Teaching resource retrieval system

Abstract

The invention discloses a teaching resource retrieval system, and belongs to the field of teaching resource management. The system comprises a data acquisition module, a knowledge storage library, a teacher teaching platform and a teaching processing module. The data acquisition module acquires teaching videos, voices and courseware; the knowledge storage library comprises a temporary database, a teaching resource library and a standard knowledge library; and the teaching processing module analyzes the difference between the classroom content and the standard knowledge points and the description difference of different teachers through a teaching content comparison unit, generates knowledge points to be confirmed through a differentiation processing unit, feeds back the knowledge points through the teachers, and updates the standard knowledge base through a standardization updating unit. The system further comprises an authority management module, a synonym processing module, an accent comparison and correction module, a class management module and the like, knowledge point standardization dynamic optimization, high-quality resource accurate recognition and teaching effect evaluation are achieved, the teaching knowledge point standardization degree and updating efficiency are improved, and high-quality teaching resource utilization and teaching quality improvement are facilitated.

Description

Technical Field

[0001] This invention relates to the field of educational information technology, and in particular to a teaching resource retrieval system. Background Technology

[0002] In the current field of teaching resource management and retrieval, the standardization and dynamic optimization of teaching knowledge points face three major technical bottlenecks: First, traditional systems lack intelligent difference recognition mechanisms. Due to differences in individual teaching styles, regional cultural backgrounds, or perspectives of understanding, different teachers often exhibit semantic deviations or terminological inconsistencies in their descriptions of the same knowledge point. Existing technologies can only perform simple keyword matching and cannot achieve automated identification and comparison of expression differences through semantic analysis. This results in a large accumulation of redundant or contradictory teaching content in the knowledge base. Second, the updating and maintenance of the standard knowledge base relies entirely on manual operation. Newly emerging high-quality teaching resources or knowledge point additions require a cumbersome offline review process before being incorporated into the standard system. This lag makes it difficult for the standard knowledge base to adapt to innovative content generated in real time during teaching practice. Especially in cross-regional teaching scenarios, pronunciation deviations caused by dialect accents further exacerbate the confusion in the expression of knowledge points. Third, existing systems lack objective criteria for evaluating the value of teaching resources. They cannot automatically identify high-quality teaching cases based on class performance data, nor can they dynamically optimize stored resources based on the frequency of knowledge point usage. This results in high-frequency, high-quality resources being mixed with low-frequency, outdated content, severely impacting retrieval efficiency. Furthermore, when classroom teaching contains content that deviates from standard knowledge points, traditional systems are unable to provide timely feedback to teachers or establish a closed-loop processing mechanism for multi-teacher collaborative verification. This passive response model severely hinders the speed of teaching standardization. These systemic deficiencies make it difficult for existing teaching resource systems to achieve the three core functions of maintaining consistent knowledge representation, timely incorporating teaching innovations, and accurately recommending high-quality resources. Summary of the Invention

[0003] The purpose of this application is to provide a teaching resource retrieval system that has the advantages of automatically identifying differences in teaching descriptions through semantic analysis and dynamically updating the standard knowledge base.

[0004] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a teaching resource retrieval system, comprising: a data acquisition module, equipped with a camera in the classroom, a voice receiving device, and an interface connected to the classroom teaching computer, for acquiring teacher lecture video streams, teacher lecture voice streams, and current teaching courseware content; The knowledge repository includes a temporary database for storing temporary information, a teaching resource repository, and a standard knowledge repository, wherein the standard knowledge repository stores several standard teaching knowledge point data. The teacher teaching platform is connected to the knowledge repository and is equipped with a teacher client for teachers to upload personal teaching materials to the teaching resource repository and for downloading data from the standard knowledge repository. The teaching processing module connects to the data acquisition module and the knowledge repository. The teaching processing module includes a teaching content comparison unit, which is used to perform semantic analysis on the classroom teaching content collected by the acquisition module and compare it with the standard teaching knowledge point data in the standard knowledge base, while also comparing the descriptions of the same knowledge point by different teachers. The differentiation processing unit generates a knowledge point to be confirmed when the classroom teaching content is inconsistent with the standard teaching knowledge point data or when different teachers describe the same knowledge point in different ways. The knowledge point to be confirmed is sent to the corresponding or different teacher clients and the unit receives feedback information from the corresponding or different teachers on the knowledge point to be confirmed. The standardization update unit determines the standardized expression of knowledge points based on feedback information and updates the content of standard teaching knowledge point data in the standard knowledge base.

[0005] By adopting the above technical solution, and through a collaborative mechanism of data collection, teaching content comparison, differentiated processing, and standardized updates, an automated closed loop of classroom teaching content from collection to dynamic updates of standardized knowledge points has been achieved. This effectively solves the problems of inconsistent descriptions of the same knowledge point by different teachers and the lag in updating the standard knowledge base, and improves the standardization and updating efficiency of teaching knowledge points.

[0006] The present invention is further configured such that: the teacher teaching platform also includes a permission management unit, which is used to distinguish between ordinary permissions and senior permissions of the teacher client. The teacher client with senior permissions has access to modify and add standard teaching knowledge point data.

[0007] By adopting the above technical solution, and distinguishing between ordinary and senior permissions through the permission management unit, senior teachers are only granted the permission to modify and add standard knowledge points. This ensures the authority and accuracy of the modification of standard knowledge points, avoids the chaos in the knowledge base caused by arbitrary modifications, and improves the reliability of standardization processing.

[0008] The invention is further configured such that the teaching processing module is also equipped with an annotation unit, which is mainly used to annotate the data in the standard knowledge base where there are differences in the historical teaching content. When the number of annotations for a certain standard teaching knowledge point reaches a preset value, the standardized teaching knowledge point is sent to the teacher client with senior permissions for modification or content expansion.

