A teaching data resource retrieval matching processing method and system

Abstract

The application discloses a kind of teaching data resource's retrieval matching processing method and system, the input information of student is analyzed and parsed, finds the most relevant resource with student demand by matching algorithm;Get a target resource and a series of target resource list, realize personalized, intelligent learning resource recommendation;And according to the input information of student and recommended target resource, generate learning task work order and integrate into electronic version resource learning manual, help student systematically carry out learning;The optimal resource flow path corresponding target resource is also pushed to the current student and the student on other nodes on the optimal resource flow path, improve the learning efficiency of student.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of electronic information database, and in particular to a teaching data resource retrieval matching processing method and system. BACKGROUND

[0002] With the rapid development of education technology, traditional paper teaching materials have been difficult to meet the needs of modern students for diversity and interactivity of teaching resources.

[0003] Although the existing paper loose-leaf teaching materials are convenient to carry and use, they have many limitations in resource acquisition, question answering and personalized learning. In the learning process, if students encounter questions, they need to manually search for relevant information, which is time-consuming and laborious, affecting learning efficiency.

[0004] With the rapid development of electronic information technology, many campuses have established electronic information teaching databases that are special and characteristic of their own teaching. Electronic information teaching databases can integrate rich digital teaching resources for retrieval, but further research has found that when students encounter problems in the learning process, they can input problem information, and according to the student's problem information, an interactive digital textbook system is constructed by searching the teaching resources in the electronic information teaching database and intelligently recommending to improve the learning experience and professional skills of students.

[0005] However, the electronic information database in the prior art still has many deficiencies in this regard. First of all, the degree of interest of students is an important basis for personalized learning experience, and the interest points of different students for the same resource may differ significantly, but the existing system often uses a general recommendation algorithm and fails to fully exploit and utilize the interest preferences of students.

[0006] Secondly, the degree of attention to the target resource directly affects the learning efficiency and resource utilization rate, and how to accurately measure and dynamically adjust the recommendation frequency of the resource is the key to improving the teaching effect. In addition, the correlation degree of other students is also an important factor affecting the learning process. Therefore, the electronic database can only achieve more accurate resource recommendation and learning planning by analyzing the interaction and cooperation relationship between students (and fully considering the flowability of learning data in the community), to promote knowledge sharing and cooperative learning. SUMMARY

[0007] The purpose of the present application is to provide a teaching data resource retrieval matching processing method and system, which solves the above technical problems pointed out in the prior art.

[0008] The present application provides a teaching data resource retrieval matching processing method, comprising the following operation steps:

[0009] Verify the identity information of the current student, and after verification, obtain the input information of the student;

[0010] searching and matching based on the input information from a preset digital teaching resource library through multi-dimensional hierarchical analysis;

[0011] The recommended information includes target resources and a target resource list.

[0012] Based on the input information and the target resources, a learning task work order is generated, and an electronic version resource learning manual is obtained by integrating the target resources and the learning task work order. Meanwhile, an optimal resource flow path and target resources corresponding to the optimal resource flow path are obtained, and the target resources corresponding to the optimal resource flow path are pushed to the current student and other students on the optimal resource flow path. A system request for establishing a community exchange group is sent to the current student.

[0013] The system request for establishing a community exchange group refers to a system request for establishing a community exchange group composed of multiple other students on the optimal resource flow path.

[0014] The ID number of a key node student corresponding to the optimal resource flow path is obtained, and all resources collected by the current key node student are obtained, and the resources collected by the current key node student are pushed to the current student based on the system request for establishing a community exchange group.

[0015] Correspondingly, the application also provides a teaching data resource searching and matching processing system, which comprises an initial information acquisition module, a recommendation module, a learning manual generation module and a pushing module.

[0016] The initial information acquisition module is used for verifying the identity information of a current student, and after verification, the input information of the student is obtained.

[0017] The recommendation module is used for searching and matching based on the input information from a preset digital teaching resource library through multi-dimensional hierarchical analysis to obtain recommended information.

[0018] The recommended information includes target resources and a target resource list.

[0019] The learning manual generation module is used for generating a learning task work order based on the input information and the target resources, and obtaining an electronic version resource learning manual by integrating the target resources and the learning task work order. Meanwhile, an optimal resource flow path and target resources corresponding to the optimal resource flow path are obtained, and the target resources corresponding to the optimal resource flow path are pushed to the current student and other students on the optimal resource flow path. A system request for establishing a community exchange group is sent to the current student.

[0020] The system request for establishing the community exchange group refers to a system request for establishing a community exchange group composed of other students on multiple nodes on an optimal resource flow path;

[0021] The pushing module is used for obtaining the ID number of the key node student corresponding to the optimal resource flow path, and obtaining all resources collected by the current key node student, and pushing all resources collected by the current key node student to the current student based on the system request for establishing the community exchange group.

[0022] Compared with the prior art, the embodiments of the present application have at least the following technical advantages:

[0023] It can be known from the above-mentioned teaching data resource retrieval matching processing method and system that, in specific application, the input information of students is collected to provide basic data for subsequent resource matching and personalized recommendation, the students input relevant information through a system interface to provide accurate basis for subsequent matching and recommendation of the system, and by obtaining the input information, it is ensured that the system can provide personalized resource recommendation according to the specific needs of students.

[0024] Further, the optimal resource flow path and the target resource corresponding to the optimal resource flow path are obtained, the target resource corresponding to the optimal resource flow path is pushed to the current student and the students on other nodes on the optimal resource flow path, the input information of the students is analyzed (the resource most related to the needs of the students is found through a matching algorithm), and the optimal resource flow path and the target resource corresponding to the optimal resource flow path are obtained (the target resource corresponding to the optimal resource flow path with high participation of the students is obtained through flow analysis on the student nodes), so that personalized and more interactive learning resource recommendation between students is realized through one target resource and a series of target resource lists, the students are prevented from wasting time to find irrelevant or unsuitable materials, and the learning efficiency is improved and the learning pertinence is enhanced.

