Teaching resource management optimization system based on artificial intelligence large model

Abstract

The invention particularly relates to a teaching resource management optimization system based on an artificial intelligence large model. The teaching resource management optimization system based on the artificial intelligence large model comprises an abstract generation module, a cover generation module, a resource directory automatic construction module, a video dotting module, a sensitive word monitoring module and a central control unit. Interaction and cooperative work among the modules are coordinated through the central control unit, and processing results of the modules are integrated and updated to a resource library. The teaching resource management optimization system based on the artificial intelligence large model is simple and efficient, the teaching resource management efficiency is greatly improved, teaching resources can be put into use more quickly, retrieval and use are facilitated, the actual utilization value of the resources is improved, and a healthy content environment is created for online education by detecting sensitive words.

Description

Technical Field

[0001] The present invention relates to the technical field of educational resource management, and in particular to a teaching resource management optimization system based on an artificial intelligence large model. Background Art

[0002] In today's era of booming digital education, the amount of teaching resources is increasing day by day, covering a large number of videos, documents and other forms. However, traditional teaching resource management methods have many drawbacks:

[0003] In the past, educators could only watch the entire video and manually summarize the key points based on their personal understanding. This process was not only time-consuming, but also prone to missing important information due to limited human attention, making it difficult to ensure the accuracy and comprehensiveness of the summary.

[0004] When generating resource covers, there is a lack of effective intelligent means, and it is often difficult to accurately locate the most representative key scenes, which makes the resource covers unable to attract users well and unable to intuitively reflect the core content of the resources.

[0005] Building a resource directory has always been a tedious task. The traditional model requires users to manually select directory information step by step. The operation process is lengthy and errors are easy to occur if you are not careful, which seriously affects the efficiency of resource uploading and slows down the update and deployment of the entire teaching resource.

[0006] Video marking is an even more arduous task. Watching each frame and manually marking important events and key frames requires a huge effort from staff, is inefficient, and often results in missing key nodes, which brings great inconvenience to subsequent users in retrieving and using video resources.

[0007] In terms of sensitive word monitoring, traditional review methods rely on manual word-by-word checking of voice and text content. Faced with massive teaching resources, the review workload grows exponentially, and the efficiency is extremely low, making it difficult to meet the requirements of rapidly developing online education for content security and compliance.

[0008] Based on the above problems, the present invention proposes a teaching resource management optimization system based on an artificial intelligence large model. Summary of the Invention

[0009] In order to remedy the deficiencies of the prior art, the present invention provides a simple and efficient teaching resource management optimization system based on an artificial intelligence large model.

[0010] The present invention is achieved through the following technical solutions:

[0011] A teaching resource management optimization system based on an artificial intelligence large model, comprising:

[0012] The summary generation module is used to connect to the video resource library of the teaching resource platform, extract the voice and image information of the video, and use the artificial intelligence model to generate a summary text covering the core points of the video;

[0013] The cover generation module connects to video resources, locates key scenes based on the big model's understanding of the video content, and custom-selects a representative frame as the resource cover;

[0014] The resource directory automatic construction module works with the large model when resources are uploaded to automatically identify the information of video resources, accurately classify them, and generate the corresponding resource directory;

[0015] The video marking module relies on the intelligent analysis of video content by large models to automatically identify important events and key frames and generate markers and node information;

[0016] The sensitive word monitoring module uses the natural language processing capabilities of large models to monitor the voice and text content of teaching resources for sensitive words to ensure content security and compliance;

[0017] The central control unit is used to coordinate the interaction and collaboration between the above modules, integrate the processing results of each module and update them to the resource library.

[0018] The summary generation module is connected to the video resource library of the teaching resource platform. When a new video is uploaded or a user requests to generate a summary, it uses the powerful speech recognition and image analysis capabilities of the large model to automatically extract the voice information in the video and convert the extracted voice information into text. At the same time, it analyzes the image content, integrates the text and image content information, and accurately generates a summary text covering the core points of the video, so as to avoid the time-consuming, labor-intensive and error-prone problems of manual summarization.

