class review, personalized teaching is achieved, and the whole teaching process is throughout.

Description

Technical Field

[0001] This application relates to the field of textbook system research and development, and particularly to a method for designing an interactive rich media textbook and courseware based on artificial intelligence. Background Art

[0002] Currently, certain developments have been made in the method for designing an interactive rich media textbook and courseware based on artificial intelligence (AI) and big data. Such systems usually digitize the content of traditional paper textbooks and embed rich media resources (such as micro-lesson videos, document materials, and PPT courseware), and provide rich learning resources and interactive experiences through online courses and digital service platforms.

[0003] Although certain achievements have been made in the prior art, although AI can analyze students' learning habits, it is still challenging to achieve true personalized teaching; moreover, the technology of AI analysis throughout the entire process before, during, and after class still needs to be improved. Summary of the Invention

[0004] The purpose of this application is to provide a method for designing an interactive rich media textbook and courseware based on artificial intelligence, so as to solve the problems that the existing method for designing an interactive rich media textbook and courseware based on artificial intelligence cannot achieve personalized teaching and cannot penetrate the entire process before, during, and after class.

[0005] To achieve the above purpose, this application provides the following solutions:

[0006] In a first aspect, this application provides a method for designing an interactive rich media textbook and courseware based on artificial intelligence, including: a resource center, a creative classroom module, and a teaching and research ecological platform;

[0007] Based on the resource center, provide the required resources to the creative classroom module and the teaching and research ecological platform;

[0008] Based on the creative classroom module, utilize the resources provided by the resource center, introduce virtual simulation scenarios, 3D models, model hotspots, interactive scenarios, VR experiences, and classroom recordings, and combine with teachers' experience to design teaching courseware, forming a personalized teaching mode of human-computer collaboration;

[0009] Based on the teaching and research ecological platform, according to the personalized teaching mode, using data mining and artificial intelligence analysis technologies to conduct emotional computing and social analysis on students' responses, analyzing the degree of students' responses to the teaching courseware, and generating a visualized knowledge point map according to the resources provided by the resource center; the visualized knowledge point map includes a course map, a professional map, and a job skill map; the visualized knowledge point map is used to cover the entire teaching process of pre-class preview, in-class interaction, and after-class review.

[0010] According to the specific embodiments provided by the present application, the present application discloses the following technical effects:

[0011] The present application not only integrates basic multimedia such as pictures, audio, and video, but also breakthroughly introduces advanced elements such as virtual simulation scenarios, 3D models, model hotspots, interactive scenarios, VR experiences, and classroom recordings, greatly enriching the expression form and interactivity of teaching materials content, and significantly enhancing students' learning interest and understanding depth.

[0012] Using AI analysis technology to conduct more accurate student data analysis, comprehensively considering factors such as students' mental state, learning environment, and social background; introducing emotional computing and social analysis technologies to enhance AI's ability to understand and respond to interpersonal relationships; combining teachers' experience to form a personalized teaching mode of human-computer collaboration.

[0013] Based on data mining and artificial intelligence analysis technologies, analyzing the degree of students' responses to the teaching courseware, and generating a visualized knowledge point map to cover the entire teaching process such as pre-class preview, in-class interaction, and after-class review, improving the teaching quality and effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required to be used in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application, and those of ordinary skill in the art can obtain other drawings without creative efforts based on these drawings.

[0015] Figure 1 It is a schematic flowchart of the design method of an AI-based interactive rich media teaching material and courseware provided by the present application.

[0016] Figure 2 It is an architecture diagram of an educational digital teaching material system constructed by the design method of an AI-based interactive rich media teaching material and courseware provided by the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0017] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0018] To make the above objects, features, and advantages of the present application more obvious and understandable, the present application will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0019] The present application provides an artificial intelligence-based interactive rich media teaching material and courseware design method. The system uses the B / S architecture style, the backend uses the microservices architecture style, and the software architecture style is MVVM; the front-end framework uses vuejs + typescript.

[0020] As Figure 1 - Figure 2 shown, the artificial intelligence-based interactive rich media teaching material and courseware design method includes the following steps.

[0021] S1: Based on the resource center, provide the required resources to the creative classroom module and the teaching and research ecological platform.

[0022] S2: Based on the creative classroom module, utilize the resources provided by the resource center to introduce virtual simulation scenarios, 3D models, model hotspots, interactive scenarios, VR experiences, and classroom recordings, and combine with teachers' experience to design teaching courseware, forming a personalized teaching mode of human-computer collaboration.

[0023] S3: Based on the teaching and research ecological platform, according to the personalized teaching mode, use data mining and artificial intelligence analysis technologies to perform emotional computing and social analysis on students' responses, analyze the response degree of students to the teaching courseware, and generate a visual knowledge point map according to the resources provided by the resource center; the visual knowledge point map includes a course map, a professional map, and a job skill map; the visual knowledge point map is used to cover the entire teaching process of pre-class preview, in-class interaction, and after-class review.