[0025] Further, according to the input information of the student and the recommended target resource, a learning task work order is generated and integrated into an electronic version resource learning manual. The electronic version resource learning manual not only integrates the recommended learning resource, but also includes the learning task of the target resource and the resource set composed of the target resource corresponding to the learning task of the plurality of target resources, a learning task list and the like, to help the student to systematically carry out learning. The learning task work order generation is to combine the input information obtained in the foregoing with the recommended target resource, and is converted into a specific learning task or learning plan. The learning task work order can be understood as a guide book for the learning process of the student, which specifies the steps, sequence and specific learning task list of the learning of the target resource. By integrating a plurality of resources into the form of a work manual, the work manual not only contains learning materials, but also can be attached with learning objectives, time arrangement, task decomposition and the like, to ensure that the student completes the learning in an orderly manner. BRIEF DESCRIPTION OF DRAWINGS

[0026] Figure 1 A main flowchart of a kind of teaching data resource retrieval matching processing method;

[0027] Figure 2 A resource flow path text real-time simulation diagram in a kind of teaching data resource retrieval matching processing method;

[0028] Figure 3 A resource flow path simulation schematic diagram in a kind of teaching data resource retrieval matching processing method;

[0029] Figure 4 A schematic diagram of the operation flow of the target resource list updated in a kind of teaching data resource retrieval matching processing method;

[0030] Figure 5 A schematic diagram of the overall architecture of a kind of teaching data resource retrieval matching processing system.

[0031] Label: initial information acquisition module 10, recommendation module 20, learning manual generation module 30, push module 40. DETAILED DESCRIPTION

[0032] The technical solutions of the present application will be described clearly and completely in combination with the drawings. Obviously, the described embodiments are part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor belong to the scope of protection of the present application.

[0033] The present application will be described in further detail below by specific embodiments and in combination with the drawings.

[0034] Embodiment one

[0035] As Figure 1 Embodiment one of the present application provides a teaching data resource retrieval matching processing method, comprising the following operation steps:

[0036] Step S10: verifying the identity information of the current student (or current user), and obtaining the input information of the student after verification;

[0037] Step S20: based on the input information, searching and matching in the preset digital teaching resource library through multi-dimensional hierarchical analysis to obtain recommended information;

[0038] The recommended information includes target resources and a target resource list;

[0039] Step S30: generating a learning task work order based on the input information and the target resources, and integrating the target resources and the learning task work order to obtain an electronic version of the resource learning manual; at the same time, obtaining the optimal resource flow path and the target resources corresponding to the optimal resource flow path, pushing the target resources corresponding to the optimal resource flow path to the current student and other students on the optimal resource flow path; sending a system request for establishing a community exchange group to the current student;

[0040] The system request for establishing a community exchange group refers to a system request for establishing a community exchange group composed of multiple other students on the optimal resource flow path;

[0041] In this way, after the current student or the multiple other students agree, the community exchange group can be directly established for learning communication.

[0042] Step S40: obtaining the ID number of the key node student (the key node student is the node student with the most likes for learning comments and problem solving, or the student with the most accurate system evaluation) corresponding to the optimal resource flow path; obtaining all resources (i.e. learning resources) collected by the current key node student, and pushing all resources collected by the current key node student to the current student based on the system request for establishing a community exchange group (i.e. after the current student accepts the system request for establishing a community exchange group, a community exchange group is established, and then all resources (i.e. learning resources) collected by the current key node student are pushed to the current student through the community exchange group).

[0043] It should be noted that the resource flow path (or resource flow trajectory) describes the propagation path of the target resource in the student group and the change of its timeliness. It can also reflect the change of the target resource selection tendency in different learning stages (or can be understood as: different learning stage information), which is of great significance for optimizing the recommended target resource, and also has an impact on the feedback in step 30 of sending a system request to the current student to establish a community exchange group composed of other students on multiple nodes on the optimal resource flow path.

[0044] The above embodiment of the present application collects the input information of the students to provide basic data for subsequent resource matching and personalized recommendation. The students input relevant information through a system interface (such as a Web form, a mobile application, etc.). The information can be collected by selection, filling, etc. The collected data can be structured data (such as course name, difficulty level) or unstructured data (such as specific problem description). These data constitute a comprehensive understanding of the current needs of the students and provide accurate basis for subsequent matching and recommendation of the system. By obtaining these input information, the system can ensure personalized resource recommendation according to the specific needs of the students;

[0045] Further, the optimal resource flow path and the target resource corresponding to the optimal resource flow path are obtained, and the target resource corresponding to the optimal resource flow path is pushed to the current student and the students on other nodes on the optimal resource flow path. By analyzing the input information of the students (finding the most relevant resources to the needs of the students through a matching algorithm) and obtaining the optimal resource flow path and the target resource corresponding to the optimal resource flow path (obtaining the target resource corresponding to the optimal resource flow path with high student engagement through flow analysis on the student node), a target resource and a series of target resource lists are obtained to realize personalized and more interactive learning resource recommendation among students, avoid wasting time searching for irrelevant or inapplicable materials, and thus improve learning efficiency and enhance learning targeting. The above technical solution fully considers the flow of learning data in the community to provide technical foundation support for more accurate interactive resource recommendation.

[0046] The recommendation result can include a target resource and a list of target resources, the target resource is the optimal recommendation, and the list of target resources provides other possible useful resources, thereby realizing personalized and intelligent learning resource recommendation, avoiding students wasting time searching for irrelevant or unsuitable materials, and improving learning efficiency. Students can quickly obtain learning resources that match their needs, enhance the relevance and effectiveness of learning; further, the system will generate a learning task work order and integrate it into an electronic version of a resource learning manual according to the input information of the student and the recommended target resource. The electronic version of the resource learning manual not only integrates the recommended learning resources (i.e., target resources), but also can include learning tasks of the target resources and a resource set composed of a series of target resources corresponding to the learning tasks of the target resources, a learning task list and the like, to help students systematically carry out learning; the learning task work order generation is to combine the input information obtained in the foregoing with the recommended target resources to convert into specific learning tasks or learning plans. The learning task work order can be understood as a guide book for the student's learning process, which specifies the steps, order and specific learning task list of the learning of the target resources; by integrating multiple resources into the form of a work manual, the work manual not only contains learning materials (i.e., target resources), but also can be attached with learning objectives, time schedules, task decomposition and the like, to ensure that the student completes the learning in an orderly manner.

[0047] At the same time, the target resources corresponding to the optimal resource flow path are pushed to the current student and the students on other nodes on the optimal resource flow path.

[0048] The above embodiments of the present application provide students with comprehensive and systematic learning tools. Students can not only obtain specific learning resources, but also get guided information to help them efficiently and orderly learn, enhance the learning experience of students, and avoid confusion or information overload in the learning process; through automatic recommendation and integration, students can obtain learning materials and community exchanges (such as changeable collaborative data) that meet their needs in a short time, reducing the time cost of resource searching and screening.

