Education resource sharing method and system based on education cloud platform
By using information upload, data preprocessing, classification and management modules on the education cloud platform, the natural language processing and support vector machine algorithms are used to realize the automatic classification and label generation of educational resources, and the problem of low manual management efficiency is solved. It improves the intelligence and retrieval efficiency of resource management, and ensures the accuracy and completeness of resource information.
Patent Information
- Application Number
- CN202410947612.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-16
- Publication Date
- 2024-11-12
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The manual classification and management methods of educational resources in the existing education cloud platform are time-consuming and labor-intensive, and are prone to errors, resulting in inaccurate resource classification and low retrieval efficiency, increasing the complexity of resource management, and it is difficult to meet the needs of efficient management and rapid retrieval.
The educational resource sharing system based on the education cloud platform includes information upload module, data preprocessing module, classification module and management module. It realizes the automatic classification and label generation of resources through natural language processing technology and support vector machine classification algorithm, and provides multi-dimensional The search function supports keyword search, classification filtering, label filtering and metadata filtering.
It realizes efficient management and intelligent classification of educational resources, improves resource utilization and user experience, ensures the accuracy and completeness of resource information, simplifies the resource import process, and reduces system maintenance costs.
Smart Images

Figure CN118941423A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of cloud platforms, and in particular to an educational resource sharing method and sharing system based on an educational cloud platform. Background Art
[0002] The educational cloud platform is one of the important applications of educational technology. It aims to improve the utilization efficiency and teaching effectiveness of educational resources through cloud computing technology. The educational resource sharing system includes teaching materials in various forms such as video, audio, and documents. How to efficiently manage and classify these resources has become a key issue in the system.
[0003] At present, many educational institutions rely on the traditional method of manual classification and management of teaching resources. This method is not only time-consuming and labor-intensive, but also prone to errors, resulting in inaccurate resource classification and inefficient retrieval. Manual management may also lead to omissions or duplicate storage of resources due to human factors, increasing the complexity of resource management. With the continuous increase in the number of teaching resources, traditional methods are obviously difficult to meet the needs of efficient management and rapid retrieval, affecting the utilization of resources by teachers and students.
[0004] Due to the shortcomings of manual classification and management of teaching resources, educational institutions face many challenges in resource management. For example, inaccurate classification makes it difficult for teachers and students to find the required resources, wasting a lot of time. In addition, the omission and duplicate storage of resources not only takes up a lot of storage space, but also increases the maintenance cost of the system. What is more serious is that low retrieval efficiency may cause students to miss key learning materials and affect learning outcomes. Traditional resource management methods are becoming increasingly incapable in the modern educational environment, and there is an urgent need to introduce intelligent classification and management technologies to improve them. Summary of the invention
[0005] In view of the deficiencies in the prior art, the present invention provides an educational resource sharing method and sharing system based on an educational cloud platform, which solves the problems mentioned in the background technology.
[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: an educational resource sharing system based on an educational cloud platform, including an information uploading module, a data preprocessing module, a classification module, a metadata module and a management module;
[0007] The information upload module provides multiple upload interfaces, supports uploading of different types of educational resource data, and supports users to upload educational resource data in batches;
[0008] The data preprocessing module preprocesses the uploaded educational resource data, including data clarity, data conversion, data standardization and multimedia processing, and simultaneously performs tagging processing and marking to form processed educational resource data;
[0009] The classification module uses natural language processing technology to analyze the content and title of text resources, extract keywords, topics and feature vectors, and then uses the support vector machine classification algorithm to establish a resource classification recognition model, and simultaneously trains the resource classification recognition model to obtain the classification result f, and then generates a classification label for the resource according to the classification result f, and associates it with the resource tag;
[0010] The metadata module extracts metadata of educational resource data with a tag association relationship, including title, author, creation time, keywords and classification label information, and then associates the extracted metadata with the resource data file, so that users can obtain complete metadata information and associated educational resource data when searching and using resources;
[0011] The management module performs life cycle management on educational resource data, including resource uploading, classification, storage, retrieval, downloading and deletion operations, and provides educational resource retrieval functions, supporting keyword search, classification screening, tag screening, metadata screening and related educational resource data retrieval.
[0012] Preferably, the information uploading module includes a collection unit;
[0013] The acquisition unit supports uploading of different types of educational resource data through multiple upload interfaces provided, supports users to upload educational resource data in batches, and simultaneously performs format verification on the uploaded educational resource data to filter out files that do not conform to the specified format;
[0014] The upload interface includes a document file upload interface, a video file upload interface, an audio file upload interface, and a picture file upload interface;
[0015] The upload interface for document files includes PDF, DOCX and TXT document file formats; the upload interface for video files includes MP4, AVI and MKV video file formats; the upload interface for audio files includes MP3, WAV and AAC audio file formats; the upload interface for image files includes JPEG, PNG and GIF image file formats.