[0024] The present application also includes a registration and login module; the platform supports three user roles, namely administrator, teacher, and student; each role corresponds to different data permissions.

[0025] In practical applications, users can register by scanning the code, log in using an account and password, or scan the WeChat code.

[0026] In an exemplary embodiment, the resource center specifically includes: creative classroom resources and teaching and research ecological resources.

[0027] In practical applications, all uploaded resources in the resource center are labeled. The purpose of labeling is to enrich the feature codes of the resources (such as name, classification, tags, description, etc.); then the feature codes are vectorized for subsequent resource recommendation; for the resources of other subsystems, a low-coupling timing update method is adopted, and in principle, no intrusion is allowed; the resource changes of other subsystems are monitored regularly. If there are new resources, resource pulling and vectorization are performed to continuously expand the resources in the resource center and form normalized and unified resources.

[0028] The creative classroom resources include recommended resources, my resources, team resources, and question bank center.

[0029] The recommended resources are system resources. The system resources include classification feature codes, tag feature codes, and introduction feature codes; the method for designing an interactive rich media teaching material and courseware based on artificial intelligence vectorizes the user behavior according to the user's resource type preference and behavior record, and compares the vectorized user behavior with the feature vectors of the system resources to determine the system resources preferred by the user; the user behavior includes the tags searched, the resource feature codes searched, and the search conditions.

[0030] In practical applications, the recommended resources are system resources. Each resource will have relevant feature codes such as classification, tags, and introduction; the system will record the user's preference for the resource type used on the platform; the behavior records of relevant information such as the tags frequently searched by the user, the resource feature codes searched, and the search conditions are made, and then these user behaviors are vectorized and compared with the vectorization of the system resource features for search; thus, it is analyzed that the user tends to preferentially recommend certain types of resources or resources in certain fields; the vector database uses the Milvus database, and the text vectorization technology uses the Embedding vector model to generate 1024-dimensional vector data for storage in the database; vector search uses the principle of the shortest cosine distance in Milvus for search.

[0031] My resources are personal exclusive resources; users can classify the resources they own into folders and upload files; the file types include pictures, audio, video, 2D interaction, 3D resources, VR resources, and virtual simulation scenarios.

[0032] In practical applications, my resources are personal and exclusive resources; the resources can be classified into folders, and files can be uploaded by oneself; the file types include: pictures, audio, video, 2D interaction, 3D, VR resources, and virtual simulation, a total of seven types of resources; when a user uploads a file, it will be automatically classified according to the file suffix format; only pictures, audio, video, and 3D resources are supported for uploading files; the remaining classified resources are imported resources; 3D resources are 3D model resources, supporting three formats: fbx, glb, and gltf; after the user uploads 3D resources, the system uses 3D rendering technology and GIF technology to automatically generate a GIF image of the model rotating one week; thus, when the model is displayed, there is both a dynamic effect and the performance overhead is further reduced; the 3D rendering technology is implemented using webgl + threejs, and the GIF generation technology uses gif.js; the file storage uses Qiniu Cloud Storage, and the files are CDN accelerated; VR resources use 720 panoramic rendering technology, and according to the pictures uploaded by the user, threejs is used for panoramic rendering to form a 720 panoramic view; 2D interaction and virtual simulation resources use an external link form, and ifram is used for embedded preview in the system to achieve the convenience of cross-platform and cross-system resources.

[0033] The team resources are team-shared resources; the team resources can be edited and deleted by the administrator and the resource uploader, and the remaining members within the team can save the team resources to their own folders.

[0034] In practical applications, team resources, also known as team-shared resources; team resources are defined in the system as resources shared by all users of the current team; regardless of who uploads the resources, as long as they are placed under team resources, all users in this team can share them; at the same time, only the administrator and oneself can edit and delete the resources; the resources in the team can be saved to one's own folder.

[0035] In practical applications, the question bank center currently supports subjective questions and objective questions. As the question data source for digital textbooks, teachers can manually add questions; they can also quickly batch import questions using an Excel template; they can also use an AI large model or documents provided by users to automatically generate multiple questions that meet the requirements.

[0036] For each question, the system records the error rate of the question. If this value is higher than a preset threshold, it will be pushed to the teacher as a key question, providing a detailed statistical analysis report to help students and teachers understand the learning situation and teaching effect.

[0037] For short answer questions, an AI large model is used to score the short answer questions. After three rounds of scoring, the average value is taken as the final score rate.

[0038] The teaching and research ecological resources can obtain required resources from the Chuangxiang classroom resources, and the teaching and research ecological resources also include a custom resource library and a custom question library.

[0039] In actual applications, the custom resource library contains documents and pictures uploaded by users, uses Qiniu Cloud as OSS, and uses CDN to accelerate file access to ensure the user experience.

[0040] In addition, the custom resource library also includes a custom question bank. Users can add custom questions to the custom question bank, or use the AI ​​big model to analyze and extract the content based on the user's wishes or uploaded files, and generate multiple test questions that meet the requirements.