[0019] The cover generation module connects to video resources and uses scene recognition technology based on the big model's understanding of video content to automatically and quickly locate several key scenes that can represent the video theme or key knowledge points. It then customizes and selects the most visually attractive and information-representative frame as the resource cover, so that the cover can accurately convey the resource content and improve the resource's recognition.

[0020] The resource directory automatic construction module automatically identifies the subject, grade and knowledge point of the video resources based on the massive knowledge learned by the large model and the classification model of educational resources, and generates a resource directory according to the preset directory structure rules.

[0021] The video marking module relies on the intelligent analysis capabilities of the large model to automatically parse the video content. By identifying multi-dimensional features, including event logic, picture changes, and knowledge point conversion dimensions, it accurately locates important events and key frames, and automatically generates corresponding tags and node information, so that subsequent users can quickly retrieve the required video clips based on the tags, greatly improving the user experience.

[0022] The sensitive word monitoring module uses a pre-trained sensitive word recognition model to comprehensively cover the voice and text content of teaching resources, screen teaching resources for sensitive words in real time or according to preset cycles, and mark resources containing sensitive words in real time and feed them back to management personnel.

[0023] In the daily management process, the sensitive word monitoring module scans the resource library at custom-set time intervals. When sensitive words are found, an alarm is sent to the central control unit. The central control unit sends a control signal to suspend access to relevant resources and notify management personnel to conduct manual review and modification to ensure content security and compliance.

[0024] During the resource upload process, the central control unit triggers the resource directory automatic construction module, summary generation module, cover generation module and video marking module to work in parallel. After each module is completed, the results are fed back to the central control unit for integration and update.

[0025] A teaching resource management optimization method based on an artificial intelligence large model, based on the above teaching resource management optimization system, includes the following steps:

[0026] Step S1: System initialization

[0027] During the initial deployment of the system, the integration and adaptation of the AI model were completed. The model was then fine-tuned and trained based on the specific characteristics of the education field to familiarize it with the terminology, knowledge system, and custom common formats related to teaching resources. Furthermore, the operating parameters of the summary generation module, cover generation module, resource directory automatic construction module, video dotting module, and sensitive word detection module were configured, including the word limit for summary generation, visual priority rules for cover selection, the default hierarchical structure of the resource directory, and the selection rules for important events and keyframes, as well as sensitive words.

[0028] Step S2: Automatically build resource directory

[0029] Educators select local teaching resources and upload them to the system. After the system receives the files, the central control unit activates the automatic resource directory construction module, analyzes the video file name, metadata, and preliminary parsed video content, determines its subject, applicable grade, and course theme, and generates the corresponding directory structure according to the preset directory template;

[0030] Step S3: Parallel implementation of summary generation, cover generation and video punctuation

[0031] The central control unit notifies the summary generation module, cover generation module, and video dotting module to work in parallel to perform summary, cover creation, and video dotting processing on the resource respectively. The results of the summary generation module, cover generation module, and video dotting module are fed back to the central control unit, which integrates and updates them to the corresponding location in the resource library.

[0032] Step S4: Sensitive word monitoring

[0033] The sensitive word monitoring module scans the resource library at custom intervals. The large model screens the speech-to-text and original text documents in the resource for sensitive words. Once a sensitive word is found, an alert is immediately sent to the central control unit, along with the resource location and sensitive word details.

[0034] The central control unit sends a control signal to suspend access to relevant resources to prevent the spread of negative information, and notifies management personnel to conduct manual review and modification to ensure content security and compliance.

[0035] Based on the analysis results provided by the big model, managers can quickly locate the problem and decide to modify, delete or otherwise handle the resources to ensure that the resources are always compliant.