[0049] Specifically, in step S20, the input information is searched and matched from a preset digital teaching resource library through multi-dimensional hierarchical analysis to obtain recommendation information, including the following operation steps:

[0050] Step S21: pre-processing the input information to obtain pre-processed input information;

[0051] It should be noted that the above preprocessing operation includes information cleaning operation (information cleaning operation is denoising and word segmentation processing, because the original information input by the student may contain spelling errors, irrelevant words or redundant punctuation marks, and needs to be preprocessed, and the useless information is cleaned by using regular expressions, spelling correction algorithm, etc., and then the student's input text is cut into words by using a word segmentation tool (such as jieba segmentation, NLTK, etc.), word segmentation is the basis of text analysis, and can ensure that the key content in the text can be extracted and processed), information normalization operation (information normalization operation is synonym processing and grammar structure simplification operation, standardize the synonyms input by the student, for example, “electronic product detection and maintenance” and “electronic technology” may point to the same type of resources, which can be unified through a synonym dictionary, and then use natural language processing (NLP) technology to simplify the grammar structure of the input text, so that the text can effectively express the needs of the students, and facilitate subsequent matching), and information classification processing operation (information classification processing operation refers to keyword extraction and category label matching, important words in the student input information are identified through TF-IDF or other keyword extraction methods (such as TextRank algorithm), which facilitates subsequent retrieval and matching, and according to the content input by the student, it is classified into related categories of teaching resources, such as “electronic product detection and maintenance”, “analog electronic technology”, “printed circuit board design”, etc., to ensure that the corresponding field or discipline can be quickly located during retrieval.

[0052] The above embodiment of the present application processes the input information through preprocessing operation, ensures that the preprocessed input information obtained is clear, standardized and can effectively reflect the needs of the students, and at the same time provides an accurate data basis for subsequent retrieval and matching operations.

[0053] Step S22: searching to obtain a plurality of initial screening preselected resources through full-text retrieval processing based on the preprocessed input information;

[0054] It should be noted that the full-text retrieval processing in the above embodiment of the present application refers to generating an input information keyword set after word segmentation processing and stop word removal processing on the preprocessed input information, and then determining a plurality of initial screening preselected resources by calculating the matching degree (the higher the matching degree, the easier it is to be selected as an initial screening preselected resource) of the inverted index of all resource information (resource information refers to the resource document described below) corresponding to the pre-established digital teaching resource library and the input information keyword set.

[0055] The higher the frequency of the index of the keyword in the inverted index of each document corresponding to the preselected resource in the initial screening corresponding to the keyword in the input information keyword set, the higher the matching degree of the resource document and the input information.

[0056] Step S23: searching for recommended information from the digital teaching resource library based on the initial screening of the preselected resource in combination with the learning tendency data of the current student through multi-algorithm fusion retrieval optimization matching processing operation;

[0057] The learning tendency data includes interest preference information (usually refers to the interested learning course category), historical learning record, and learning stage information (the learning stage information such as primary, intermediate, and advanced is obtained through the progress information of the current student's learning course, for example, chapters 1-5 are the primary stage, chapters 6-8 are the intermediate stage, and chapters 9-12 are the advanced stage. The above learning stage information can indicate the student's knowledge mastery, so as to recommend resources according to the student's knowledge mastery in the subsequent resource retrieval and matching process. At the same time, the above learning stage information can be obtained by the system according to the historical learning record or input by the student independently).

[0058] It should be noted that the above embodiment of the present application first cleans and optimizes the data input by the student through various technical means to ensure that the information meets the retrieval requirements in terms of quality. Specifically, on the one hand, the information input by the student may contain spelling errors, irrelevant words, or redundant punctuation marks. These "noise" information will affect the subsequent analysis. Through regular expressions, spelling correction algorithms (such as automatic correction), etc., the input information is ensured to be accurate and not redundant. On the other hand, some words input by the student may express the same meaning, such as "electronic product detection and maintenance" and "mathematics". These words need to be standardized. Through a thesaurus, the system can unify these words to avoid ambiguity. Thirdly, the natural language processing (NLP) technology is used to simplify the grammar structure, so that the grammar of the input text is more straightforward, which is conducive to the subsequent processing. Also, through the keyword extraction method (such as TF-IDF or TextRank algorithm), the keywords in the student's input are extracted. These keywords not only help to understand the student's needs, but also help the system to classify the subsequent retrieval operation, which is classified into related teaching resource categories (such as "electronic product detection and maintenance", "analog electronic technology", "printed circuit board design", etc.), so as to realize more accurate resource positioning. Through denoising, word segmentation, synonym standardization, etc., the input information is ensured to have higher quality and can reflect the actual needs of the student. After standardization, the information becomes more standardized, which is helpful for subsequent retrieval and matching and reduces unnecessary errors or ambiguities. Further, through full-text retrieval, the most matched multiple initial screening preselected resources from the digital teaching resource library are found.

[0059] Specifically, after the input information is processed by word segmentation, stop words (such as irrelevant words such as "of", "is", etc.) are removed, a keyword set is generated, and an inverted index is created for each document in the resource library, that is, each keyword contained in each document has a corresponding index. When the keyword set of the input information is generated, the inverted index is used to find the documents containing the relevant keywords. The matching degree of each document with the keyword set of the student input information is calculated by the system, and the documents with high matching degrees are selected as the initial screening preselected resources. The matching degree is measured by calculating the frequency of the inverted index keywords. The higher the frequency, the higher the matching degree of the document with the input information. Through the inverted index, the documents containing the relevant keywords are quickly located, thereby improving the retrieval speed and accuracy. By calculating the matching degree, the most relevant initial screening preselected resources to the student's needs are selected, thereby improving the relevance and accuracy of the recommendations.

[0060] Further, by combining the student's learning tendency data, the recommendation results are optimized using various algorithms, and finally the recommended information suitable for the student is generated. The learning tendency data includes the student's interest preference information, historical learning records, and learning stage information, which helps to further refine the recommendations and avoid pure text-based matching. Instead, the recommendations are optimized based on the student's personalized learning characteristics. The student's interest preference information, historical learning records, and learning stage information are used to provide personalized background data. For example, a student may have a special interest in a certain subject or have a high level of knowledge in certain areas. By using various recommendation algorithms (such as collaborative filtering, content-based recommendation, deep learning, etc.), not only the student's input information is considered, but also the learning tendency data is optimized, making the recommended results more in line with the student's personalized learning needs.