[0041] In an exemplary embodiment, the creative classroom module 2 specifically includes: a digital intelligence courseware module, a digital intelligence e-book module, a digital intelligence teaching material module and a 3D model editor.

[0042] The digital courseware module is used to utilize SVG technology and WebRTC technology to import resources from the resource center while editing PPT online, and to generate teaching courseware using PPT animation effects.

[0043] In actual applications, the digital courseware module is similar to the PPT editing function in Windows Office. Through software technology, the traditional desktop office software functions are migrated to the web page system.

[0044] The digital courseware module can realize PPT online editing while seamlessly importing various images, videos, audio, 2D, 3D and other resources from the resource center, and then produce PPT animation special effects to form an online teaching courseware. The system introduces Scalable Vector Graphics (SVG) technology to support online rendering of high-quality graphics and animation effects, and uses Web Real-Time Communication (WebRTC) technology to support preview and editing of real-time video and audio resources.

[0045] Using WebAssembly technology, the PPT parsing and rendering engine is embedded in the Web application, supporting the online import and editing of local PPT files.

[0046] The PDF.js open source library is introduced to achieve compatibility and conversion of PDF format courseware. Based on the original PPT content, public resources in the system resource library are introduced to enrich the courseware content.

[0047] During the PPT production process, the system integrates an animation effect library, providing a variety of preset animation templates that users can easily apply to the PPT. It supports customizing animation paths, speeds, trigger conditions, etc., meeting personalized teaching needs.

[0048] Using the Canvas API, it realizes the real-time drawing and preview of complex animation effects. All types of courseware are stored in Qiniu Cloud Object Storage Service (OSS) using real-time streaming transmission. By introducing the content delivery network (CDN) acceleration function of OSS, it ensures the fast access and distribution of courseware globally, forming a high-performance and directly editable cloud-based courseware.

[0049] After the courseware is made, you can directly select the student classes associated with the current account for online teaching. The system uses the Netty high-performance and asynchronous event-driven network application framework to achieve real-time transmission of teaching data.

[0050] The intelligent e-book module is used to introduce intelligent interaction functions, video interaction functions, canvas functions, and screen recording functions based on the intelligent courseware module, using DeepSpeech technology, MMAction2 action recognition framework, Two-Stream ConvNets model, and MediaRecorder API technology.

[0051] In practical applications, the intelligent e-book module introduces functions such as intelligent interaction, video interaction, canvas, and screen recording on the basis of intelligent courseware. Intelligent interaction includes adding voice following questions, requiring students to use the microphone of the terminal device to record following voice data. The background server uses DeepSpeech technology to perform text recognition on the collected voice data, and uses the spaCy modern natural language processing library for sentiment analysis, extracts the key information of the voice data, and then calculates the score of this sub-question according to the preset scoring rules of the question.

[0052] Video interaction is to upload or introduce a video resource in the e-book and distribute it to students, requiring students to imitate and learn the actions in the video and then be scored.

[0053] Through the action recognition framework of the Multimedia Action Recognition Toolbox (Version 2) (OpenMMLab Action Recognition Toolbox and Benchmark Version 2, MMAction2), combined with the Two-Stream ConvNets model, the system performs skeleton action recognition and captures dynamic features of the temporal flow for the videos introduced in the e-books and the videos recorded by the students' cameras, and calculates the scores of the students based on the recognized data.

[0054] The system also supports the canvas function and adopts a responsive design to ensure that the canvas function can be well adapted on both mobile and computer terminals.

[0055] The canvas function uses the Konva.js technology for teachers to draw and create complex graphics and animations on the page to explain complex knowledge points and various evolution processes to students.

[0056] For the need to demonstrate the verification process of certain exercises, the system also provides a screen recording function, which uses the MediaRecorder application programming interface (API) technology to record media streams (such as videos and audios) and present them to students as e-book content.

[0057] The digital textbook module is used to online edit the teaching courseware based on the AI large model, combined with the LTP technology and the pre-trained model.

[0058] In practical applications, the digital textbook module is similar to the Word editing function in Windows Office and can edit textbooks online.

[0059] AI Xiaoyun is introduced in the teaching materials. The large AI model can be used to polish, correct errors, continue writing, infer context, etc. for the teaching material content. It can also use the large AI model to analyze the meaning of the currently selected content, and then combine with the Language Technology Platform (LTP) to perform Chinese word segmentation and dependency syntax analysis on the data in the question bank center. Finally, the best-matched data of the two are integrated, and the best recommended questions for the selected content in the teaching material are inserted into the teaching material. According to the chapter information and other texts provided in the teaching material, a pre-trained model - Bidirectional Encoder Representations from Transformers (BERT) is used for keyword extraction and summary generation. Then, based on the extracted key data, the entire teaching material is converted into a knowledge graph to realize the change of data form. After the teaching material is produced, qrcode-generator is used to generate a QR code and share it with students. Scanning the code with a mobile phone can preview the teaching material content.