[0036] The beneficial effects of the present invention are: the teaching resource management optimization system based on the artificial intelligence large model is simple and efficient. It not only greatly improves the efficiency of teaching resource management, allowing teaching resources to be put into use more quickly, but also facilitates retrieval and use, thereby increasing the actual utilization value of resources. By detecting sensitive words, it creates a healthy content environment for online education. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0038] Attachment Figure 1 Schematic diagram of the teaching resource management optimization method based on the artificial intelligence large model of the present invention. DETAILED DESCRIPTION

[0039] In order to enable those skilled in the art to better understand the technical solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work should fall within the scope of protection of the present invention.

[0040] The teaching resource management and optimization system based on the artificial intelligence large model includes:

[0041] The summary generation module is used to connect to the video resource library of the teaching resource platform, extract the voice and image information of the video, and use the artificial intelligence model to generate a summary text covering the core points of the video;

[0042] The cover generation module connects to video resources, locates key scenes based on the big model's understanding of the video content, and custom-selects a representative frame as the resource cover;

[0043] The resource directory automatic construction module works with the large model when resources are uploaded to automatically identify the information of video resources, accurately classify them, and generate the corresponding resource directory;

[0044] The video marking module relies on the intelligent analysis of video content by large models to automatically identify important events and key frames and generate markers and node information;

[0045] The sensitive word monitoring module uses the natural language processing capabilities of large models to monitor the voice and text content of teaching resources for sensitive words to ensure content security and compliance;

[0046] The central control unit is used to coordinate the interaction and collaboration between the above modules, integrate the processing results of each module and update them to the resource library.

[0047] The summary generation module connects to the video resource library of the teaching resource platform. When a new video is uploaded or a user requests a summary, it leverages the powerful speech recognition and image analysis capabilities of the large model to automatically extract the audio information from the video and convert it into text. It also analyzes the image content and combines the text and image content to accurately generate a summary covering the core points of the video. For example, for a physics lab video, the summary might include the experiment's objectives, key steps, and conclusions. This avoids the time-consuming, labor-intensive, and error-prone manual summarization.

[0048] The cover generation module connects to video resources and uses scene recognition technology based on the big model's understanding of video content to automatically and quickly locate several key scenes that can represent the video theme or key knowledge points. It then custom-selects the most visually appealing and information-representative frame as the resource cover, so that the cover can accurately convey the resource content, improve the resource's recognition, and enhance the resource's visual appeal.

[0049] The resource directory automatic construction module automatically identifies the subject, grade and knowledge point information of video resources based on the massive knowledge learned by the large model and the classification model of resources in the educational field, and generates a resource directory according to the preset directory structure rules, which greatly saves the user's manual operation time and reduces the probability of errors.

[0050] The video marking module relies on the intelligent analysis capabilities of the large model to automatically parse the video content. By identifying multi-dimensional features, including event logic, picture changes, and knowledge point conversion dimensions, it accurately locates important events and key frames, and automatically generates corresponding tags and node information, so that subsequent users can quickly retrieve the required video clips based on the tags, greatly improving the user experience.

[0051] The sensitive word monitoring module uses a pre-trained sensitive word recognition model to comprehensively cover the voice and text content of teaching resources. Whether it is newly uploaded resources or regular reviews of existing resources on the platform, it screens teaching resources for sensitive words in real time or according to preset cycles, and marks resources containing sensitive words in real time and feeds back to management personnel, effectively reducing the workload of manual review and ensuring the security and compliance of teaching content.

[0052] In the daily management process, the sensitive word monitoring module scans the resource library at custom-set time intervals. When sensitive words are found, an alarm is sent to the central control unit. The central control unit sends a control signal to suspend access to relevant resources and notify management personnel to conduct manual review and modification to ensure content security and compliance.

[0053] During the resource upload process, the central control unit triggers the resource directory automatic construction module, summary generation module, cover generation module and video marking module to work in parallel. After each module is completed, the results are fed back to the central control unit for integration and update.