[0016] Preferably, the data preprocessing module includes a processing unit and a marking unit;
[0017] The processing unit pre-processes the uploaded educational resource data, including data clarity, data conversion, data standardization and multimedia processing;
[0018] Among them, data cleaning includes noise removal, error correction and duplicate data processing; data conversion includes format conversion, encoding conversion and file compression; data standardization includes text standardization, metadata standardization and numerical standardization; multimedia processing includes audio processing, video processing and image processing;
[0019] The marking unit marks the pre-processed educational resource data, including text marking, audio marking, video marking, image marking and upload log information marking, to form the processed educational resource data.
[0020] Preferably, the classification module includes an extraction unit and a modeling unit;
[0021] The extraction unit uses natural language processing technology to analyze the content and title of the text resource, extract keywords, topics and feature vectors, wherein keyword extraction includes extraction using TF-IDF and TextRank algorithms; topic extraction includes analysis and extraction of the topic of the text using a latent Dirichlet allocation topic model; feature vector extraction includes conversion of text into feature vector representation using a bag of words model;
[0022] The modeling unit forms a feature vector matrix X and a corresponding label vector Y as input data based on the extracted keywords, topics and feature vector information, and then uses a support vector machine classification algorithm to establish a resource classification and recognition model, and simultaneously trains the resource classification and recognition model to obtain a classification result f, and then generates a classification label for the resource based on the classification result f, and associates it with a resource tag, including a positive resource tag association relationship and a negative resource tag association relationship.
[0023] Preferably, the feature vector matrix X and the corresponding label vector Y are obtained by the following steps:
[0024] Step 1: Calculate the term frequency (TF) and inverse document frequency (IDF) of each document to obtain the TF-IDF value of each word;
[0025] TF-IDF(t,d)=TF(t,d)*IDF(t);
[0026] In the formula, TF(t,d) represents the frequency of keyword t appearing in document d, and IDF(t) is calculated by The calculation formula is obtained, which represents the inverse document frequency of keyword t, N represents the total number of documents, and |d∈D:t∈d| represents the number of documents containing keyword t;
[0027] Step 2: Construct the eigenvector matrix X:
[0028]
[0029] Among them, X represents the feature vector matrix, which is specifically an N*m matrix, N represents the amount of document data, and X N Specifically, [TF-IDF(t1,d N )TF-IDF(t2,d N )....TF-IDF(t m ,d N )], m represents the size of the vocabulary, TF-IDF(t1,d N ) represents the TF-IDF value of keyword t1 in the Nth document;
[0030] Step 3: Construct the label vector Y, which contains the classification labels L of N documents:
[0031]
[0032] In the formula, Y represents the label vector, which is a vector of length N. N Represents document d N The label value can be 1, 2, 3, ..., L.
[0033] Preferably, the classification result f is obtained by the following calculation steps:
[0034] Step 1: Obtain a hyperplane to maximize the interval between categories through the support vector machine classification algorithm:
[0035] w*X i +b=0;
[0036] In the formula, w represents the weight vector, X i Represents the i-th sample in the input feature vector matrix X i , b represents the bias vector;
[0037] Step 2: Maximize the gap between the two categories through the support vector machine classification algorithm, while minimizing the classification error:
[0038]
[0039] In the formula, ||w|| 2 represents the square of the second norm of the weight vector, C represents the regularization parameter, which is specifically used to balance the weight ratio of the maximum interval and classification error, ξ i represents the slack variable;
[0040] Among them, the slack variable ξ i Specifically represents the classification error of the i-th sample, ξ i ≥0; when sample i is misclassified, ξ i >0;
[0041] Step 3: Set constraints, including setting correct classification constraints:
[0042]
[0043] Where Y i *(w*X i +b) represents the i-th sample X in the eigenvector matrix i The projection value on the hyperplane is multiplied by the i-th label value Y in the label vector i , 1-ξ i Represents the error tolerance range introduced by the slack variable. Indicates that the slack variable must be non-negative;
[0044] Among them, when Y i =1 indicates positive sample X i , forming w*X i +b≥1-ξ i ;
[0045] When Y i =-1 indicates negative sample X i , forming w*X i +b≥-1+ξ i ;
[0046] When i = 0, indicating no misclassification. i >0 indicates sample X i misclassified;
[0047] Step 4: Execute the classification decision function:
[0048] f(X i )=w*X i +b;
[0049] In the formula, f(X i ) represents the output value of the classification decision function, specifically the i-th sample X in the input feature vector matrix i The classification result, w*X i Represents the weight vector and the i-th sample X in the input feature vector matrix i The inner product of b represents the bias vector.