[0061] Specifically, in step S23, based on the initial screening preselected resources and the current student's learning tendency data, the recommended information is searched and matched from the digital teaching resource library through multi-algorithm fusion retrieval optimization matching processing operations, including the following operation steps:

[0062] Step S231: The learning tendency data is uniformly processed to obtain a first learning tendency data score. A learning tendency matrix L is constructed based on the first learning tendency data scores corresponding to each student. Each element in the learning tendency matrix L represents the first learning tendency data score of the ith student in the jth learning tendency data.

[0063] It should be noted that the uniform processing of the learning tendency data by the embodiments of the present application means that the student's learning tendency data is processed through format unification and quantization to obtain the first learning tendency data score of the student for each learning tendency data, which facilitates subsequent calculation and analysis.

[0064] Step S232: obtaining the access frequency of the initial screening preselected resource ; based on the access frequency An initial screening preselected resource prediction score matrix is established based on the learning tendency matrix L and the access frequency; and the top n initial screening preselected resources are selected as the preferred resources after sorting the elements in the initial screening preselected resource prediction score matrix from high to low according to the scores.

[0065] Each element in the initial screening preselected resource prediction score matrix represents the prediction score of the jth first learning tendency data score of the ith student corresponding to the access frequency of the mth initial screening preselected resource.

[0066] It should be noted that, in the above embodiments of the present application, the initial screening preselected resource prediction score matrix is established based on the access frequency and the learning tendency matrix L, which is a weighted combination of the learning tendency of the student and the access frequency of the resource by means of weighted summation. For example, each element in the matrix L represents the interest intensity of the student in a certain learning field, and the access frequency reflects the general use of the resource in the student group. By combining the two, the possible interest degree of a student in a specific resource can be predicted. For example, if a student has a high learning tendency data score in the field of "electronic product detection and maintenance", and the access frequency of the "electronic product detection and maintenance" resource is also high, the prediction score of the student for this resource will be high, and vice versa.

[0067] The learning tendency data of the student is closely related to the access frequency of each resource in the digital teaching resource library. For example, a student majoring in electronic information engineering technology will usually use the "electronic product detection and maintenance" related resources in the digital teaching resource library. At the same time, his learning habits and historical learning records usually contain "electronic product detection and maintenance" related resource information, so the "electronic product detection and maintenance" resource will be frequently accessed by the student and students of the same major. Therefore, the access frequency of the "electronic product detection and maintenance" resource for students majoring in electronic information engineering technology is higher than that of other resources (such as resources related to economics and marketing statistics for students majoring in marketing). Based on this, the present application establishes an initial screening preselected resource prediction score matrix for each student by means of the learning tendency matrix and the access frequency, so as to score and sort the initial screening preselected resources based on the learning tendency data of each student, and further screen the resource information that is more consistent with the input information of the current student.

[0068] Step S233: obtaining the abstract information of the preferred resources in advance; performing keyword feature extraction on the abstract information of the preferred resources to obtain the keyword feature information; performing feature extraction on the input information and the sequence information of the input information to obtain the sequence feature information of the input information; and performing filtering based on the keyword feature information and the sequence feature information of the input information by calculating the similarity to obtain a plurality of target resources.

[0069] It should be noted that the keyword feature extraction of the preferred resources is performed in the embodiments of the present application, and the subsequent search matching operation is simplified.

[0070] The sequence information of the input information refers to the text sequence of the input information, that is, for example, the current student inputs "how to correctly use the single-chip microcomputer physical structure in the electronic product detection and maintenance to correspond to the single-chip microcomputer instruction application", and after the preprocessing such as word segmentation and stop word removal, the keyword results such as "single-chip microcomputer", "programming", "physical structure", and "instruction" are obtained. Based on the input sequence information of the current input information (that is, the order position information of each keyword), it can be known that the input information of the current student is mainly to learn "single-chip microcomputer instruction application", and if the sequence of the current input information is not used, resources such as "single-chip microcomputer physical structure" or resources related to "instruction control single-chip microcomputer physical structure" will be recognized. Therefore, the sequence information of the input information is extracted in the embodiments of the present application, so that the preferred resources that are more in line with the needs of the current student input search matching corresponding resource information are more accurately searched and matched.

[0071] The sequence feature of the input information in the embodiments of the present application refers to the sequence information of the input text formed by the keywords in the original order. The input sequence feature not only considers the keywords themselves, but also considers their positions and orders in the input text, because the order of the keywords helps to better understand the specific details of the student's demand. After the keyword features of the preferred resources and the sequence features of the input information are obtained, the similarity between the input information and each preferred resource can be evaluated by using a similarity calculation technology (such as cosine similarity, Euclidean distance, etc.). For example, if the input information contains "single-chip microcomputer instruction application", and the keywords of a certain resource include "single-chip microcomputer", "programming", and "instruction", the target resource that is most in line with the student's demand can be selected by calculating the similarity score between the input information and these resources. It should be particularly noted that during feature extraction, not only keyword matching is performed, but also the context of the input information is combined, for example, the input information is "how to correctly use the single-chip microcomputer physical structure in the electronic product detection and maintenance to correspond to the single-chip microcomputer instruction application", and the "single-chip microcomputer instruction application" part needs to be focused on, so that mis-matching of resources in other fields can be avoided, and the resource matching is more accurate.

[0072] Step S234: obtaining a target resource list by optimizing sorting of the target resources in combination with the learning tendency matrix L and the variability collaboration data of the current student, and outputting the target resource list.

[0073] It should be noted that the embodiments of the present application standardize and quantify the learning tendency data of the students, so as to facilitate subsequent analysis and calculation. By constructing the learning tendency matrix L, the system can clearly express the learning interest or tendency of each student in each field or knowledge point, thereby laying a data foundation for personalized recommendation. Through the construction of the learning tendency matrix, the learning interest of each student in each subject or field is accurately represented. Further, based on the learning tendency matrix L of the student and the access frequency of the resource, the interest degree of the student to each initial screening preselected resource is predicted, and the most relevant resource is screened out. Specifically, the access frequency reflects the popularity or popularity of different resources in the overall student group. For example, a certain learning resource (such as “electronic product detection and maintenance”) may have a high access frequency in students majoring in electrical engineering, while students of other majors may use it less. Based on the learning tendency matrix L and the access frequency of the resource, a personalized resource prediction score matrix is generated by weighted summation of the student learning tendency and the resource access frequency, so as to accurately screen out the preselected resource that meets the interest of the student, and optimize the accuracy of resource recommendation. Further, through keyword feature extraction and sequence feature extraction of input information, resource matching is further refined, so that the recommended resource is more in line with the actual needs of the student. Finally, according to the optimized sorting result, the target resource and the target resource list that meet the needs of the student are outputted.