[0060] Real-time collaborative editing technology:

[0061] Digital intelligence courseware, digital intelligence e-books, and digital intelligence teaching materials can all support online real-time editing and multi-person real-time collaborative editing.

[0062] Execution entities: users (perform editing operations), clients (capture editing operations and convert them into data), and servers (receive data and synchronize it to all collaborators).

[0063] Implementation method: When a user performs an editing operation on the client, the client immediately captures these operations and converts them into data instructions. Subsequently, the client sends these data instructions to the server through real-time communication technology (such as WebSocket). After receiving the data, the server updates the content of the cloud document and broadcasts the updated content to the clients of all collaborators through the same communication technology, realizing a seamless real-time collaborative editing experience.

[0064] The 3D model editor uses Threejs 3D graphics rendering technology to edit 3D models.

[0065] In practical applications, the 3D model editor:

[0066] Uses Threejs 3D graphics rendering technology to edit 3D models; supports formats such as gla, gltf, and fbx.

[0067] The functions mainly include: generating gif animated images, setting hotspots, explosion effects, and scene rendering.

[0068] The generation of Gif animations uses gif.js and interpolates frames for the images in the canvas, obtaining 45 frames with a 75 - millisecond delay to generate smooth animations.

[0069] Hotspot setting: Using threejs technology, obtain the child nodes in the model, and attach hotspots to all child nodes. The attached hotspot information includes: text descriptions, pictures, audio, videos, etc.; further introduce a certain part of the model in detail; and in this process, adopt AI technology to further optimize the text descriptions, pictures, videos, etc.

[0070] Explosion effects: Use a self - developed physical simulation particle system to simulate and implement various explosion effects and fluid effects.

[0071] Scene rendering: Use threejs technology to render complex 3D scenes, which supports multiple lighting models and shadow casting, and renders multiple materials and textures, providing rich visual effects for 3D scene rendering.

[0072] In an exemplary embodiment, the teaching and research ecological platform specifically includes: a curriculum map module, a professional map module, and a job skill map module.

[0073] The curriculum map module includes a knowledge source map unit, a teaching intervention unit, and a teaching data unit, which are used to generate a knowledge source map, create teaching plans, and summarize learning data.

[0074] The professional map module includes a professional overview unit, a professional curriculum topology map unit, and an interdisciplinary integration exploration unit, which are used to overview professional information, generate a professional curriculum topology map, and provide an interdisciplinary integration exploration function.

[0075] The job skill map module includes a job management unit, a skill library unit, a student skill assessment unit, a learning plan formulation unit, and a job development path planning unit, which are used to generate job courses, create a skill library, conduct skill assessments based on the majors and courses studied by students, formulate learning results based on skill assessment results, and plan job development paths.

[0076] In an exemplary embodiment, the knowledge source map unit uses G6 as a graph visualization engine to describe the self - coordinates, shapes, and inter - relationships of each knowledge point; uses swiper as a content touch tool to build an interaction channel between users and knowledge points; uses a force - directed layout algorithm to formulate the arrangement rules of knowledge points and determine the positions of knowledge point nodes; uses horizontal and vertical depth - first search algorithms to intelligently identify and aggregate knowledge points with inherent relevance and allocate learning weeks; for a single course with more than a set number of knowledge points, adopt a GPU hardware acceleration scheme to improve rendering efficiency.

[0077] In practical applications, the knowledge source map unit describes the self - coordinates, shapes, and inter - relationships of each knowledge point. The swiper, as a content touch - control tool, constructs an interaction channel between users and knowledge points. The knowledge point arrangement rule uses the Force - Directed Layout algorithm to determine the positions of nodes, making the connections between knowledge points more intuitive. When allocating study weeks, the horizontal and vertical depth - first search algorithm (Depth - first search, DFS) is used, enabling the system to intelligently identify and aggregate knowledge points with inherent relevance and reasonably allocate them to the same study week. For a single course with a large number of knowledge points, due to the lag during rendering caused by hardware performance, a GPU hardware acceleration solution is adopted to improve the rendering efficiency and ensure smooth page loading.

[0078] In an exemplary embodiment, the teaching intervention unit notifies the created teaching plans, exams, and assignments to the corresponding student terminals; among them, the teaching plans are automatically rotated according to the study weeks using a timed task framework, and different knowledge points are pushed to students every week; after students learn the knowledge points, based on the collected student learning data, Apache Spark is used to analyze the learning data, find the students' weak knowledge points, and push questions or knowledge points with a relevance higher than a set threshold to the students for learning.

[0079] In practical applications, data such as teaching plans, exams, and assignments created in the teaching intervention unit are all notified to the corresponding student terminals. The teaching plans are automatically rotated according to the study weeks using a timed task framework, so that different knowledge points are pushed to students every week; after students learn the knowledge points, based on the student learning data collected by the system, Apache Spark is used to analyze the learning data, find the students' weak knowledge points, and push highly relevant questions or knowledge points to the students for learning to improve their learning level.