[0050] Preferably, the resource classification label is generated according to the classification result f, and the resource classification label is generated according to the following classification rules:
[0051] When f(X i )≥0, then the i-th sample X in the input eigenvector matrix i Belongs to the positive class label;
[0052] When f(X i )<0, then the i-th sample X in the input eigenvector matrix i Belongs to the negative class label;
[0053] Among them, 0 represents the threshold of the hyperplane, specifically represents the i-th sample X in the input feature vector matrix i The distance relative to the hyperplane in feature space.
[0054] Preferably, the metadata module includes an association unit;
[0055] The association unit extracts metadata of the educational resource data with a tag association relationship, including title, author, creation time, keywords and classification tag information, and then associates the extracted metadata with the resource data file, including classifying the resources according to the obtained positive and negative tags, and storing and associating the classification results as part of the metadata, so that users can obtain complete metadata information and associated educational resource data when retrieving and using resources;
[0056] Among them, positive class labeling processing: for resources marked as positive class, add corresponding classification labels, including: Category 1;
[0057] Negative labeling processing: For resources marked as positive, add corresponding classification labels, including: Category 2.
[0058] Preferably, the management module includes a control unit;
[0059] The control unit performs life cycle management on educational resource data, including resource upload, classification, storage, retrieval, download and deletion operations, and provides educational resource retrieval functions, supporting keyword search, classification screening, tag screening, metadata screening and related educational resource data retrieval;
[0060] Among them, classification includes automatically classifying resources according to extracted metadata and classification tags, supporting users to manually classify resources, providing flexible classification management functions and updating existing classifications.
[0061] The educational resource sharing method based on the educational cloud platform includes the following steps:
[0062] Step 1: The information upload module provides multiple upload interfaces, supports uploading different types of educational resource data, and supports users to upload educational resource data in batches;
[0063] Step 2: The data preprocessing module preprocesses the uploaded educational resource data, including data clarity, data conversion, data standardization and multimedia processing, and simultaneously performs tagging and marking to form the processed educational resource data;
[0064] Step 3: The classification module uses natural language processing technology to analyze the content and title of text resources, extract keywords, topics and feature vectors, and then uses the support vector machine classification algorithm to establish a resource classification recognition model. The resource classification recognition model is trained simultaneously to obtain the classification result f, and then the classification label of the resource is generated according to the classification result f, and the association relationship is marked with the resource.
[0065] Step 4: The metadata module extracts metadata of educational resource data with tag association relationships, including title, author, creation time, keywords and classification label information, and then associates the extracted metadata with the resource data file, so that users can obtain complete metadata information and associated educational resource data when searching and using resources;
[0066] Step 5: The management module performs life cycle management on educational resource data, including resource uploading, classification, storage, retrieval, downloading and deletion operations. It also provides educational resource retrieval functions, supporting keyword search, classification filtering, tag filtering, metadata filtering and related educational resource data retrieval.
[0067] The present invention provides an educational resource sharing method and sharing system based on an educational cloud platform, which has the following beneficial effects:
[0068] (1) When the system is running, the information upload module, data preprocessing module, classification module, metadata module and management module work together to achieve efficient management of educational resource data. By providing upload interfaces for various types of educational resource data and supporting batch upload, the resource import process is simplified. The multi-level processing of the data preprocessing module ensures the quality and consistency of the data. The classification module uses natural language processing technology and support vector machine classification algorithm to achieve automatic classification and label generation of resources, thereby improving the intelligent level of resource management. The metadata module extracts and associates metadata information of resources to ensure that users can obtain complete resource information and associated data when retrieving and using resources. The management module comprehensively covers the life cycle management of resources, from uploading, classification, storage, retrieval to downloading and deletion, providing comprehensive resource operation support. Through multi-dimensional retrieval functions, it greatly improves the user's resource access and usage experience, thereby improving the efficiency and accuracy of educational resource management. At the same time, it also enhances user experience and resource utilization, and achieves efficient sharing and intelligent management of educational resources.
[0069] (2) Based on the extracted information, the feature vector matrix X and label vector Y are constructed, and the support vector machine classification algorithm is used to establish a resource classification and recognition model, which is trained and optimized to obtain the classification result f. Based on the classification result f, the classification label of the resource is generated and associated with the resource tag to distinguish between positive and negative resources. The system ensures high accuracy and effectiveness of classification through strict steps, including calculating word frequency and inverse document frequency, constructing feature vector matrix and label vector, using support vector machine to maximize category interval and minimize classification error, setting classification constraints, and executing classification decision function. The final classification label generation rule is based on the output value of the classification decision function, which enables the system to automatically and efficiently classify and manage educational resources accurately.