[0009] By adopting the above technical solution, the knowledge points that repeatedly show differences in history teaching are cumulatively annotated through annotation units. When the preset value is reached, it triggers senior teachers to review and optimize. This solution addresses the standardization problem of high-frequency difference knowledge points and improves the adaptability and accuracy of the standard knowledge base to teaching practice.

[0010] The present invention is further configured such that: when classroom teaching content is not included in the standard teaching knowledge point data, the differentiation processing unit generates the unincluded content and sends it to a teacher client with senior privileges. After the teacher on the teacher client with senior privileges approves the unincluded content, it is stored in the standard teaching knowledge point data.

[0011] By adopting the above technical solution, the non-included content is generated through the differentiated processing unit and then included in the standard library after being approved by senior teachers. This achieves the standardized and authoritative inclusion of new teaching content, avoids the omission of valuable new teaching content, enriches the coverage of the standard knowledge base, and ensures its timeliness.

[0012] The invention is further configured such that: the teaching processing module is also equipped with a synonym processing unit, which, when comparing teaching content, prioritizes synonym processing when inconsistencies are found; when experienced teacher clients input standard teaching knowledge point data, the synonym processing unit compares the data, and when the same knowledge points are found, the content input by the experienced teacher clients is summarized and processed.

[0013] By adopting the above technical solution, the synonym processing unit prioritizes processing synonyms and summarizing the same knowledge points, reducing misjudgments in content comparison caused by differences in expression, avoiding duplicate entry of the same knowledge points, and improving the accuracy of content comparison and the simplicity of the standard knowledge base.

[0014] The invention is further configured such that: the data acquisition module acquires the classroom teaching content taught by the teacher in the classroom in real time, and displays the relevant knowledge points on the classroom teaching computer to further expand the teaching content; when the teaching content comparison unit finds that the classroom teaching content acquired by the acquisition module is inconsistent with the standard teaching knowledge point data, it sends the inconsistent information to the teacher's teaching computer.

[0015] By adopting the above technical solution, the extended content can be displayed in real time through the data acquisition module, and information on inconsistencies between the teaching content and the standard knowledge points can be fed back to the teacher's teaching computer in real time. This helps the teacher to be aware of teaching deviations in a timely manner and make adjustments, while also expanding the depth of teaching content and improving the pertinence and richness of classroom teaching.

[0016] The present invention is further configured such that: the system also includes an accent comparison and correction module, which is connected to the data acquisition module and the standard knowledge base. The accent comparison and correction module is used to acquire the teacher's lecture voice stream collected by the data acquisition module, perform semantic and speech feature analysis on the pronunciation in the voice stream, compare it with the standard speech data stored in the standard knowledge base, identify the expression differences or pronunciation deviations caused by accents, generate accent correction prompt information, and send it to the teacher's teaching computer and teacher client to assist the teacher in adjusting pronunciation; The standard speech data includes standard pronunciations and terminological readings.

[0017] By adopting the above technical solution, the accent comparison and correction module analyzes the teacher's lecture speech stream and compares it with standard speech data. It identifies pronunciation deviations caused by accents and generates correction prompts, which solves the problem of non-standard pronunciation of terms caused by accents, improves the accuracy of teaching speech, and helps teachers improve their pronunciation to reduce students' comprehension deviations.

[0018] The present invention is further configured such that: the system also includes a class management module, which configures an independent class database for each class and records the class's grades each time. When the grade of a certain class in a certain lesson is higher than that of other classes, the system obtains the video, audio and courseware of that lesson for that class and improves its ranking in the knowledge base. The class management module generates an evaluation report on the teaching effectiveness of the teacher in charge of the class based on chapter test scores and average mastery of knowledge points in class.

[0019] By adopting the above technical solution, the class management module records grades, improves the ranking of high-scoring class resources, and generates teaching effectiveness evaluation reports. Based on objective grade data, it accurately identifies high-quality teaching resources and improves resource retrieval efficiency. At the same time, by evaluating teaching effectiveness through multi-dimensional data, it provides data support for teachers to improve teaching and promotes the improvement of teaching quality.

[0020] The present invention is further configured such that: the knowledge repository is connected to a storage management unit for recording the call frequency of each knowledge point and classifying and managing the data in the knowledge repository according to the call frequency; the knowledge repository includes: The high-frequency storage repository stores data that is accessed at a preset frequency, as well as classroom teaching content from high-scoring classes in that subject. Low-frequency repository, storing data whose access frequency does not reach a preset value; Archive repository, storing data that has not been accessed within a preset time period.

[0021] By adopting the above technical solution, storage resources are classified and stored according to the frequency of access through the storage management unit. High-frequency content and high-scoring class resources are included in the high-frequency storage repository, which optimizes the knowledge storage structure and improves the retrieval speed of high-frequency and high-quality resources. At the same time, low-frequency and archived content is managed in a reasonable manner, saving storage resources and improving system operating efficiency.

[0022] The present invention is further configured such that: the knowledge storage repository is configured with an error database; when the teaching processing module confirms the standardized expression of the knowledge point based on the feedback information, it stores the identified differences and the knowledge point expressions that are finally determined to be incorrect into the error database. The teaching processing module also includes a test question generation unit, which generates test questions based on standard teaching knowledge point data in the knowledge base.

[0023] By adopting the above technical solution, the error database stores discrepancies and incorrect knowledge points, and the test question generation unit generates questions based on standard knowledge points. This focuses on common mistakes in teaching to assist targeted teaching, while ensuring the consistency between the test questions and standard knowledge points, thereby improving the accuracy of the test questions and the effectiveness of teaching assessment.