[0074] Specifically, in step S234, the target resources are sorted by optimization in combination with the learning tendency matrix L and the variability collaboration data of the current student, and a target resource list is outputted, including the following operation steps:

[0075] Step S2341: based on the key feature information of the target resources, a target resource feature matrix R is established (each element in the target resource feature matrix R represents the kth key feature information of the mth target resource); according to the learning tendency matrix L and the target resource feature matrix R, the matching degree of the learning tendency matrix L and the target resource feature matrix R (target resource and learning tendency data) is calculated, and the first target resource list is generated by sorting from high to low according to the matching degree (to ensure that the output resource is more in line with the actual needs of the student, and the learning efficiency is improved).

[0076] The calculation method of the matching degree is:

[0077] ;

[0078] It should be noted that the above embodiment of the present application uses matrix multiplication to combine the learning tendency matrix of the student and the resource feature matrix to obtain the matching degree score of each student and all resources (here, all resources refer to the target resources in the target resource feature matrix R), thereby sorting all resources, thereby generating a first target resource list; by matching the learning tendency of the student and the resource content, the generated target resource list will be more in line with the personalized needs of the student, ensuring that the recommended resources match the learning goals of the student, and improving the learning efficiency of the student.

[0079] Step S2342: obtaining the update time of each target resource , based on the update time and the access frequency of the target resource , calculating to obtain the target resource heat; updating the first target resource list based on the target resource heat to obtain a second target resource list (ensuring the timeliness and practicality of the resources);

[0080] The calculation method of the target resource heat is:

[0081] ;

[0082] In the formula, is the access frequency of the mth target resource; is the update time of the mth target resource; and is a weight coefficient;

[0083] It should be noted that the updating of the first target resource list based on the target resource heat in the above embodiment of the present application is to use the target resource heat and the matching degree to perform weighted summation, then obtain the comprehensive score of the target resource heat and the matching degree of the target resource, and then sort according to the comprehensive score from high to low to obtain the second target resource list;

[0084] The above embodiment of the present application reorders the target resources by the access frequency and the update time of the target resources to obtain the second target resource list, ensuring that the output resources not only meet the needs of the student, but also have high timeliness and practicality (because generally, the closer the update time is to the current time, the higher the completeness and credibility of the content of the target resource, because the update of the target resource is updated with the progress of technology and the further completeness of knowledge); in the generated second target resource list, the heat and update of the target resource will be fully considered, ensuring that the recommended resources not only meet the needs of the student, but also provide the most popular learning materials.

[0085] Generally speaking, the learning behavior time matching of individuals and groups is more common in conventional technology, but it has obvious technical limitations; especially without fully considering the content of the target resource itself and the flow and replenishment of learning resources among students, as well as the attention of peer students (peer students are other students in the process of community exchange and resource flow path in addition to the current student) on the resource flow path.

[0086] Step S2343: Obtain changeable collaboration data, update the second target resource list based on the changeable collaboration data to obtain a target resource list (analyze common needs and provide group learning resources);

[0087] Resource flow trajectory (or resource flow path), which describes the propagation path of the target resource in the student group and the change of its timeliness. At the same time, it can also reflect the change of the target resource selection tendency in different learning stages (or can be understood as: different learning stage information), which is of great significance to the optimization of the recommended target resource, and also has an impact on the feedback in step 30 "sending a system request to the current student to establish a community exchange group composed of other students on multiple nodes on the optimal resource flow path".

[0088] The changeable collaboration data represents the changeable target resource (i.e. resource) of the node student corresponding to the different time nodes on the different resource flow paths of all students except the current student; because the changeable collaboration data reflects the group learning event; wherein, the group learning event is, for example, a learning event such as an upcoming exam, a test, a group exchange, a preview, etc.

[0089] In this way, when the group learning event occurs, the corresponding target resource will be frequently consulted, and researchers have found that the target resource will change after being processed by other students, and as the target resource is passed among students (i.e. forums and communities and resource posts), the target resource may be enriched and added by student feedback;

[0090] For example, as Figure 2As shown, assuming that there are 30 students in a class, a student A uploads an electronic product detection and repair learning material to the learning community in the electronic database, and after the group learning event, the analysis process of the resource flow trajectory is as follows: students B and C download the material (i.e., target resource) and share it with D and E. However, student E makes a summary note based on the material and enriches it with some learning annotations and knowledge point questions and shares it with G and H. At the same time, student D only learns after obtaining the target resource and does not improve or change the target resource. As time goes on, student G also improves the summary note and enriches the learning content and answers the knowledge point questions of student E and shares it with other students, and then the above other students update the target resource, assuming that at this time other students no longer add "learning annotations and knowledge point questions", and the subsequent propagation path is not described again; if other students continue to add "learning annotations and knowledge point questions", the resource flow path needs to be tracked continuously.

[0091] In this way, by constructing the propagation network, it is found that student E and student G are important propagation nodes because they generate derivative resources and expand the propagation range.

[0092] Therefore, the changeable collaborative data represent the current student and other students in different resource flow paths corresponding to the changeable target resource, so he has multiple paths corresponding to the changeable target resource (such as Figure 3 As shown, Figure 3 The nodes corresponding to the bold boxes in the figure are key nodes that make annotations, questions or answers to changeable target resources); among them, the resource flow path "A-B-C" and "A" and "A-B" are the most original target resources, but the resource flow path "A-B-C-E" is the most optimal target resource after being enriched and improved once, and the resource flow path "A-B-C-E-G" is the most optimal target resource after being enriched and improved multiple times (2 times in this case); there will be many kinds of resource flow paths corresponding to the changeable collaborative data; therefore, different resource flow paths will produce different results of changeable target resources, so that the update data of the target resource is greatly enriched.

[0093] Research shows that, therefore, the changeable collaborative data corresponding to the target resource not only reflect the heat of the group learning event, but most importantly, it also has the feedback of the resource update and the speciality of the resource flow path; therefore, considering the changeable collaborative data, the target resource recommendation in the target resource list is of great significance. Compared with the traditional system server-side bottom resource update, it supplements the capture of target resource annotations and questions and answers in the student node end data update, and better supplements the readability of learning resources.