[0070] (3) By classifying and storing positively labeled and negatively labeled resources, it is ensured that users can obtain complete and accurate metadata information and related educational resource data when retrieving and using resources. The management and control unit performs comprehensive life cycle management of educational resource data, covering upload, classification, storage, retrieval, download and deletion operations, and provides multi-dimensional retrieval functions, including keyword search, category screening, tag screening, metadata screening and related resource retrieval. Through the flexible management methods of automatic classification and manual classification, as well as the timely update of existing classifications, the accuracy and efficiency of resource management are improved. BRIEF DESCRIPTION OF THE DRAWINGS
[0071] Figure 1 It is a schematic diagram of the flow chart of the educational resource sharing system based on the educational cloud platform of the present invention;
[0072] Figure 2 It is a schematic diagram of the steps of the educational resource sharing method based on the educational cloud platform of the present invention. DETAILED DESCRIPTION
[0073] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.
[0074] Example 1
[0075] The present invention provides an educational resource sharing system based on an educational cloud platform. Figure 1 , including information uploading module, data preprocessing module, classification module, metadata module and management module;
[0076] The information upload module provides multiple upload interfaces, supports uploading of different types of educational resource data, and supports users to upload educational resource data in batches;
[0077] The data preprocessing module preprocesses the uploaded educational resource data, including data clarity, data conversion, data standardization and multimedia processing, and simultaneously performs tagging processing and marking to form processed educational resource data;
[0078] The classification module uses natural language processing technology to analyze the content and title of text resources, extract keywords, topics and feature vectors, and then uses the support vector machine classification algorithm to establish a resource classification recognition model, and simultaneously trains the resource classification recognition model to obtain the classification result f, and then generates a classification label for the resource according to the classification result f, and associates it with the resource tag;
[0079] The metadata module extracts metadata of educational resource data with a tag association relationship, including title, author, creation time, keywords and classification label information, and then associates the extracted metadata with the resource data file, so that users can obtain complete metadata information and associated educational resource data when searching and using resources;
[0080] The management module performs life cycle management on educational resource data, including resource uploading, classification, storage, retrieval, downloading and deletion operations, and provides educational resource retrieval functions, supporting keyword search, classification screening, tag screening, metadata screening and related educational resource data retrieval.
[0081] In this embodiment, efficient management of educational resource data is achieved through the collaborative work of the information upload module, data preprocessing module, classification module, metadata module and management module. By providing upload interfaces for various types of educational resource data, batch uploading is supported, and the resource import process is simplified. The quality and consistency of the data are ensured through multi-level processing of the data preprocessing module. The classification module uses natural language processing technology and support vector machine classification algorithm to achieve automatic classification and label generation of resources, thereby improving the intelligent level of resource management. The metadata module extracts and associates metadata information of resources to ensure that users can obtain complete resource information and associated data when retrieving and using resources. The management module comprehensively covers the life cycle management of resources, from uploading, classification, storage, retrieval to downloading and deletion, providing comprehensive resource operation support, and through multi-dimensional retrieval functions, it greatly improves the user's resource access and usage experience, thereby improving the efficiency and accuracy of educational resource management, while also enhancing user experience and resource utilization, and achieving efficient sharing and intelligent management of educational resources.
[0082] Example 2
[0083] This embodiment is explained in Example 1, please refer to Figure 1,Specifically: the information uploading module includes a collection unit;
[0084] The acquisition unit supports uploading of different types of educational resource data through multiple upload interfaces provided, supports users to upload educational resource data in batches, and simultaneously performs format verification on the uploaded educational resource data to filter out files that do not conform to the specified format;
[0085] The upload interface includes a document file upload interface, a video file upload interface, an audio file upload interface, and a picture file upload interface;
[0086] The upload interface for document files includes PDF, DOCX and TXT document file formats; the upload interface for video files includes MP4, AVI and MKV video file formats; the upload interface for audio files includes MP3, WAV and AAC audio file formats; the upload interface for image files includes JPEG, PNG and GIF image file formats.
[0087] The data preprocessing module includes a processing unit and a marking unit;
[0088] The processing unit pre-processes the uploaded educational resource data, including data clarity, data conversion, data standardization and multimedia processing;
[0089] Among them, data cleaning includes noise removal, error correction and duplicate data processing; data conversion includes format conversion, encoding conversion and file compression; data standardization includes text standardization, metadata standardization and numerical standardization; multimedia processing includes audio processing, video processing and image processing;
[0090] The marking unit marks the pre-processed educational resource data, including text marking, audio marking, video marking, image marking and upload log information marking, to form the processed educational resource data.