[0024] The present invention has significant technical effects due to the adoption of the above technical solutions: The teaching resource retrieval system and its differential processing and standardization update method provided in this application include a data acquisition module, a knowledge storage repository, a teacher teaching platform and a teaching processing module. By comparing classroom teaching content with standard knowledge point data through semantic analysis, it generates knowledge points to be confirmed and updates the standard knowledge base based on feedback. It has the advantage of improving the uniformity and accuracy of teaching knowledge point expression through an automated difference identification and collaborative confirmation mechanism. Attached Figure Description

[0025] Figure 1 This is a schematic diagram of the structure of a teaching resource retrieval system. Detailed Implementation

[0026] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments.

[0027] Example: In existing technologies, the standardization and dynamic optimization of teaching knowledge points mainly rely on manual organization and maintenance. Different teachers, due to differences in teaching styles and perspectives, often describe the same teaching knowledge point differently. Current systems lack the ability to automatically identify, compare, and standardize these discrepancies, leading to chaotic descriptions of knowledge points in the knowledge base. New teaching content or supplementary expansions to knowledge points require manual screening and input into the standard system, a cumbersome and outdated process that struggles to adapt to real-time changes in teaching content. Furthermore, when discrepancies arise, existing teaching resources cannot be used for immediate correction, resulting in an overall deficiencies in the teaching system. To address the aforementioned issues, an efficient confirmation and processing mechanism is needed. First, it's crucial to automate the identification of discrepancies in teaching knowledge points. This involves analyzing teacher lecture videos, audio streams, and courseware content, combined with semantic analysis techniques to extract core knowledge points. Second, a dynamic feedback mechanism is necessary to send identified discrepancies to teachers for confirmation, avoiding the inefficiency of manual screening. Finally, a standardized update process must be built to automatically correct the knowledge base content based on feedback from multiple sources, forming a closed-loop optimization system. The key to this process lies in integrating multi-source data collection, semantic comparison algorithms, and collaborative feedback mechanisms to create an adaptive, standardized knowledge management architecture.

[0028] This application proposes a system comprising a data acquisition module, a knowledge repository, a teacher teaching platform, and a teaching processing module. The data acquisition module is equipped with a camera, a voice receiver, and a teaching computer interface in the classroom to collect video streams, audio streams, and courseware content from the teacher's lectures. The knowledge repository includes a temporary database, a teaching resource database, and a standard knowledge base storing standard teaching knowledge points. The teacher teaching platform connects to the knowledge repository, providing functions for uploading courseware and downloading standard knowledge points. The teaching processing module includes a teaching content comparison unit, a differentiation processing unit, and a standardization update unit. Through semantic analysis, it compares classroom teaching content with standard knowledge points, generates knowledge points to be confirmed, processes teacher feedback, and finally updates the standard knowledge base.

[0029] The data acquisition module refers to the device that acquires teaching site data through multimodal equipment. Specifically, it can be implemented using a network camera array, directional microphone group, and teaching computer API interface. It is used to synchronously capture visual, auditory, and courseware information during the teacher's teaching process. The knowledge repository refers to the hierarchical storage structure for managing teaching data. Specifically, it can be implemented using a hybrid architecture of relational database and document database. Real-time data acquisition is achieved through temporary database caching. The teaching resource repository stores courseware uploaded by teachers, and the standard knowledge repository maintains standardized knowledge points. The teaching content comparison unit refers to the processor that performs semantic analysis. Specifically, it can use a natural language processing model to vectorize text content and calculate the matching degree with standard knowledge points using a cosine similarity algorithm. The difference processing unit refers to the logic controller that generates knowledge points to be confirmed. Specifically, it can use a message queue mechanism to distribute the difference content to the corresponding teacher client and record the feedback status. The standardization update unit refers to the decision module that updates the knowledge base. Specifically, it can use a majority voting algorithm or weight evaluation model to determine the final standardized expression.

[0030] During system operation, the data acquisition module captures in real time the blackboard writing in the teacher's lecture video stream, the explanation content in the audio stream, and the text content in the teaching courseware. The teaching content comparison unit performs semantic analysis on these three types of data, extracts the core knowledge point descriptions, and compares them with the corresponding entries in the standard knowledge base. When a semantic deviation is detected between the classroom teaching content and the standard expression, the differentiation processing unit generates a comparison report containing the original content and the standard content, which is sent to the teacher or other relevant teacher clients through the teacher's teaching platform. Teachers can view the difference details on the client interface, choose to confirm the validity of the standard expression or submit correction suggestions. After collecting all valid feedback, the standardization update unit generates a new version of the standard expression through a semantic fusion algorithm and automatically updates it to the standard knowledge base. For controversial expression differences, the system will trigger multiple rounds of feedback processes until a consensus is reached.

[0031] This solution solves the problem of low efficiency in manual comparison by using automated data collection and semantic analysis technology. Compared with the traditional static knowledge base management method, the dynamic feedback mechanism established by this system can capture the differences in expression in teaching practice in real time, and ensure the accuracy of standard knowledge points through collaborative confirmation process. In addition, the closed-loop update system breaks through the time delay defect of the traditional manual input mode, enabling the standard knowledge base to be continuously optimized with teaching practice.

[0032] This application realizes the automated identification and collaborative confirmation of differences in teaching knowledge points, effectively eliminating the problem of confused expression caused by individual differences among teachers. The system significantly improves the response speed of knowledge point standardization processing by collecting multi-source teaching data in real time and performing intelligent comparison. The established dynamic update mechanism enables the standard knowledge base to absorb new content generated in teaching practice in a timely manner, ensuring the timeliness and applicability of the knowledge system. The closed-loop design of the feedback processing flow provides a reliable verification method for knowledge point standardization, avoiding the deviation that may be caused by single manual judgment.