[0054] The teaching resource management optimization method based on the artificial intelligence large model is based on the above-mentioned teaching resource management optimization system and includes the following steps:

[0055] Step S1: System initialization

[0056] During the initial deployment of the system, the integration and adaptation of the AI model were completed. The model was then fine-tuned and trained based on the specific characteristics of the education field to familiarize it with the terminology, knowledge system, and custom common formats related to teaching resources. Furthermore, the operating parameters of the summary generation module, cover generation module, resource directory automatic construction module, video dotting module, and sensitive word detection module were configured, including the word limit for summary generation, visual priority rules for cover selection, the default hierarchical structure of the resource directory, and the selection rules for important events and keyframes, as well as sensitive words.

[0057] Step S2: Automatically build resource directory

[0058] Educators select local teaching resources (including text, pictures, and videos) and upload them to the system. After the system receives the file, the central control unit starts the resource directory automatic construction module, analyzes the video file name, metadata, and preliminary parsed video content, determines its subject, applicable grade, and course theme, and generates the corresponding directory structure according to the preset directory template, such as the "subject / grade / course / chapter" format.

[0059] Step S3: Parallel implementation of summary generation, cover generation and video punctuation

[0060] The central control unit notifies the summary generation module, cover generation module, and video dotting module to work in parallel to perform summary, cover creation, and video dotting processing on the resource respectively. The results of the summary generation module, cover generation module, and video dotting module are fed back to the central control unit, which integrates and updates them to the corresponding location in the resource library.

[0061] Step S4: Sensitive word monitoring

[0062] The sensitive word monitoring module scans the resource library at a customizable time interval (e.g., once an hour). The large model screens the speech-to-text and original text documents in the resource for sensitive words. Once a sensitive word is found, an alert is immediately sent to the central control unit, along with the resource location and sensitive word details.

[0063] The central control unit sends a control signal to suspend access to relevant resources to prevent the spread of negative information, and notifies management personnel to conduct manual review and modification to ensure content security and compliance.

[0064] Based on the analysis results provided by the big model, managers can quickly locate the problem and decide to modify, delete or otherwise handle the resources to ensure that the resources are always compliant.

[0065] Compared with existing technologies, this teaching resource management and optimization system based on artificial intelligence large models has the following characteristics:

[0066] First, it greatly improves the efficiency of teaching resource management: through automated summary generation, directory construction, video marking and other functions, compared with traditional manual methods, it saves a lot of manpower and time costs, allowing resources to be put into use more quickly to meet teaching needs.

[0067] Second, improve resource quality and usability: accurate summaries, attractive covers, and convenient video annotations allow educators and students to understand resource content more quickly and accurately, facilitate retrieval and use, and increase the actual utilization value of resources.

[0068] Third, ensure content compliance: The sensitive word monitoring module uses the efficient recognition capabilities of large models to ensure that teaching resources comply with relevant regulations and educational standards, thereby creating a healthy content environment for online education.

[0069] The embodiment described above is only one specific implementation of the present invention. Common changes and substitutions made by those skilled in the art within the scope of the technical solution of the present invention should be included in the protection scope of the present invention.

Claims

1. A teaching resource management optimization system based on an artificial intelligence large model, characterized by: include: The summary generation module is used to connect to the video resource library of the teaching resource platform, extract the voice and image information of the video, and use the artificial intelligence model to generate a summary text covering the core points of the video; The cover generation module connects to video resources, locates key scenes based on the big model's understanding of the video content, and custom-selects a representative frame as the resource cover; The resource directory automatic construction module works with the large model when resources are uploaded to automatically identify the information of video resources, classify them, and generate the corresponding resource directory; The video marking module relies on the intelligent analysis of video content by large models to automatically identify important events and key frames and generate markers and node information; The sensitive word monitoring module uses the natural language processing capabilities of large models to monitor the voice and text content of teaching resources for sensitive words to ensure content security and compliance; The central control unit is used to coordinate the interaction and collaboration between the above modules, integrate the processing results of each module and update them to the resource library.