[0094] It should be noted that the above embodiment of the present application can find the resource information that the current student needs more urgently by analyzing the changeable collaborative data, and mobilize and adjust the resource recommendation according to the urgent needs of the student; at the same time, the above changeable collaborative data can find the common needs of the student group, and adjust the resource recommendation according to these needs;

[0095] The above embodiment of the present application realizes the accuracy and timeliness of resource matching through multi-dimensional feature fusion and dynamic sorting mechanism, effectively improves the personalized learning experience, and ensures that the student obtains the most suitable learning resources. Through this multi-dimensional feature fusion and dynamic sorting mechanism, not only the accuracy of resource matching is improved, but also the personalized needs of students and the changes of learning environment are fully considered, so that the recommended learning resources are more accurate and practical, and the learning effect and satisfaction of students are further improved.

[0096] At the same time, it needs to be specially pointed out that, as shown in the execution process of step S30, the embodiment mechanism of the present application can also capture the dynamic changes in the learning process in real time, adjust the resource recommendation strategy in time, and even accept the feedback of the number of likes of the student to obtain the optimal resource flow path and the corresponding target resource (i.e. knowledge content), so as to ensure that the student can obtain the most suitable learning support in different learning stages, thereby continuously optimizing and determining the optimal resource flow path and the corresponding optimal peer interaction (for example, if the optimal resource flow path "A-B-C-E-G" corresponds to the target resource, the student "G" (even the student "E") on the optimal resource flow path can be pushed to the current student for offline communication, or an independent communication community can be established by multiple students on the optimal resource flow path "A-B-C-E-G", and the optimal target resource after the optimal resource flow path is enriched and improved can be recommended to the current student and other students on the resource flow path), and the overall learning effect is improved. Through this intelligent resource matching system, students not only can efficiently obtain the required knowledge, but also can cultivate self-learning ability, form good learning habits, and help overall development.

[0097] Specifically, as shown in Figure 4 In step S2343, the second target resource list is updated based on the changeable collaborative data to obtain a target resource list, including the following operation steps:

[0098] Step S23431: obtaining the changeable target resource heat of the gth changeable target resource in the changeable collaborative data ; based on the changeable target resource heat , the changeable target resource heat variance is calculated ;

[0099] The changeable target resource heat The changeable target resource heat The higher the changeable target resource heat is, the more likely the changeable target resource is to be borrowed by the current student.

[0100] It should be noted that the embodiments of the present application measure the attention degree (attention degree refers to reading, commenting or asking questions, etc.) of each node student to the changeable target resource in the resource flow trajectory process, and calculate the heat and heat variance of the changeable target resource, so as to understand the frequency of reading or using the resource by peers, thereby providing a reference for the recommendation of the current student; specifically, the resource with higher heat of the changeable target resource means that more students pay attention to and use the resource, and the content of the resource is also more complete and reliable; the variance of the heat of the changeable target resource reflects the fluctuation of the heat, if the variance is large, it means that the attention of each node student to the resource is greatly different, which may indicate that the resource is suitable for some students, but not suitable for other students, on the contrary, if the variance is small, it means that the popularity of the resource is relatively consistent among the node students; the above embodiments of the present application provide quantitative analysis of the node student to the changeable target resource, which helps the system to understand the tendency and distribution of the peer node student in the selection of learning resources, and provides basic data for subsequent calculation.

[0101] Step S23432: calculating the collaboration similarity based on the changeable collaboration data and each target resource in the second target resource list The collaboration similarity The collaboration similarity

[0102] It should be noted that the embodiments of the present application calculate the similarity between the variability collaborative data and the target resources based on the variability collaborative data and each target resource in the second target resource list of the current student, find out the variability collaborative data close to the current student's demand, and further enhance the accuracy of personalized recommendation. Specifically, by comparing the similarity between the target resources of the current student and the variability collaborative data, the higher the similarity, and the higher the popularity of the variability collaborative data, the more the variability collaborative data meets the demand of the current student for searching and matching resources, and it should be listed in the front end of the target resource list, thereby providing more accurate resource recommendation.

[0103] Step S23433: calculating the second target resource popularity of each target resource in the second target resource list based on the variability target resource popularity , the variability target resource popularity variance , and the collaborative similarity Step S23434: combining the target resource popularity of each target resource in the second target resource list Step S23435: calculating the second target resource popularity of each target resource in the second target resource list ;

[0104] The calculation method of the second target resource popularity is as follows:

[0105]

[0106] In the formula, , represents the variability target resource popularity;

[0107] , represents the variability target resource popularity variance;

[0108] , represents the second target resource popularity of the mth target resource in the second target resource list of the target resources of the current student;

[0109] , represents the second target resource popularity of the qth target resource in the second target resource list of the target resources of the current student;

[0110] , represents the collaborative similarity;

[0111] N represents the number of students (i.e. the number of students corresponding to the variability collaborative data and the current student);

[0112] It should be noted that the above embodiments of the present application calculate the second target resource heat of each target resource based on the variability target resource heat, heat variance and collaboration similarity, optimize the resource recommendation list of the current student according to the learning behavior of the variability collaboration data and its similarity. Specifically, by combining the variability target resource heat, heat variance and collaboration similarity, the predicted heat of each target resource for the current student is obtained through weighted calculation. By comprehensively considering the variability target resource heat, heat variance and similarity, the target resource heat more in line with the needs of students is generated according to the learning needs of the current student and the behavior of similar collaboration data. Through multi-dimensional weighted calculation, the personalization and accuracy of resource recommendation are optimized.

[0113] Step S23434: based on the second target resource heat The final target resource heat is obtained through iterative optimization judgment and reconfirmation ;

[0114] It should be noted that the above embodiments of the present application reconfirm the second target resource heat obtained by preliminary calculation through iterative optimization, so as to ensure that the recommended resources are more accurate and more in line with the needs of students. Specifically, after the second target resource heat is obtained by preliminary calculation, multiple iterations and optimization are performed to further adjust the heat of the resources. Through iterative adjustment, the heat of the resources is dynamically corrected, and the accuracy and personalization level of the recommendation are further improved. By continuously adjusting the heat of the recommended resources, it is ensured that the final output resource list meets the needs of students as much as possible, so that the recommendation system is more intelligent and flexible.

[0115] Step S23435: based on the final target resource heat The target resource list is obtained by sorting from high to low.

[0116] It should be noted that the above embodiments of the present application fully play the advantages of group wisdom and personalized recommendation by using variability collaboration data (including variability target resource heat and heat variance) and each target resource of the student. The student not only obtains recommendation according to his own learning tendency, but also optimizes the recommended content according to similar variability collaboration data. Moreover, through comprehensive calculation of multi-dimensional factors (variability target resource heat, heat variance and collaboration similarity), the one-sidedness of single-dimensional recommendation is avoided. Through iterative optimization mechanism, the heat of the resources is continuously adjusted and corrected, so as to ensure that the recommended result gradually tends to be optimal over time, and the accuracy and adaptability of the recommendation are improved. Finally, through sorting of the final heat, the most relevant and suitable resources are ranked in the front row, helping the student to quickly find the most matched learning resources and improving the learning effect.