[0033] The teacher teaching platform also includes a permission management unit to differentiate between ordinary and advanced permissions for teacher clients. Teacher clients with advanced permissions have access to modify and add standard teaching knowledge point data. The permission management unit is a module that controls the operation permissions of teacher clients in a hierarchical manner. Specifically, it can be implemented using a role-permission mapping table combined with an authentication mechanism. For example, it can differentiate between ordinary and advanced teacher accounts by an account type field and restrict access to operation interfaces in the database. Ordinary permissions refer to basic operation permissions that only allow teacher clients to perform tasks such as uploading courseware and downloading standard knowledge base data. This can be implemented through interface whitelists or visibility control of function menus, such as hiding data modification buttons or restricting interface calls for submitting modification requests. Advanced permissions refer to advanced operation permissions that allow modification of content or addition of entries to standard teaching knowledge point data. This can be implemented through a separate data editing interface or open API interfaces, such as providing advanced teacher accounts with knowledge point version management tools and review process configuration functions.

[0034] The access control unit dynamically assigns operation permissions based on account type when teachers log in. When a teacher's client requests to modify standard knowledge point data, the system first verifies its permission level, allowing only senior-permission accounts to enter the data editing interface. For example, a regular teacher account will be redirected to a prompt page when attempting to access the knowledge point modification page, while a senior teacher account can directly enter the editing interface to adjust content or add knowledge point entries. After the data modification request is submitted, the modified content must undergo semantic conflict detection by the standardization update unit, and the change history is recorded through a version control mechanism to ensure the integrity and traceability of the standard knowledge base.

[0035] Existing teaching systems typically lack hierarchical management of teacher access permissions, allowing any teacher to freely modify standard knowledge point data, which can easily lead to confusion in the description of knowledge points. This solution, however, achieves fine-grained control of access permissions through a permission management unit, allowing only teachers with teaching experience or qualifications to maintain core data, effectively preventing unapproved modifications from polluting the standard knowledge base.

[0036] This application addresses the problem of standard knowledge point data being arbitrarily tampered with due to the lack of access control in existing systems. It ensures the standardization and authority of knowledge base updates through a hierarchical access control mechanism, while reducing the risk of data errors caused by accidental or unauthoritative modifications.

[0037] The teaching processing module also includes an annotation unit. This unit is primarily used to annotate data in the standard knowledge base that differs from historical teaching content. When the number of annotations for a certain standard teaching knowledge point reaches a preset value, the standardized teaching knowledge point is sent to a teacher's client with senior privileges for modification or content expansion. The annotation unit is a module used to identify and mark data in the standard knowledge base that differs from historical teaching content. Specifically, it can be implemented using a keyword matching algorithm combined with a semantic difference analysis model. By comparing the historical teaching content of different teachers with the original data in the standard knowledge base, the differences in expression are identified and marked. The preset value refers to the threshold number of annotations that triggers the knowledge point review process. This can be implemented using a dynamically adjustable numerical setting method, for example, by setting different threshold ranges based on the subject type or the complexity of the knowledge point. When the number of times the same knowledge point is annotated reaches this threshold, the system automatically initiates the review process.

[0038] The annotation unit continuously analyzes the historical teaching materials and classroom recordings uploaded by teachers, extracts the descriptions of knowledge points, and compares them semantically with the corresponding entries in the standard knowledge base. When more than a preset number of discrepancies are detected for the same knowledge point, the knowledge point is marked as an entry to be optimized and pushed to the client of a teacher with senior privileges. Senior teachers can view the historical discrepancy annotation records of the knowledge point through the client interface, and judge whether the standard expression needs to be adjusted or supplemented with extended content based on their teaching practice experience, ultimately forming updated standardized knowledge point data.

[0039] Existing systems rely on manual periodic checks of historical teaching data to identify differences in knowledge points, which is inefficient and has a high risk of omissions. This solution uses an automated annotation mechanism combined with dynamic threshold settings to capture frequently occurring differences in expression in real time and accurately trigger the optimization process of standardized knowledge points, thus avoiding the subjectivity and lag of manual screening.

[0040] This application enables the automated identification and targeted optimization of controversial knowledge points in the standard knowledge base, ensuring that frequently differing content can be reviewed and corrected by experienced teachers in a timely manner. This mechanism effectively reduces the chaos of teaching resources caused by inconsistent expressions. At the same time, by controlling the review trigger conditions through preset values, it avoids unnecessary review requests generated by the system due to low-frequency occasional differences, thereby improving the efficiency and accuracy of dynamic updates to the knowledge base.

[0041] When classroom teaching content is not included in the standard teaching knowledge point data, the differentiation processing unit generates the unincluded content and sends it to a teacher client with senior privileges. After approval by the teacher on the senior privileged client, the unincluded content is stored in the standard teaching knowledge point data. The differentiation processing unit is a module used to identify the differences between classroom teaching content and the standard knowledge base. Specifically, it can be implemented through semantic analysis algorithms and knowledge base retrieval interfaces to detect unincluded teaching content. The senior privileged teacher client refers to a terminal device with data modification permissions, which can be implemented through an account permission hierarchy mechanism to ensure the accuracy of standard knowledge base updates. Unincluded content refers to classroom teaching knowledge points not covered by the standard knowledge base. Specifically, it can be identified through text matching algorithms to discover knowledge points that need to be supplemented.