2. The teaching resource management and optimization system based on the artificial intelligence large model according to claim 1 is characterized by: The summary generation module is connected to the video resource library of the teaching resource platform. When a new video is uploaded or a user requests to generate a summary, it uses the powerful speech recognition and image analysis capabilities of the large model to automatically extract the voice information in the video and convert the extracted voice information into text. At the same time, it analyzes the image content, integrates the text and image content information, and generates a summary text covering the core points of the video, so as to avoid the time-consuming, labor-intensive and error-prone problems of manual summarization.

3. The teaching resource management and optimization system based on artificial intelligence large model according to claim 1 is characterized by: The cover generation module connects to video resources and, based on the big model's understanding of the video content, uses scene recognition technology to automatically locate several key scenes that can represent the video theme or key knowledge points, and custom-selects the most visually attractive and information-representative frame as the resource cover to improve the resource's recognition.

4. The teaching resource management and optimization system based on artificial intelligence large model according to claim 1 is characterized in that: The resource directory automatic construction module automatically identifies the subject, grade and knowledge point of the video resources based on the massive knowledge learned by the large model and the classification model of educational resources, and generates a resource directory according to the preset directory structure rules.

5. The teaching resource management and optimization system based on artificial intelligence large model according to claim 1 is characterized in that: The video marking module relies on the intelligent analysis capabilities of the large model to automatically parse the video content. By identifying multi-dimensional features, including event logic, picture changes, and knowledge point conversion dimensions, it locates important events and key frames and automatically generates corresponding tags and node information so that subsequent users can quickly retrieve the required video clips based on the tags.

6. The teaching resource management and optimization system based on artificial intelligence large model according to claim 1 is characterized by: The sensitive word monitoring module uses a pre-trained sensitive word recognition model to comprehensively cover the voice and text content of teaching resources, screen teaching resources for sensitive words in real time or according to preset cycles, and mark resources containing sensitive words in real time and feed them back to management personnel.

7. The teaching resource management and optimization system based on artificial intelligence large model according to claim 6 is characterized by: In the daily management process, the sensitive word monitoring module scans the resource library at custom-set time intervals. When sensitive words are found, an alarm is sent to the central control unit. The central control unit sends a control signal to suspend access to relevant resources and notify management personnel to conduct manual review and modification to ensure content security and compliance.

8. A teaching resource management optimization method based on an artificial intelligence large model, characterized by: The teaching resource management optimization system according to any one of claims 1 to 7 comprises the following steps: Step S1: System initialization During the initial deployment of the system, the integration and adaptation of the AI model were completed. The model was then fine-tuned and trained based on the specific characteristics of the education field to familiarize it with the terminology, knowledge system, and custom common formats related to teaching resources. Furthermore, the operating parameters of the summary generation module, cover generation module, resource directory automatic construction module, video dotting module, and sensitive word detection module were configured, including the word limit for summary generation, visual priority rules for cover selection, the default hierarchical structure of the resource directory, and the selection rules for important events and keyframes, as well as sensitive words. Step S2: Automatically build resource directory Educators select local teaching resources and upload them to the system. After the system receives the files, the central control unit activates the automatic resource directory construction module, analyzes the video file name, metadata, and preliminary parsed video content, determines its subject, applicable grade, and course theme, and generates the corresponding directory structure according to the preset directory template; Step S3: Parallel implementation of summary generation, cover generation and video punctuation The central control unit notifies the summary generation module, cover generation module, and video dotting module to work in parallel to perform summary, cover creation, and video dotting processing on the resource respectively. The results of the summary generation module, cover generation module, and video dotting module are fed back to the central control unit, which integrates and updates them to the corresponding location in the resource library. Step S4: Sensitive word monitoring The sensitive word monitoring module scans the resource library at custom intervals. The large model screens the speech-to-text and original text documents in the resource for sensitive words. Once a sensitive word is found, an alert is immediately sent to the central control unit, along with the resource location and sensitive word details. The central control unit sends a control signal to suspend access to relevant resources and notifies management personnel to conduct manual review and modification to ensure content security and compliance.

Images