[0080] In an exemplary embodiment, the teaching data unit calculates the overall variance of the students' grades in each class; for the learning data of any student, the number of questions answered during the exam and the student's exam scores each time are recorded, and the Pearson correlation coefficient is used to analyze the relationship between the number of questions answered by the student and the scores to determine the student's learning efficiency; at the same time, the Spearman rank correlation coefficient is used to examine the relationship between the final exam scores and the usual scores of the students to comprehensively understand the students' learning situation.

[0081] In practical applications, the teaching data unit is a summary statistic of all the learning data generated by students in the above teaching interventions. According to different learning situations, in addition to the regular class grade ranking list, the system also calculates the Population Variance of the students' grades in each class to understand the central tendency of the class grade data. For the presentation of a student's teaching data, the number of questions x answered by the student during the exam and the student's exam scores y each time are recorded in the system. The Pearson Correlation Coefficient is used to analyze the relationship between the number of questions the student answers and their grades, so as to understand the student's learning efficiency. At the same time, the Spearman Rank Correlation Coefficient is used to examine the relationship between the student's final exam grades and their usual grades, achieving a comprehensive understanding of the student's learning situation.

[0082] In an exemplary embodiment, the professional overview unit includes aspects of practical experience, professional construction, and construction achievements to describe professional information.

[0083] In the aspect of practical experience, combining TimelineJS technology, TAIlwind CSS layout method, and GSAP technology, a timeline or narrative display method is adopted to show the key events and important turning points in the origin and development process of the profession.

[0084] In practical applications, in the aspect of practical experience, a timeline or narrative display technique is adopted to vividly show the origin of the profession and the key events and important turning points in its development process.

[0085] Through TimelineJS, remarkable visual effects are presented. At the same time, in order to ensure that the website can have a good display effect on various devices, TAIlwind CSS is used for style layout, realizing a highly customizable user interface.

[0086] In addition, animations and transition effects are added. Using GSAP (GreenSock Animation Platform) technology, high-performance transition animations are achieved, further enhancing the user experience.

[0087] In the aspect of professional construction, D3.js is used to display charts of the technical teaching staff; Lodash technology is used to process the data of the teaching staff charts.

[0088] In practical applications, in the aspect of professional construction, this application details key elements such as the teaching staff, teaching facilities, and scientific research platforms of the profession.

[0089] The chart display of the faculty strength uses D3.js technology to achieve a highly customized and interactive chart application. For complex faculty chart data, we use Lodash to process it.

[0090] In terms of the construction achievements, the Swiper component is used to dynamically display the construction achievements in a carousel, and the one-click sharing function is used to share the construction achievements.

[0091] In terms of practical applications and construction results, this application summarizes the outstanding achievements of the profession in academic research, technological innovation, and social services.

[0092] In order to better display these achievements, this application uses a web sliding plug-in (Swiper component) to achieve a dynamic carousel display of construction results.

[0093] For construction results that need to be shared, this application uses the one-click sharing function provided by the English sharing plug-in (ShareThis), which supports multiple social platforms to facilitate users to share and disseminate.

[0094] In an exemplary embodiment, the professional course topology map unit uses the G6 graph visualization engine in the form of a canvas to generate a professional course topology map of all courses of any major in the form of nodes and display it on the page; for majors with a number of courses greater than a set threshold, the Intersection Observer API lazy loading technology is used to delay the loading of content outside the viewport; at the same time, the GPU hardware acceleration solution is enabled to ensure that complex pages can also be loaded smoothly; when a search is required, the Lodash filter method is used to filter course data.

[0095] In practical applications, the professional course topology map unit uses the canvas format and the G6 graph visualization engine to clearly display all courses of a major in the form of nodes on the page.

[0096] The topology diagram is equipped with a comprehensive editing toolbar to facilitate users to customize the properties of course nodes.

[0097] For majors with a large number of courses, the system uses Intersection ObserverAPI lazy loading technology to delay the loading of content outside the viewport, thereby increasing the page loading speed; at the same time, the GPU hardware acceleration solution is enabled to ensure that complex pages can also load smoothly.

[0098] When there are a large number of courses that need to be searched, the system uses Lodash's filter method to filter the data and implement efficient search functions.

[0099] In the process of building a professional map, the system will intelligently recommend relevant content based on user behavior, and with the help of the machine learning model TensorFlow.js, help users quickly build a professional map. The course nodes on the canvas support dragging operations, which are implemented by vue-draggable; and the relationship between nodes is represented by drawing connecting lines using the link function of D3.js. Each node and the connecting line together constitute the basic elements on the canvas.

[0100] In addition, the system also supports the implementation of semester sidebars in the canvas through Vue Teleport, horizontally displaying the semester to which the course belongs; circle graphic elements are used to mark the required or optional attributes of the course. If you need to quickly and automatically arrange elements on the canvas, the system uses D3.js's force-directed layout (d3.forceSimulation) and hash coordinate distribution algorithm to automatically assign element coordinates. Finally, users can use jsPDF to export the entire professional map as an image or PDF file.