[0042] When the classroom teaching content acquired by the data acquisition module cannot match the knowledge points in the standard knowledge base, the differentiation processing unit marks it as uncollected content. The permission management unit of the teacher teaching platform then filters out senior teacher clients with approval permissions. The uncollected content that has been approved is directly stored in the standard knowledge base after being processed by the standardization update unit, without the need for manual entry.

[0043] Traditional methods rely on manual screening of unincluded content and submission for review, resulting in long update cycles and a high risk of omissions. This solution automates the identification of unincluded content and pushes it to the authorized review nodes, reducing manual intervention and enabling new knowledge points to quickly enter the standardized process. This application solves the problem of new knowledge points not being included in the standard knowledge base in a timely manner. By combining hierarchical access control with automated processes, it achieves real-time and accurate knowledge point updates, avoiding knowledge base delays caused by manual operations.

[0044] The teaching processing module is also equipped with a synonym processing unit. When comparing teaching content, if inconsistencies are found, synonym processing is performed first. When experienced teachers input standard teaching knowledge point data from their clients, the synonym processing unit compares the data. If the data belongs to the same knowledge point, the content input by the experienced teachers is summarized and processed.

[0045] The synonym processing unit refers to the unit used to identify and match different words that express the same semantic meaning. Specifically, it can be implemented using semantic similarity calculation models in natural language processing technology, such as synonym matching based on word vectors or pre-trained language models, thereby reducing misjudgments caused by differences in expression during the content comparison stage. Inductive processing refers to the process of integrating content with different expressions but consistent semantics into a unified expression. Specifically, it can be implemented using semantic clustering algorithms or rule template generation methods, such as by extracting core semantic elements and generating standardized description templates, thereby eliminating redundant expressions and optimizing the knowledge base structure.

[0046] During the comparison of classroom teaching content with standard knowledge points, the synonym processing unit performs semantic analysis on the collected text data in real time to identify words or phrases that may be synonymous. For example, it identifies "the formula for the area of ​​a triangle" and "the formula for calculating the area of ​​a triangle" as the same knowledge point. If it finds that they are not consistent but are synonymous, it will not generate a knowledge point to be confirmed, but will directly mark it as a successful match. When experienced teachers add or modify knowledge points through the client, the input content is first compared with the existing knowledge point database by the synonym processing unit. If it detects that there is a synonym in the existing knowledge point, it will trigger the inductive processing process. For example, it will merge "the two legs of an isosceles triangle are equal" and "the two sides of an isosceles triangle are the same length" into a unified expression "the two legs of an isosceles triangle are equal in length".

[0047] Traditional systems rely on manual review to judge expression differences, which leads to a high misjudgment rate and low processing efficiency due to the inability to automatically identify synonyms. In contrast, this solution can quickly identify and eliminate expression differences through automated semantic analysis and synonym matching. At the same time, it reduces knowledge point redundancy through inductive processing and improves the standardization of the knowledge base. This application can effectively reduce comparison errors caused by differences in teachers' expression habits, avoid the repeated generation of knowledge points to be confirmed, and optimize the expression structure of knowledge points through automatic induction, thereby reducing the cost of manual intervention and ensuring the accuracy and consistency of the standard knowledge base.

[0048] The data acquisition module acquires real-time classroom teaching content from the teacher and displays relevant knowledge points on the classroom computer to further expand the teaching content. When the teaching content comparison unit detects inconsistencies between the classroom teaching content acquired by the acquisition module and the standard teaching knowledge point data, it sends the inconsistency information to the teacher's computer. Real-time acquisition of classroom teaching content refers to synchronously capturing video, audio, and courseware data streams during the teaching process through cameras, voice receiving devices, and the teaching computer interface. Specifically, a streaming media transmission protocol can be used to achieve real-time data acquisition to ensure the integrity and timeliness of the teaching content. Displaying relevant knowledge points refers to dynamically presenting knowledge point information from the standard knowledge base that matches the current teaching progress on the teaching computer interface. Specifically, this can be achieved through the associated retrieval interface of the teaching resource library to assist teachers in instantly accessing standardized teaching materials. Sending inconsistency information means that when semantic analysis detects a deviation between the teacher's statement and the standard knowledge points, the difference is pushed to the teaching terminal in the form of a visual prompt. Specifically, message queue technology can be used to achieve asynchronous notification to trigger the teacher to instantly check and correct the teaching content.

[0049] During classroom teaching, the data acquisition module continuously captures the teacher's lecture video stream, audio stream, and courseware operation records, and transmits the data stream to the teaching processing module. The teaching content comparison unit analyzes the semantic information collected in real time and matches it item by item with the knowledge points in the standard knowledge base. When a difference is detected between the teacher's description of a specific knowledge point and the standard data, the system automatically generates a prompt message containing the location of the difference and the standard description, which is displayed in real time through the interactive interface of the teaching computer. The teacher can adjust the teaching content in time according to the prompt message. At the same time, the teaching computer simultaneously retrieves supplementary materials from the standard knowledge base to assist teaching, such as displaying related charts or supplementary cases. This process ensures the real-time synchronization of classroom teaching content with the standard system through a closed-loop mechanism of real-time data acquisition, dynamic comparison, and instant feedback.

[0050] Traditional teaching systems rely on manual compilation of teaching records after class for knowledge point comparison, which is time-consuming and cannot intervene in the teaching process. This solution, through embedded data acquisition and real-time processing technology, can instantly identify content deviations and trigger correction mechanisms during the teaching process, effectively solving the problem of delayed standardization implementation. At the same time, the function of dynamically linking and displaying standard knowledge points and extended materials breaks through the limitations of traditional systems that only provide static knowledge bases, and realizes the organic integration of teaching content and standard system.