[0091] Example 3
[0092] This embodiment is explained in Example 1, please refer to Figure 1 ,Specifically: the classification module includes an extraction unit and a modeling unit;
[0093] The extraction unit uses natural language processing technology to analyze the content and title of the text resource, extract keywords, topics and feature vectors, wherein keyword extraction includes extraction using TF-IDF and TextRank algorithms; topic extraction includes analysis and extraction of the topic of the text using a latent Dirichlet allocation topic model; feature vector extraction includes conversion of text into feature vector representation using a bag of words model;
[0094] The modeling unit forms a feature vector matrix X and a corresponding label vector Y as input data based on the extracted keywords, topics and feature vector information, and then uses a support vector machine classification algorithm to establish a resource classification and recognition model, and simultaneously trains the resource classification and recognition model to obtain a classification result f, and then generates a classification label for the resource based on the classification result f, and associates it with a resource tag, including a positive resource tag association relationship and a negative resource tag association relationship.
[0095] The feature vector matrix X and the corresponding label vector Y are obtained by the following steps:
[0096] Step 1: Calculate the term frequency (TF) and inverse document frequency (IDF) of each document to obtain the TF-IDF value of each word;
[0097] TF-IDF(t,d)=TF(t,d)*IDF(t);
[0098] In the formula, TF(t,d) represents the frequency of keyword t appearing in document d, and IDF(t) is calculated by The calculation formula is obtained, which represents the inverse document frequency of keyword t, N represents the total number of documents, and |d∈D:t∈d| represents the number of documents containing keyword t;
[0099] Step 2: Construct the eigenvector matrix X:
[0100]
[0101] Among them, X represents the feature vector matrix, which is specifically an N*m matrix, N represents the amount of document data, and X N Specifically, [TF-IDF(t1,d N )TF-IDF(t2,d N )....TF-IDF(t m ,d N )], m represents the size of the vocabulary, TF-IDF(t1,d N ) represents the TF-IDF value of keyword t1 in the Nth document;
[0102] Step 3: Construct the label vector Y, which contains the classification labels L of N documents:
[0103]
[0104] In the formula, Y represents the label vector, which is a vector of length N. N Represents document d N The label value can be 1, 2, 3, ..., L.
[0105] The classification result f is obtained by the following calculation steps:
[0106] Step 1: Obtain a hyperplane to maximize the interval between categories through the support vector machine classification algorithm:
[0107] w*X i +b=0;
[0108] In the formula, w represents the weight vector, X i Represents the i-th sample in the input feature vector matrix X i , b represents the bias vector;
[0109] Step 2: Maximize the gap between the two categories through the support vector machine classification algorithm, while minimizing the classification error:
[0110]
[0111] In the formula, ||w|| 2 represents the square of the second norm of the weight vector, C represents the regularization parameter, which is specifically used to balance the weight ratio of the maximum interval and classification error, ξ i represents the slack variable;
[0112] Among them, the slack variable ξ i Specifically represents the classification error of the i-th sample, ξ i ≥0; when sample i is misclassified, ξ i >0;
[0113] Step 3: Set constraints, including setting correct classification constraints:
[0114]
[0115] Where Y i *(w*X i +b) represents the i-th sample X in the eigenvector matrix i The projection value on the hyperplane is multiplied by the i-th label value Y in the label vector i , 1-ξ i Represents the error tolerance range introduced by the slack variable. Indicates that the slack variable must be non-negative;
[0116] Among them, when Y i =1 indicates positive sample X i , forming w*X i +b≥1-ξ i ;
[0117] When Y i =-1 indicates negative sample X i , forming w*X i +b≥-1+ξi ;
[0118] When i = 0, indicating no misclassification. i >0 indicates sample X i misclassified;
[0119] Step 4: Execute the classification decision function:
[0120] f(X i )=w*X i +b;
[0121] In the formula, f(X i ) represents the output value of the classification decision function, specifically the i-th sample X in the input feature vector matrix i The classification result, w*X i Represents the weight vector and the i-th sample X in the input feature vector matrix i The inner product of b represents the bias vector.
[0122] The classification label of the resource is generated according to the classification result f, and the classification label of the resource is generated according to the following classification rules:
[0123] When f(X i )≥0, then the i-th sample X in the input eigenvector matrix i Belongs to the positive class label;
[0124] When f(X i )<0, then the i-th sample X in the input eigenvector matrix i Belongs to the negative class label;
[0125] Among them, 0 represents the threshold of the hyperplane, specifically represents the i-th sample X in the input feature vector matrix i The distance in feature space relative to the hyperplane.