[0101] In practical applications, cross-disciplinary integration exploration units are used to provide users with cross-field learning suggestions and case analyses, encouraging students to explore connections and cooperation opportunities between different disciplines.

[0102] Based on the students' test and wrong question records in the course map, the Python Pandas library is used to clean the original data, eliminate invalid information, and the data is analyzed with the help of the machine learning model Scikit-Learn, so as to provide personalized learning suggestions and teaching resources.

[0103] The learning path planning function customizes interdisciplinary personalized learning paths based on students' personal information, learning background, courses taken and study time, including course schedule, study time and evaluation criteria.

[0104] To this end, this application uses Python's Scrapy framework to crawl course information and industry trends from the Internet, and then uses the Pandas library to process the collected data, including filling missing values, correcting outliers, and storing the cleaned data in the relational database MySQL.

[0105] The system uses an item-based collaborative filtering algorithm to recommend courses based on the user's past learning records. At the same time, the Dijkstra algorithm is used to combine the course prerequisites and difficulty levels to plan the optimal learning path and present it to the user through Charts. In addition, the system background will record the user's clicks, browsing, learning time and other behavioral data for subsequent analysis and optimization.

[0106] In the case analysis section, users can customize and upload case videos, and store them through Alibaba Cloud OSS and share them on the Internet.

[0107] This application uses Meta tags, Sitemap and other technical means to optimize the page search engine ranking and increase software popularity.

[0108] In terms of user interaction, the system integrates the Disqus comment system, allowing users to post comments and feedback, and is connected to the Vue-RateIt rating plug-in to facilitate users to rate cases.

[0109] In an exemplary embodiment, the position management unit is designed to help schools or students understand the detailed information, skill requirements and related course settings of different positions. For the position list data with more reading and less writing, the system uses Redis cache technology to improve the system response speed and reduce server pressure. At the same time, in order to support the two-way many-to-many binding of positions and skills, based on the G6 graph visualization engine in the canvas, the data binding (DataBinding) rules are customized to realize the binding of multiple nodes (multipleNode) and multiple edges (multipleEdge), so as to describe complex multilateral node relationships.

[0110] In an exemplary embodiment, the skill library unit uses Redis cache technology and a recursive algorithm, combined with a multi-level parent-child structure, to describe the progressive relationship of skill classification.

[0111] In actual applications, skill library units include skill classification, skill description, etc. For skill library data that is read more and written less, the system also uses Redis cache technology. For multi-level parent-child structured skill classification, the central control server uses a recursive algorithm to return a "tree" structure data to describe the progressive relationship of skill classification. Skill descriptions are automatically generated using AI big model technology, or after manual input by users, the AI ​​big model is used to help users polish and rewrite.

[0112] In an exemplary embodiment, the student skill assessment unit conducts skill assessment and ranking based on the students' majors and courses, supports multiple assessment methods such as teacher evaluation and peer evaluation, and provides multi-dimensional feedback.

[0113] When calculating the student skill evaluation score, we first use the peer evaluation scores as the basis to calculate the standard deviation (StandardDeviation). If this value is higher than the system preset threshold (indicating that the degree of dispersion of the peer evaluation scores is large), the weight of the peer evaluation scores will be dynamically reduced, and the weight of the teacher's evaluation will be correspondingly increased. Finally, the weighted average (WeightedMean) of the two is calculated to obtain a student skill score that is as objective as possible.

[0114] When generating a new score ranking each time, in order to enable students to receive data updates in a timely manner, the central control server uses SSE technology to achieve real-time update of ranking data to the client.

[0115] In an exemplary embodiment, the learning plan formulation unit manually or automatically generates a learning plan according to the skill assessment results of students.

[0116] When arranging the learning plan, students with similar skill levels and learning habits are analyzed through the Collaborative Filtering algorithm, and the learning tasks and resources they all like are recommended.

[0117] In addition, the Apriori algorithm is adopted: it is used to mine the association rules between learning tasks and resources, find out which learning tasks and resources are often selected by students together, so as to recommend a suitable learning plan.

[0118] In an exemplary embodiment, the job development path planning unit aims to help students understand information such as the core skills and internship opportunities of different jobs, and enhance students' confidence and planning ability for future career development by showing statistical data on the employment situation of graduates, such as the employment rate and average salary level.

[0119] The data analysis tool Pandas library is used to process and analyze the data stored in the MySQL database, extract data information on various dimensions of the job, such as salary trends, employment rate trends, and job demand for the number of people, and the visualization tool Matplotlib is used to intuitively present the analysis results in the form of charts, graphs, etc., to help students better understand job data.

[0120] This application also establishes a strict data protection mechanism to ensure the privacy and security of students' data. Encryption technology, anonymization processing and other means are adopted to prevent data leakage and abuse.

[0121] AI technology is used for more accurate student data analysis, comprehensively considering factors such as students' mental state, learning environment and social background. Emotional computing and social analysis technologies are introduced to enhance AI's ability to understand and respond to interpersonal relationships. Combining with teachers' intuition and experience, a personalized teaching mode of human-machine collaboration is formed.