[0051] This application can monitor the matching degree between classroom teaching content and standard knowledge points in real time. When discrepancies in expression are detected, a warning message is immediately pushed to the teacher's end to prevent the continuous spread of incorrect knowledge points. Teachers can quickly adjust the teaching content based on the instant feedback, and improve the teaching plan with the help of standard materials automatically pushed by the system. This mechanism significantly improves the efficiency of standardized teaching execution, ensures the accuracy and consistency of knowledge transmission, and enhances the richness of classroom interaction by dynamically expanding teaching content.

[0052] The teaching resource retrieval system also includes an accent comparison and correction module. This module connects the data acquisition module and the standard knowledge base. It is used to acquire the teacher's lecture audio stream collected by the data acquisition module, perform semantic and speech feature analysis on the pronunciation in the audio stream, compare it with the standard speech data stored in the standard knowledge base, identify differences in expression or pronunciation deviations caused by accents, generate accent correction prompts, and send them to the teacher's teaching computer and teacher client to assist the teacher in adjusting pronunciation. The standard speech data includes standard pronunciation and terminology pronunciation.

[0053] The accent comparison and correction module refers to a device that extracts features and matches patterns from the teacher's lecture speech using speech recognition algorithms. Specifically, it can be implemented using a deep learning-based acoustic model combined with a pronunciation rule base to separate semantic content and pronunciation features in the speech stream. Speech feature analysis refers to the quantification of pitch, intensity, duration, and timbre parameters of the speech signal. Specifically, it can be implemented using Mel-frequency cepstral coefficients combined with formant detection algorithms to construct a digital representation of pronunciation features. Standard speech data refers to a set of pronunciation samples certified by authoritative institutions. Specifically, it can be implemented using a standard terminology pronunciation database recorded by professional broadcasters combined with the International Phonetic Alphabet (IPA) annotation system to establish a benchmark reference system for pronunciation comparison.

[0054] After the data acquisition module acquires the teacher's lecture audio stream in real time, the accent comparison and correction module extracts the pure teacher's pronunciation signal through voiceprint separation technology. Then, the speech recognition engine divides the continuous speech stream into independent syllable units. Each syllable unit is converted by Mel-frequency cepstral coefficients and compared with the standard pronunciation sample of the corresponding term in the standard knowledge base through dynamic time warping to calculate the pronunciation deviation value. When the detected term pronunciation deviation exceeds the preset threshold, the system automatically generates correction prompt information including the deviation type, deviation location and standard pronunciation demonstration, and pushes it synchronously to the interactive interface of the teacher's teaching computer and the teacher's mobile terminal. During the standard knowledge base update process, the term pronunciation data is generated into multi-dialect comparison versions through a professional speech synthesis engine to ensure the comprehensiveness and applicability of the pronunciation benchmark.

[0055] Traditional systems can only identify semantic differences and cannot detect pronunciation deviations of terms caused by regional accents or individual pronunciation habits. This solution achieves real-time monitoring and accurate correction of pronunciation quality in speech streams by constructing a linkage mechanism between a pronunciation feature analysis model and a standard speech database, thus solving the problem of knowledge transfer errors caused by non-standard pronunciation.

[0056] This application can effectively identify pronunciation deviations of terminology caused by regional accents or individual habits during teachers' lectures, automatically generate targeted correction suggestions and provide timely feedback to the teaching terminal. This technical solution can help teachers standardize the pronunciation of professional terms and avoid misunderstandings among students due to inaccurate pronunciation. At the same time, by continuously accumulating standard pronunciation data, it optimizes the speech comparison model and improves the system's adaptability to diverse pronunciation features.

[0057] The system also includes a class management module, which configures an independent class database for each class and records the class's grades each time. When a class's grade in a certain lesson is higher than that of other classes, the system retrieves the video, audio, and courseware of that lesson and improves its ranking in the knowledge base. The class management module generates a teaching effectiveness evaluation report for the teacher of the class based on the chapter test scores and the average mastery of knowledge points in class.

[0058] A class database refers to a storage unit configured separately for each class to store the class's exclusive teaching process data. This can be implemented using relational databases or distributed storage technologies. Its function is to achieve isolated storage and targeted analysis of data from different classes. Grade records refer to the structured storage of the results of each class's course tests or exams. This can be achieved through data acquisition interfaces connected to the academic affairs system, providing quantitative evidence for subsequent teaching resource evaluation. Video, audio, and courseware acquisition refers to extracting teaching materials related to high-scoring classes from the data acquisition module. This can be achieved through timestamp matching or chapter tag association, establishing a relationship between teaching outcomes and teaching resources. Ranking improvement refers to adjusting the priority weights of teaching resources in the knowledge base, which can be implemented using dynamic scoring algorithms. This ensures that high-quality teaching resources are prioritized for retrieval and access. The teaching effectiveness evaluation report is a comprehensive analysis document based on exam scores and knowledge point mastery data. This can be generated into charts and reports using data visualization tools, providing data support for teachers to improve their teaching methods.

[0059] The class management module records chapter exam scores for each class through an independent database. When it detects that a class's score in a specific course is significantly higher than other classes, it automatically triggers the teaching resource collection process. At this time, the classroom teaching videos, audio recordings, and courseware content for the corresponding chapter of that class are extracted and associated with existing resources in the knowledge base. By dynamically adjusting the sorting algorithm, the teaching resources of high-scoring classes are given higher priority in the knowledge base, allowing other teachers to obtain high-quality resources first when searching. At the same time, the module combines exam score distribution data with the knowledge point mastery rate collected in real time in the classroom to generate a report containing quantitative indicators of teaching effectiveness. This report can reflect the correlation between teachers' teaching strategies and students' learning outcomes.