[0126] In this embodiment, a feature vector matrix X and a label vector Y are constructed based on the extracted information, and a support vector machine classification algorithm is used to establish a resource classification and recognition model, which is trained and optimized to obtain a classification result f. Based on the classification result f, a classification label for the resource is generated and associated with the resource tag to distinguish between positive and negative resources. The system ensures high accuracy and effectiveness of classification through strict steps, including calculating word frequency and inverse document frequency, constructing a feature vector matrix and a label vector, using a support vector machine to maximize category intervals and minimize classification errors, setting classification constraints, and executing a classification decision function. The final classification label generation rule is based on the output value of the classification decision function, so that the system can automatically and efficiently classify and manage educational resources accurately.
[0127] Example 4
[0128] This embodiment is explained in Example 1, please refer to Figure 1 ,Specifically: the metadata module includes an association unit;
[0129] The association unit extracts metadata of the educational resource data with a tag association relationship, including title, author, creation time, keywords and classification tag information, and then associates the extracted metadata with the resource data file, including classifying the resources according to the obtained positive and negative tags, and storing and associating the classification results as part of the metadata, so that users can obtain complete metadata information and associated educational resource data when retrieving and using resources;
[0130] Among them, positive class labeling processing: for resources marked as positive class, add corresponding classification labels, including: Category 1;
[0131] Negative labeling processing: For resources marked as positive, add corresponding classification labels, including: Category 2.
[0132] The management module includes a control unit;
[0133] The control unit performs life cycle management on educational resource data, including resource upload, classification, storage, retrieval, download and deletion operations, and provides educational resource retrieval functions, supporting keyword search, classification screening, tag screening, metadata screening and related educational resource data retrieval;
[0134] Among them, classification includes automatically classifying resources according to extracted metadata and classification tags, supporting users to manually classify resources, providing flexible classification management functions and updating existing classifications.
[0135] In this embodiment, by classifying and storing positively labeled and negatively labeled resources, it is ensured that users can obtain complete and accurate metadata information and associated educational resource data when retrieving and using resources. The management and control unit performs comprehensive life cycle management of educational resource data, covering upload, classification, storage, retrieval, download and deletion operations, and provides multi-dimensional retrieval functions, including keyword search, category screening, label screening, metadata screening and associated resource retrieval. Through the flexible management methods of automatic classification and manual classification, as well as the timely update of existing classifications, the accuracy and efficiency of resource management are improved.
[0136] Example 5
[0137] For educational resource sharing methods based on the educational cloud platform, please refer to Figure 2 , specifically: including the following steps:
[0138] Step 1: The information upload module provides multiple upload interfaces, supports uploading different types of educational resource data, and supports users to upload educational resource data in batches;
[0139] Step 2: The data preprocessing module preprocesses the uploaded educational resource data, including data clarity, data conversion, data standardization and multimedia processing, and simultaneously performs tagging and marking to form the processed educational resource data;
[0140] Step 3: The classification module uses natural language processing technology to analyze the content and title of text resources, extract keywords, topics and feature vectors, and then uses the support vector machine classification algorithm to establish a resource classification recognition model. The resource classification recognition model is trained simultaneously to obtain the classification result f, and then the classification label of the resource is generated according to the classification result f, and the association relationship is marked with the resource.
[0141] Step 4: The metadata module extracts metadata of educational resource data with tag association relationships, including title, author, creation time, keywords and classification label information, and then associates the extracted metadata with the resource data file, so that users can obtain complete metadata information and associated educational resource data when searching and using resources;
[0142] Step 5: The management module performs life cycle management on educational resource data, including resource uploading, classification, storage, retrieval, downloading and deletion operations. It also provides educational resource retrieval functions, supporting keyword search, classification filtering, tag filtering, metadata filtering and related educational resource data retrieval.
[0143] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.
Claims
1. Educational resource sharing system based on education cloud platform, Features: It includes information uploading module, data preprocessing module, classification module, metadata module and management module; The information upload module provides multiple upload interfaces, supports uploading of different types of educational resource data, and supports users to upload educational resource data in batches; The data preprocessing module preprocesses the uploaded educational resource data, including data clarity, data conversion, data standardization and multimedia processing, and simultaneously performs tagging processing and marking to form processed educational resource data; The classification module uses natural language processing technology to analyze the content and title of text resources, extract keywords, topics and feature vectors, and then uses the support vector machine classification algorithm to establish a resource classification recognition model, and simultaneously trains the resource classification recognition model to obtain the classification result f, and then generates a classification label for the resource according to the classification result f, and associates it with the resource tag; The metadata module extracts metadata of educational resource data with a tag association relationship, including title, author, creation time, keywords and classification label information, and then associates the extracted metadata with the resource data file, so that users can obtain complete metadata information and associated educational resource data when searching and using resources; The management module performs life cycle management on educational resource data, including resource uploading, classification, storage, retrieval, downloading and deletion operations, and provides educational resource retrieval functions, supporting keyword search, classification screening, tag screening, metadata screening and related educational resource data retrieval.