[0122] Develop a preview mechanism applicable to before class, a teacher lesson preparation mechanism, a classroom interaction mechanism, a multi-channel media collaboration mechanism, an AI adaptive push mechanism after class, an after-class effect analysis mechanism, etc.

[0123] Provide teacher training and support to help them master the use of AI tools and data analysis methods. Design a teaching model of human-computer collaboration so that teachers can give full play to their educational functions while using AI tools to improve teaching efficiency and quality.

[0124] This application can achieve diversified content innovation, resource quantification and intelligent recommendation, AI-driven e-textbooks, and full-life-cycle coverage of intelligent classrooms. This application also has an efficient 3D scene editor.

[0125] Diversified content innovation: This application can realize the construction of diversified e-textbooks, which not only integrate basic multimedia such as pictures, audio, and video, but also break through to introduce advanced elements such as virtual simulation scenes, 3D models, model hotspots, interactive scenes, VR experiences, and classroom recordings, greatly enriching the expression form and interactivity of textbook content, and significantly enhancing students' learning interest and depth of understanding.

[0126] Resource quantification and intelligent recommendation: By vectorizing educational resources, efficient organization and indexing of resources are realized, laying a solid foundation for data analysis and precise recommendation. Combining with artificial intelligence algorithms, the platform can intelligently push personalized learning resources according to students' interests and learning behaviors, improving learning efficiency and pertinence.

[0127] AI-driven e-textbooks: By deeply integrating artificial intelligence technology with vectorized resources, an intelligent and dynamically adaptable e-textbook system is constructed. This innovation not only enhances the intelligent interactivity of textbooks, but also realizes the intelligent update and adjustment of teaching content to meet the learning needs of different students.

[0128] Full-life-cycle coverage of intelligent classrooms: The platform covers all teaching links such as pre-class preview, in-class interaction, and after-class review, providing a one-stop intelligent classroom solution. Through data analysis and feedback mechanisms, the teaching process is continuously optimized to improve teaching quality and effects.

[0129] Efficient 3D scene editor: An advanced 3D scene editor is built-in, supporting the rapid import, adjustment, and rendering of multi-format models, greatly reducing the design and production threshold of complex teaching scenes, and providing teachers with powerful and flexible tools to create richer and more vivid classroom experiences.

[0130] The technical features of the above embodiments can be combined arbitrarily. For the sake of concise description, not all possible combinations of the technical features in the above embodiments are described. However, as long as the combinations of these technical features do not conflict, they should be considered as the scope recorded in this specification.

[0131] In this text, specific examples are used to illustrate the principle and implementation manner of this application. The description of the above embodiments is only used to help understand the method and its core idea of this application. At the same time, for those of ordinary skill in the art, according to the idea of this application, there will be changes in the specific implementation manner and application scope. To sum up, the content of this specification should not be construed as a limitation to this application.

Claims

1. An interactive rich media teaching material and courseware design method based on artificial intelligence, characterized in that: The interactive rich media teaching materials and courseware design method based on artificial intelligence includes: a resource center, a creative classroom module and a teaching and research ecological platform; Based on the resource center, provide the required resources to the Creative Classroom module and the teaching and research ecological platform; Based on the Creative Classroom module, using the resources provided by the resource center, virtual simulation scenes, 3D models, model hot spots, interactive scenes, VR experiences and classroom records are introduced, combined with teacher experience, teaching courseware is designed to form a personalized teaching mode of human-computer collaboration; Based on the teaching and research ecological platform, according to the personalized teaching model, data mining and artificial intelligence analysis technology are used to perform emotional calculation and social analysis on students' responses, analyze the degree of students' response to the teaching courseware, and generate a visual knowledge point map based on the resources provided by the resource center; the visual knowledge point map includes course maps, professional maps and job skill maps; the visual knowledge point map is used to cover the entire teaching process including pre-class preparation, in-class interaction and after-class review.

2. The interactive rich media teaching material and courseware design method based on artificial intelligence according to claim 1 is characterized in that: The resource center specifically includes: creative classroom resources and teaching and research ecological resources; The Chuangxiang classroom resources include recommended resources, my resources, team resources and question bank center; The recommended resources are system resources, which include classification feature codes, label feature codes, and introduction feature codes; according to the user's resource type tendency and behavior records, the user behavior is vectorized, and the vectorized user behavior is compared with the feature vector of the system resource to determine the system resource that the user tends to prefer; the user behavior includes the searched label, the searched resource feature code, and the search condition; My resources are personal resources. Users can classify their resources into folders and upload files. File types include pictures, audio, video, 2D interaction, 3D resources, VR resources and virtual simulation scenes. The team resource is a team shared resource; the team resource can be edited and deleted by the administrator and the resource uploader, and the remaining members of the team can save the team resource to their own folders; The teaching and research ecological resources can obtain required resources from the Chuangxiang classroom resources, and the teaching and research ecological resources also include a custom resource library and a custom question library.