[0060] Traditional teaching systems rely on human experience to select high-quality resources and lack an automated evaluation mechanism based on objective data. This solution establishes a data collection and analysis system at the class level, which can directly link teaching outcomes and teaching process data to achieve an objective quantitative evaluation of the value of teaching resources. Existing technologies cannot dynamically adjust the knowledge base ranking based on class performance, while this solution enables the knowledge base to continuously absorb effective teaching resources that have been verified in practice through automated data association and priority calculation.

[0061] This application can automatically identify teaching resources associated with high-scoring teaching outcomes and recommend them to other teachers, solving the problem of inaccurate resource identification caused by the reliance on subjective judgment in traditional systems. The teaching effectiveness evaluation report provides teachers with improvement directions based on objective data, avoiding the subjectivity and lag of manual evaluation. The dynamic sorting mechanism of the knowledge base ensures the continuous optimization and integration of high-quality resources, improving the practicality and update efficiency of the teaching resource system.

[0062] The knowledge repository is connected to a storage management unit, which records the frequency of access to each knowledge point and classifies the data in the knowledge repository according to the access frequency. The knowledge repository includes a high-frequency repository, a low-frequency repository, and an archive repository. The high-frequency repository stores data whose access frequency reaches a preset value and classroom teaching content of high-scoring classes in the subject. The low-frequency repository stores data whose access frequency does not reach the preset value. The archive repository stores data that has not been accessed within a preset time.

[0063] The storage management unit is a functional module used to monitor and count the number of times knowledge points are accessed. This can be achieved by combining database log analysis technology with scheduled statistical tasks. By recording the timestamp and source information of each knowledge point retrieval or access, it generates access frequency data. Access frequency refers to the statistical value of the number of times a knowledge point is retrieved or used per unit of time. This can be achieved using a sliding window algorithm or a fixed-period statistical method to quantify the activity level of knowledge points. The high-frequency repository is an independent storage partition that prioritizes storing highly active data. This can be achieved using in-memory databases or caching technology to quickly respond to high-frequency access requests. The low-frequency repository is an independent storage partition that stores low-activity data. This can be achieved using conventional disk storage technology to balance access efficiency by reducing storage costs. The archive repository is an offline storage partition that stores data that has not been accessed for a long time. This can be achieved using cold backup storage devices or cloud storage services to free up main storage space and retain historical data.

[0064] The storage management unit continuously monitors the call records of knowledge points and counts the number of calls for each knowledge point according to a preset period. When the number of calls for a knowledge point reaches a preset threshold, its associated teaching content and classroom data of high-scoring classes will be migrated to the high-frequency storage repository to ensure fast access to hot content. Data with a recent call record but a lower call count is retained in the low-frequency storage repository, while data that has not been called for more than a set time is transferred to the archive storage repository. For example, when the trigonometric function knowledge point is called more than 200 times in the monthly exam period, its courseware, teaching videos of high-scoring classes, and associated test questions will be centrally stored in the high-frequency storage repository, and the content of this repository can be displayed first when the teacher's client searches.

[0065] Existing systems typically manage teaching resources using fixed classifications or manual annotations, which cannot dynamically adjust storage strategies based on actual usage. This solution achieves intelligent lifecycle management of teaching resources through automated call frequency statistics and hierarchical storage mechanisms. This not only improves the access efficiency of high-frequency resources and optimizes the utilization rate of storage resources, but also enhances the dissemination effect of high-quality teaching resources by associating them with high-scoring class data.

[0066] This application can automatically identify hot knowledge points in teaching practice, centrally manage frequently used content and high-quality class teaching resources, and significantly improve teachers' retrieval efficiency; reduce the storage cost of low-frequency data through a hierarchical storage strategy to avoid invalid resources occupying system performance; and release storage space through an archiving mechanism to ensure the long-term stability of the system.

[0067] The knowledge repository is equipped with an error database. When the teaching processing module confirms the standardized expression of knowledge points based on feedback information, it stores the identified discrepancies and the knowledge point expressions that are ultimately determined to be incorrect into the error database. The teaching processing module also includes a test question generation unit, which generates test questions based on the standard teaching knowledge point data in the knowledge repository.

[0068] An error database is a database used to store errors or inaccuracies found during the teaching process that deviate from the standard knowledge points. It can be implemented using a distributed database or a relational database. By setting up a data table structure to record the error content, the time of discovery, and the related knowledge points, it is convenient for subsequent teaching analysis and error correction. The test question generation unit is a functional module that automatically generates test questions based on the standard knowledge points. It can be implemented using natural language processing technology combined with preset test question templates. For example, multiple-choice or fill-in-the-blank questions can be generated through keyword extraction and grammatical recombination to test students' mastery of knowledge points.

[0069] Once the teaching processing module generates knowledge points to be confirmed through the differentiation processing unit and receives teacher feedback, if it confirms that there are errors or differences in expression for the knowledge point, it links and stores the corresponding original teaching content and the corrected standardized expression in the error question bank. The question generation unit extracts core concepts through semantic parsing based on the knowledge point data in the standard knowledge base and automatically generates questions based on preset question type rules. For example, for mathematical formula knowledge points, it can generate practice questions that include variable substitution; for historical event knowledge points, it can generate time sequence questions or cause-and-effect relationship judgment questions. The generated questions can be pushed to the teacher's client for classroom testing or stored in the teaching resource library as shared resources.

[0070] Traditional systems lack a systematic recording and analysis mechanism for incorrect knowledge points. Teachers need to manually compile error-prone points, which cannot form structured data to support teaching improvement. This solution achieves automated archiving of incorrect statements through an error question bank. Combined with the test question generation function, it transforms standard knowledge points into quantifiable teaching assessment tools, solving the problems of low efficiency in manual question generation and difficulty in tracing incorrect knowledge points.