2. According to the educational resource sharing system based on the educational cloud platform according to claim 1, Features: The information uploading module includes a collection unit; The acquisition unit supports uploading of different types of educational resource data through multiple upload interfaces provided, supports users to upload educational resource data in batches, and simultaneously performs format verification on the uploaded educational resource data to filter out files that do not conform to the specified format; The upload interface includes a document file upload interface, a video file upload interface, an audio file upload interface, and a picture file upload interface; The upload interface for document files includes PDF, DOCX and TXT document file formats; the upload interface for video files includes MP4, AVI and MKV video file formats; the upload interface for audio files includes MP3, WAV and AAC audio file formats; the upload interface for image files includes JPEG, PNG and GIF image file formats.
3. According to the educational resource sharing system based on the educational cloud platform according to claim 1, Features: The data preprocessing module includes a processing unit and a marking unit; The processing unit pre-processes the uploaded educational resource data, including data clarity, data conversion, data standardization and multimedia processing; Among them, data cleaning includes noise removal, error correction and duplicate data processing; data conversion includes format conversion, encoding conversion and file compression; data standardization includes text standardization, metadata standardization and numerical standardization; multimedia processing includes audio processing, video processing and image processing; The marking unit marks the pre-processed educational resource data, including text marking, audio marking, video marking, image marking and upload log information marking, to form the processed educational resource data.
4. According to the educational resource sharing system based on the educational cloud platform of claim 1, Features: The classification module includes an extraction unit and a modeling unit; The extraction unit uses natural language processing technology to analyze the content and title of the text resource, extract keywords, topics and feature vectors, wherein keyword extraction includes extraction using TF-IDF and TextRank algorithms; topic extraction includes analysis and extraction of the topic of the text using a latent Dirichlet allocation topic model; feature vector extraction includes conversion of text into feature vector representation using a bag of words model; The modeling unit forms a feature vector matrix X and a corresponding label vector Y as input data based on the extracted keywords, topics and feature vector information, and then uses a support vector machine classification algorithm to establish a resource classification and recognition model, and simultaneously trains the resource classification and recognition model to obtain a classification result f, and then generates a classification label for the resource based on the classification result f, and associates it with a resource tag, including a positive resource tag association relationship and a negative resource tag association relationship.
5. According to the educational resource sharing system based on the educational cloud platform according to claim 4, Features: The feature vector matrix X and the corresponding label vector Y are obtained by the following steps: Step 1: Calculate the term frequency (TF) and inverse document frequency (IDF) of each document to obtain the TF-IDF value of each word; TF-IDF(t,d)=TF(t,d)*IDF(t); In the formula, TF(t,d) represents the frequency of keyword t appearing in document d, and IDF(t) is calculated by The calculation formula is obtained, which represents the inverse document frequency of keyword t, N represents the total number of documents, and |d∈D:t∈d| represents the number of documents containing keyword t; Step 2: Construct the eigenvector matrix X: Among them, X represents the feature vector matrix, which is specifically an N*m matrix, N represents the amount of document data, and X N Specifically, [TF-IDF(t 1 ,d N )TF-IDF(t 2 ,d N )....TF-IDF(t m ,d N )], m represents the size of the vocabulary, TF-IDF(t 1 ,d N ) indicates the keyword t 1 The TF-IDF value of the Nth document; Step 3: Construct the label vector Y, which contains the classification labels L of N documents: In the formula, Y represents the label vector, which is a vector of length N. N Represents document d N The label value can be 1, 2, 3, ..., L.