[0117] The technical solution described above in this invention is a multi-dimensional user behavior analysis data recommendation method. It fully considers factors such as interest level, attention level, and resource popularity among classmates. It not only accurately captures students' personalized needs but also comprehensively considers key factors such as the popularity of variable target resources and collaboration similarity. This enables the construction of a more efficient, personalized, and interactive electronic information teaching database, thereby enhancing students' learning experience.

[0118] Specifically, in step S23434, based on the second target resource heat index... The final target resource popularity is obtained by reconfirming the judgment through iterative optimization. The operation includes the following steps:

[0119] Step S234341: Initialize the iteration parameters, which include an iteration counter and a maximum iteration threshold; the initial value of the iteration counter is 0.

[0120] Step S234342: Based on the second target resource popularity With the preset ideal temperature The first error of the heat was obtained through calculation. ;

[0121] ;

[0122] In the formula, The ideal level of interest is preset based on factors such as students' historical behavior and interests.

[0123] Step S234343: Determine the first error in the heat. Is it less than or equal to a preset heat error threshold? If so, output the current heat of the second target resource. For the ultimate goal of resource popularity If not, increment the iteration counter by 1 to obtain the current iteration count; determine if the current iteration count is greater than or equal to the maximum iteration count threshold; if so, output the current second target resource heat index. For the ultimate goal of resource popularity If not, then based on the first error of the heat... Combining the current students' learning tendency matrix L with the update time of the target resources The heat adjustment factor is calculated. ;

[0124] The heat adjustment factor The calculation method is as follows:

[0125] L;

[0126] wherein, is a weight coefficient of the adjustment factor;

[0127] Step S234344: updating the second target heat based on the heat adjustment factor to obtain a third target heat ; taking the third target heat as the second target heat and returning to the above operation (i.e., the above step S234342) for reiteration until a final target resource heat is output ;

[0128] The third target heat is calculated in the following manner:

[0129] + ;

[0130] It should be noted that the above embodiment of the present application reconfirms the second target resource heat by means of triple judgment, thereby obtaining the final target resource heat . Specifically, the final target resource heat is determined through the calculation of the heat first error , the number of iterations, and the third target heat . More specifically, the above embodiment of the present application first provides a control mechanism for the subsequent iteration process by initializing the iteration parameter, ensures that the algorithm is executed within the set number of times, and avoids unlimited iteration. Then, the error is calculated to provide a measurement standard for subsequent heat adjustment. If the error is large, it indicates that the current resource heat deviates from the ideal target and needs to be adjusted. Furthermore, the error and the number of iterations are used to determine whether to end the iteration and output the final result. In general, the heat adjustment factor introduced by the learning tendency matrix and the resource update time in the above embodiment of the present application is not only adjusted based on the error, but also considers the personalized learning behavior of students and the new changes of resources, making the heat adjustment more accurate, personalized, and dynamic. The key of the above embodiment of the present application lies in the calculation of the heat first error, the number of iterations, and the heat adjustment factor, which dynamically optimizes the resource heat to make it as close to the ideal heat as possible.

[0131] Embodiment Two

[0132] Based on the same concept of the above method embodiment, the present embodiment also provides a teaching data resource retrieval matching processing system for implementing the above method of the present application. Since the system embodiment solves the problem in a similar principle and method, it at least has all the beneficial effects brought by the technical solutions of the above embodiments, and thus will not be described here in detail.

[0133] As Figure 5As shown, the present application also provides a retrieval matching processing system of teaching data resources, comprising an initial information acquisition module 10, a recommendation module 20, a learning manual generation module 30 and a pushing module 40.

[0134] The initial information acquisition module 10 is used for verifying the identity information of the current student, and after verification, the input information of the student is acquired.

[0135] The recommendation module 20 is used for searching and matching in the preset digital teaching resource library based on the input information through multi-dimensional hierarchical analysis to obtain recommendation information.

[0136] The recommendation information includes target resources and a target resource list.

[0137] The learning manual generation module 30 is used for generating a learning task work order based on the input information and the target resources, and integrating the target resources and the learning task work order to obtain an electronic version resource learning manual. At the same time, the optimal resource flow path and the target resources corresponding to the optimal resource flow path are acquired, and the target resources corresponding to the optimal resource flow path are pushed to the current student and other students on the optimal resource flow path. A system request for establishing a community exchange group is sent to the current student.

[0138] The system request for establishing a community exchange group refers to a system request for establishing a community exchange group composed of other students on multiple nodes on the optimal resource flow path.

[0139] The pushing module 40 is used for acquiring the ID number of the key node student corresponding to the optimal resource flow path, and acquiring all the resources collected by the current key node student. All the resources collected by the current key node student are pushed to the current student based on the system request for establishing a community exchange group.

[0140] In summary, the retrieval matching processing method and system of teaching data resources provided by the present application collect the input information of the student to provide basic data for subsequent resource matching and personalized recommendation. The student inputs relevant information through the system interface to provide accurate basis for subsequent matching and recommendation of the system. By acquiring the input information, the system can provide personalized resource recommendation according to the specific needs of the student.

[0141] Further, the optimal resource flow path and the target resource corresponding to the optimal resource flow path are obtained, and the target resource corresponding to the optimal resource flow path is pushed to the current student and the students on other nodes on the optimal resource flow path; in this way, by analyzing the input information of the student (finding the most relevant resource to the demand of the student through a matching algorithm) and obtaining the optimal resource flow path and the target resource corresponding to the optimal resource flow path (obtaining the target resource corresponding to the optimal resource flow path with high student participation through flow analysis on the student node), the personalized and more interactive learning resource recommendation between students is realized through a target resource and a series of target resource lists, avoiding the waste of time of the student in searching for irrelevant or inapplicable materials, thereby improving the learning efficiency and enhancing the learning pertinence; at the same time, the target resource corresponding to the optimal resource flow path is pushed to the current student and the students on other nodes on the optimal resource flow path; the above technical solution provides a technical basis for realizing more accurate interactive resource recommendation by fully considering the flow of learning data in the community.