3. The interactive rich media teaching material and courseware design method based on artificial intelligence according to claim 1 is characterized in that: The Creative Sharing Classroom module specifically includes: a digital intelligence courseware module, a digital intelligence e-book module, a digital intelligence teaching material module, and a 3D model editor; Based on the digital courseware module, SVG technology and WebRTC technology are used to import resources from the resource center while editing PPT online, and PPT animation effects are used to generate teaching courseware; Based on the digital intelligence e-book module and the digital intelligence courseware module, the DeepSpeech technology, MMAction2 motion recognition framework, Two-Stream ConvNets model and MediaRecorder API technology are used to introduce intelligent interaction function, video interaction function, canvas function and screen recording function; Based on the digital teaching material module, according to the AI ​​big model, combined with LTP technology and pre-training model, the teaching courseware is edited online; Based on the 3D model editor, the 3D model is edited using Threejs3D graphics rendering technology.

4. The interactive rich media teaching material and courseware design method based on artificial intelligence according to claim 1 is characterized in that: The teaching and research ecological platform specifically includes: a course map module, a professional map module and a job skill map module; Based on the course map module, a knowledge source map is generated, a teaching plan is created, and learning data is summarized; the course map module includes a knowledge source map unit, a teaching intervention unit, and a teaching data unit; Based on the professional map module, summarize professional information, generate professional course topology map and provide cross-professional integration exploration function; the professional map module includes a professional overview unit, a professional course topology unit and a cross-professional integration exploration unit; Based on the job skill map module, job courses are generated, a skill library is created, skill assessments are conducted based on students' majors and courses, learning outcomes are formulated based on the skill assessment results, and job development paths are planned; the job skill map module includes a job management unit, a skill library unit, a student skill assessment unit, a learning plan formulation unit, and a job development path planning unit.

5. The interactive rich media teaching material and courseware design method based on artificial intelligence according to claim 4 is characterized in that: The knowledge source graph unit uses G6 as a graph visualization engine to describe the coordinates, shape and mutual relationships of each knowledge point; uses swiper as a content touch tool to build an interactive channel between users and knowledge points; uses a force-directed layout algorithm to formulate knowledge point arrangement rules and determine the positions of knowledge point nodes; uses horizontal and vertical depth-first search algorithms to intelligently identify and aggregate knowledge points with intrinsic correlations and allocate learning weeks; for a single course with more than a set number of knowledge points, uses a GPU hardware acceleration solution to improve rendering efficiency.

6. The interactive rich media teaching material and courseware design method based on artificial intelligence according to claim 4 is characterized in that: The teaching intervention unit notifies the corresponding student end of the created teaching plan, exams and homework; the teaching plan is automatically rotated according to the learning week using a timed task framework to push different knowledge points to students every week; after students learn the knowledge points, Apache Spark is used to analyze the learning data based on the collected student learning data, find the students' knowledge weaknesses, and push test questions or knowledge points with relevance higher than the set threshold to the students for learning.

7. The interactive rich media teaching material and courseware design method based on artificial intelligence according to claim 4 is characterized in that: The teaching data unit counts the overall variance of the grades of students in each class; for any student's learning data, the number of questions answered during the exam and the student's exam score for each time are recorded, and the Pearson correlation coefficient is used to analyze the relationship between the number of questions answered and the score, so as to determine the student's learning efficiency; at the same time, the Spearman rank correlation coefficient is used to examine the relationship between the student's final score and his or her usual score, so as to fully understand the student's learning situation.

8. The interactive rich media teaching material and courseware design method based on artificial intelligence according to claim 1 is characterized in that: The professional overview unit includes the practice process, professional construction and construction results; In terms of the practical process, TimelineJS technology, Tailwind CSS layout method and GSAP technology are combined to use a timeline or narrative display method to show the origin and key events and important turning points in the development of the profession; In terms of the professional construction, D3.js is used to display the chart of technical faculty strength; Lodash technology is used to process the data of faculty chart; In terms of the construction achievements, the Swiper component is used to dynamically display the construction achievements in a carousel, and the one-click sharing function is used to share the construction achievements.

9. The interactive rich media teaching material and courseware design method based on artificial intelligence according to claim 1, characterized in that: The professional course topology map unit uses the G6 graph visualization engine in the form of canvas to generate a professional course topology map of all courses of any major in the form of nodes and display it on the page; for majors with more courses than the set threshold, the Intersection ObserverAPI lazy loading technology is used to delay the loading of content outside the viewport; at the same time, the GPU hardware acceleration solution is enabled to ensure that complex pages can also be loaded smoothly; when searching is required, the filter method of Lodash is used to filter course data.

10. The interactive rich media teaching material and courseware design method based on artificial intelligence according to claim 1, characterized in that: The skill library unit adopts Redis cache technology and recursive algorithm, combined with a multi-level parent-child structure, to describe the progressive relationship of skill classification.

Images