6. According to claim 4, the educational resource sharing system based on the educational cloud platform, Features: The classification result f is obtained by the following calculation steps: Step 1: Obtain a hyperplane to maximize the interval between categories through the support vector machine classification algorithm: w*X i +b=0; In the formula, w represents the weight vector, X i Represents the i-th sample in the input feature vector matrix X i , b represents the bias vector; Step 2: Maximize the gap between the two categories through the support vector machine classification algorithm, while minimizing the classification error: In the formula, ||w|| 2 represents the square of the second norm of the weight vector, C represents the regularization parameter, which is specifically used to balance the weight ratio of the maximum interval and classification error, ξ i represents the slack variable; Among them, the slack variable ξ i Specifically represents the classification error of the i-th sample, ξ i ≥0; when sample i is misclassified, ξ i >0; Step 3: Set constraints, including setting correct classification constraints: Where Y i *(w*X i +b) represents the i-th sample X in the eigenvector matrix i The projection value on the hyperplane is multiplied by the i-th label value Y in the label vector i , 1-ξ i represents the error tolerance range introduced by the slack variable, when ξ i ≥0 Indicates that the slack variable must be non-negative; Among them, when Y i =1 indicates positive sample X i , forming w*X i +b≥1-ξ i ; When Y i =-1 indicates negative sample X i , forming w*X i +b≥-1+ξ i ; When i = 0, indicating no misclassification. i >0 indicates sample X i misclassified; Step 4: Execute the classification decision function: f(X i )=w*X i +b; In the formula, f(X i ) represents the output value of the classification decision function, specifically the i-th sample X in the input feature vector matrix i The classification result, w*X i Represents the weight vector and the i-th sample X in the input feature vector matrix i The inner product of b represents the bias vector.
7. According to claim 4, the educational resource sharing system based on the educational cloud platform, Features: The classification label of the resource is generated according to the classification result f, and the classification label of the resource is generated according to the following classification rules: When f(X i )≥0, then the i-th sample X in the input eigenvector matrix i Belongs to the positive class label; When f(X i )<0, then the i-th sample X in the input eigenvector matrix i Belongs to the negative class label; Among them, 0 represents the threshold of the hyperplane, specifically represents the i-th sample X in the input feature vector matrix i The distance in feature space relative to the hyperplane.
8. According to claim 1, the educational resource sharing system based on the educational cloud platform, Features: The metadata module includes an association unit; The association unit extracts metadata of the educational resource data with a tag association relationship, including title, author, creation time, keywords and classification tag information, and then associates the extracted metadata with the resource data file, including classifying the resources according to the obtained positive and negative tags, and storing and associating the classification results as part of the metadata, so that users can obtain complete metadata information and associated educational resource data when retrieving and using resources; Among them, positive class labeling processing: for resources marked as positive class, add corresponding classification labels, including: Category 1; Negative labeling processing: For resources marked as positive, add corresponding classification labels, including: Category 2.
9. According to claim 1, the educational resource sharing system based on the educational cloud platform, Features: The management module includes a management and control unit; The control unit performs life cycle management on educational resource data, including resource upload, classification, storage, retrieval, download and deletion operations, and provides educational resource retrieval functions, supporting keyword search, classification screening, tag screening, metadata screening and related educational resource data retrieval; Among them, classification includes automatically classifying resources according to extracted metadata and classification tags, supporting users to manually classify resources, providing flexible classification management functions and updating existing classifications.
10. An educational resource sharing method based on an educational cloud platform, comprising the educational resource sharing system based on an educational cloud platform as described in any one of claims 1 to 9 above, Features: The following steps are involved: Step 1: The information upload module provides multiple upload interfaces, supports uploading different types of educational resource data, and supports users to upload educational resource data in batches; Step 2: The data preprocessing module preprocesses the uploaded educational resource data, including data clarity, data conversion, data standardization and multimedia processing, and simultaneously performs tagging and marking to form the processed educational resource data; Step 3: The classification module uses natural language processing technology to analyze the content and title of text resources, extract keywords, topics and feature vectors, and then uses the support vector machine classification algorithm to establish a resource classification recognition model. The resource classification recognition model is trained simultaneously to obtain the classification result f, and then the classification label of the resource is generated according to the classification result f, and the association relationship is marked with the resource. Step 4: The metadata module extracts metadata of educational resource data with tag association relationships, including title, author, creation time, keywords and classification label information, and then associates the extracted metadata with the resource data file, so that users can obtain complete metadata information and associated educational resource data when searching and using resources; Step 5: The management module performs life cycle management on educational resource data, including resource uploading, classification, storage, retrieval, downloading and deletion operations. It also provides educational resource retrieval functions, supporting keyword search, classification filtering, tag filtering, metadata filtering and related educational resource data retrieval.
Citation Information
Patent Citations
Text categorization feature selection and weight computation method based on field knowledge
CN101290626A
Text event classification method based on CHI feature selection
CN104965867A
Knowledge point-oriented educational resource collection and classification method and system
CN111414336A
Education resource sharing method and system based on education big data platform
CN115293724A
Education informatization management platform based on cloud computing
CN115601204A
Cited By
RAG-based multi-format data rapid query method
CN119415544A
Teaching resource library management system and method
CN119807179A
A teaching resource library management system and method
CN119807179B
Method and system for integrating and sharing digital media content resources
CN120091025A