[0142] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; a person of ordinary skill in the art can modify the technical solutions described in the above embodiments, or make equivalent replacement for part or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solution deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A search matching processing method of a teaching data resource, characterized by, The method comprises the following steps: Verify the identity information of the current student, and obtain the input information of the student after verification; Based on the input information, search and match in the preset digital teaching resource library through multi-dimensional hierarchical analysis to obtain recommended information, comprising the following steps: Preprocessing the input information to obtain preprocessed input information; Based on the preprocessed input information, search through full-text retrieval processing to obtain a plurality of initial screening preselected resources; Based on the initial screening preselected resources and the learning tendency data of the current student, search and match in the digital teaching resource library through multi-algorithm fusion retrieval optimization matching processing operation to obtain recommended information; The learning tendency data includes interest preference information, historical learning records, and learning stage information; The recommended information includes target resources and a target resource list, including: quantifying the learning tendency data of the student to construct a learning tendency matrix L; based on the learning tendency matrix L of the student and the resource access frequency, predicting the interest degree of the student to each initial screening preselected resource, and screening to obtain target resources; combine the learning tendency matrix L and the variability collaboration data of the current student to obtain a target resource list through optimization sorting, and output; the variability collaboration data represents the variability target resources of the node students corresponding to different time nodes on different resource flow paths of all students except the current student; Based on the input information and the target resources, generate a learning task work order, and integrate the target resources and the learning task work order to obtain an electronic version of the resource learning manual; at the same time, obtain the optimal resource flow path and the target resources corresponding to the optimal resource flow path, and push the target resources corresponding to the optimal resource flow path to the current student and other students on the optimal resource flow path; Send a system request to the current student to establish a community exchange group; The system request for establishing a community exchange group refers to a system request for establishing a community exchange group composed of a plurality of other students on the optimal resource flow path; Obtain the ID number of the key node student corresponding to the optimal resource flow path; obtain all resources collected by the current key node student, and push all resources collected by the current key node student to the current student based on the system request for establishing a community exchange group.

2. The method of claim 1, wherein, The resource flow path represents the propagation path of the target resources in the student group and the change of timeliness.

3. The method of claim 2, wherein the method further comprises: Based on the initial screening preselected resources and the learning tendency data of the current student, search and match in the digital teaching resource library through multi-algorithm fusion retrieval optimization matching processing operation to obtain recommended information, comprising the following steps: Uniformly process the learning tendency data to obtain a first learning tendency data score; based on the first learning tendency data score corresponding to each student, a learning tendency matrix L is constructed; each element in the learning tendency matrix L represents the i-th student at the j-th first learning tendency data score; obtaining the consulting frequency of the initial screening preselected resource ; based on the consulting frequency establishing an initial screening preselected resource prediction score matrix in combination with the learning tendency matrix L; selecting the top n initial screening preselected resources as preferred resources after sorting each element in the initial screening preselected resource prediction score matrix from high to low according to the score The initial screening pre-selected resource prediction score matrix is used to represent the prediction score of the first j learning tendency data score of the i student corresponding to the m initial screening pre-selected resource reading frequency; The abstract information of the preferred resource is obtained in advance, the keyword feature extraction operation is performed on the abstract information of the preferred resource to obtain the key feature information, and the feature extraction operation is performed based on the input information and the sequence information of the input information to obtain the input sequence feature information; The key feature information and the input sequence feature information are used to perform the similarity calculation to obtain a plurality of target resources.

4. The method of claim 3, wherein, The target resource list is obtained by optimizing the sorting of the target resource combined with the learning tendency matrix L and the variability collaboration data of the current student, and is output, including the following operation steps: The target resource feature matrix R is established based on the key feature information of the target resource; The matching degree of the learning tendency matrix L and the target resource feature matrix R is calculated according to the learning tendency matrix L and the target resource feature matrix R, the learning tendency matrix L and the target resource feature matrix R are sorted from high to low according to the matching degree, and a first target resource list is generated; acquiring update times of the target resources , calculating target resource hotness based on the update times and the frequencies of accessing the target resources ​ The first target resource list is updated based on the target resource heat to obtain a second target resource list. The variability collaboration data is obtained, and the second target resource list is updated based on the variability collaboration data to obtain a target resource list.

5. The method of claim 4, wherein, The second target resource list is updated based on the variability collaboration data to obtain a target resource list, including the following operation steps: acquiring a change target resource popularity of a gth change target resource in change collaboration data ; based on the changeable target resource heat computing the changeable target resource heat variance ; a similarity degree of cooperation is calculated between the changeable cooperation data and each of the target resources in the second target resource list ; the similarity degree of cooperation represents a similarity degree of the a-th changeable cooperation data and the m-th target resource of the current student based on the change target resource heat , change target resource heat variance , and the cooperation similarity combining the target resource heat of each target resource in the second target resource list to calculate the second target resource heat of each target resource ; based on the second target resource heat The reconfirmation is performed in a manner of iterative optimization judgment to obtain a final target resource heat ; based on the final target resource heat perform high-to-low ordering to obtain a target resource list.

6. The method of claim 5, wherein, The second target resource heat The calculation method is: ； In the formula, represents the variability target resource heat; denotes the variability target resource hotness variance; a target resource hotness corresponding to the mth target resource in the second target resource list representing target resources of the current student; a target resource heat degree corresponding to the qth target resource in the second target resource list representing target resources of the current student; represents a cooperative similarity; N represents the number of students.

7. The method of claim 6, wherein the method further comprises: The second target resource popularity is determined based on the first target resource popularity The final target resource popularity is obtained by reconfirming in an iterative optimization judgment manner The method comprises the following operation steps: An iteration parameter is initialized, the iteration parameter including an iteration counter and an iteration maximum number threshold value; the iteration number of the iteration counter is initially 0; based on the second target resource popularity with a pre-set ideal popularity a first error of the popularity is calculated ; ； In the formula, T is the ideal temperature; determining whether the first error of the hotness is less than or equal to a preset hotness error threshold value; if yes, outputting the second target resource hotness as the final target resource hotness ;​​ If not, the iteration number of the iteration counter is added by 1 to obtain a current iteration number; it is judged whether the current iteration number is greater than or equal to an iteration maximum number threshold; if yes, the current second target resource heat is output For the final target resource heat If not, the heat first error Combined with the current student's learning tendency matrix L and the update time of the target resource The heat adjustment factor is calculated ; updating the second target resource hotness based on the hotness adjustment factor to obtain a third target resource hotness ; reiterating the above operation with the third target resource heat as a second target resource heat and returning until a final target resource heat is output .

Images