[0071] This application can systematically accumulate erroneous statements generated during the teaching process, providing teachers with targeted teaching improvement basis. At the same time, it reduces teachers' workload through automated test question generation, improves the timeliness of knowledge point assessment and teaching effectiveness evaluation, and the establishment of the error question bank further optimizes the data quality of the standard knowledge base and avoids the recurrence of erroneous statements. The test question generation unit strengthens the closed-loop verification mechanism of knowledge point mastery.

Claims

1. A teaching resource retrieval system, characterized by comprising: The application relates to a teaching system, which comprises the following parts: a data collection module, which is configured with a camera in a classroom, a voice receiving device and an interface connected with a classroom teaching computer, and is used for collecting teacher teaching video stream, teacher teaching voice stream and current teaching courseware content; a knowledge storage, which comprises a temporary database for storing temporary information, a teaching resource library and a standard knowledge base, and the standard knowledge base stores a plurality of standard teaching knowledge point data; a teacher teaching platform, which is connected with the knowledge storage and is configured with a teacher client for uploading personal teaching courseware to the teaching resource library and downloading data in the standard knowledge base; a teaching processing module, which is connected with the data collection module and the knowledge storage; the teaching processing module comprises the following parts: a teaching content comparison unit, which is used for performing semantic analysis on classroom teaching content collected by the collection module and comparing the classroom teaching content with standard teaching knowledge point data in the standard knowledge base, and comparing description contents of different teachers on the same knowledge point; a differential processing unit, which generates a to-be-confirmed knowledge point when the classroom teaching content is inconsistent with the standard teaching knowledge point data or the description contents of different teachers on the same knowledge point are inconsistent, sends the to-be-confirmed knowledge point to a corresponding or different teacher client and receives feedback information of the corresponding or different teacher on the to-be-confirmed knowledge point; a standardization updating unit, which determines a standardized expression of the knowledge point according to the feedback information and updates the content of the standard teaching knowledge point data in the standard knowledge base.

2. The educational resource retrieval system of claim 1, wherein, The teacher teaching platform further comprises a permission management unit, which is used for distinguishing between ordinary permission and senior permission of the teacher client, and the teacher client with the senior permission has an entry for modifying and adding the standard teaching knowledge point data.

3. A teaching resources retrieval system according to claim 2, wherein The teaching processing module is further configured with a marking unit, which is mainly used for marking data in the standard knowledge base which has differences in historical teaching content, and when the marking number of a certain standard teaching knowledge point data reaches a preset value, the standard teaching knowledge point is sent to the teacher client with the senior permission for modification or content expansion.

4. The educational resource retrieval system of claim 2, wherein, When the classroom teaching content is not included in the standard teaching knowledge point data, the differential processing unit generates non-included content and sends the non-included content to the teacher client with the senior permission, and after the non-included content is approved by the teacher with the senior permission, the non-included content is stored in the standard teaching knowledge point data.

5. The educational resource retrieval system of claim 2, wherein, The teaching processing module is further configured with a synonym processing unit, which is used for processing synonyms when the teaching content comparison unit finds that the content is inconsistent; when the teacher client with the senior permission inputs the standard teaching knowledge point data, the synonym processing unit is used for comparison, and when it is found that the input content belongs to the same knowledge point, the input content of the teacher client with the senior permission is processed.

6. The educational resource retrieval system of claim 1, wherein, The data collection module acquires classroom teaching content of teacher teaching in a classroom in real time, displays related knowledge point content in the classroom teaching computer to further expand the teaching content, and when the teaching content comparison unit finds that the classroom teaching content collected by the collection module is inconsistent with the standard teaching knowledge point data, the inconsistency information is sent to the teacher teaching computer.

7. A teaching resources retrieval system according to claim 6, wherein The system further comprises an accent comparison correction module connected to the data acquisition module and the standard knowledge base, configured to obtain the teacher's teaching voice stream collected by the data acquisition module, perform semantic and voice feature analysis on the pronunciation in the voice stream, and compare with the standard voice data stored in the standard knowledge base, identify the expression difference or pronunciation deviation caused by the accent, generate accent correction prompt information, and send it to the teacher's teaching computer and the teacher's client to assist the teacher in adjusting the pronunciation.

8. The educational resource retrieval system of claim 1, wherein, The system further comprises a class management module, which configures an independent class database for each class and records the scores of each class. When the score of a certain class for a certain course is higher than that of other classes, the video, audio and courseware of the class for the chapter are obtained, and the sorting of the knowledge base is improved. The class management module generates a teaching effect evaluation report for the teacher of the class according to the chapter examination scores and the average knowledge point mastery data in the classroom.

9. A teaching resources retrieval system according to claim 8, wherein, The knowledge storage library is connected with a storage management unit for recording the calling frequency of each knowledge point and classifying and managing the data in the knowledge storage library according to the calling frequency. The knowledge storage library comprises: A high-frequency storage library for storing data with a calling frequency reaching a preset value and classroom teaching content of high-score classes in the subject; A low-frequency storage library for storing data with a calling frequency not reaching a preset value; An archive storage library for storing data not called within a preset time.

10. The educational resource retrieval system of claim 1, wherein, The knowledge storage library is configured with a mistake bank. When the teaching processing module confirms the standardized expression of the knowledge point according to the feedback information, the knowledge point expression content with differences and finally determined as errors is stored in the mistake bank. The teaching processing module further comprises a test question generation unit for generating test questions based on the standard teaching knowledge point data in the knowledge base.

Images