Learning friend recommendation method based on smart education terminal

Abstract

The invention provides a learning friend recommendation method based on an intelligent education terminal, and the method comprises the steps: obtaining learning data of a first student from the intelligent education terminal, so as to determine a student behavior tag of the first student; performing standardization processing and aggregation processing on the student behavior tags to obtain at least one of a learning participation degree parameter corresponding to the first tag, a learning efficiency parameter corresponding to the second tag and a learning type parameter corresponding to the third tag; and based on at least one of the learning participation degree parameter, the learning efficiency parameter and the learning type parameter, friend recommendation is carried out on the first student. According to the method disclosed by the invention, the learning data of the first student is acquired, so that the labels of the first student are generated from multiple aspects such as learning behavior conditions, knowledge mastering degrees and learning styles, and then similarity matching is performed on the labels of different students, so that friend recommendation is performed on the first student according to the similarity matching result; and the objectivity and accuracy of friend recommendation are improved.

Description

Technical Field

[0001] The present disclosure relates to the field of computer technology, and in particular to a learning friend recommendation method based on a smart education terminal. Background Art

[0002] In recent years, with the continuous advancement of computer technology and educational informatization, computer and artificial intelligence technology have gradually been applied to daily educational and teaching activities. More and more intelligent devices are gradually being applied to teaching scenarios. Students can learn through intelligent devices, providing students with more convenient learning services.

[0003] When students use intelligent devices to study, they can usually only manually select their own learning tags, which is subjective. Therefore, when students are recommended friends based on the learning tags, the acceptance rate is not high, and it is not possible to accurately recommend learning partners to students. Therefore, a solution is needed to objectively recommend friends to students. Summary of the invention

[0004] The present invention provides a learning friend recommendation method based on a smart education terminal.

[0005] According to a first aspect of the present disclosure, a method for recommending learning friends based on a smart education terminal is provided, the method comprising: obtaining learning data of a first student from a smart education terminal to determine a student behavior label of the first student, the student behavior label comprising at least one of the following: a first label, a second label and a third label, wherein the first label is used to indicate the learning behavior of the first student, the second label is used to indicate the knowledge mastery of the first student, and the third label is used to indicate the learning style of the first student; standardizing and aggregating the student behavior labels to obtain at least one of a learning engagement parameter corresponding to the first label, a learning effectiveness parameter corresponding to the second label, and a learning type parameter corresponding to the third label; and recommending friends to the first student based on the learning engagement parameter, the learning effectiveness parameter, and at least one of the learning type parameters.

[0006] In some embodiments of the present disclosure, obtaining the learning data of the first student to determine the student behavior label includes at least one of the following: determining the chat data label of the first student based on the chat content of the first student in the learning data; determining the learning understanding label of the first student based on the first student's understanding of the teaching material in the learning data; determining the knowledge point proficiency label of the first student based on the answer data of the first student to the first question corresponding to the same knowledge point in the learning data; determining the global ability evaluation parameters of the comprehensive evaluation of the first student based on the answer data of the first student to questions on different knowledge points in the learning data, and using the global ability evaluation parameters to compare with a preset threshold to determine the learning type label of the first student, wherein the learning type label includes at least one of the following: an advanced challenge label, a balanced development label, and a basic cultivation label.

[0007] In some embodiments of the present disclosure, based on the answer data of the first student for the first question corresponding to the same knowledge point in the learning data, determining the knowledge point proficiency label of the first student includes: determining the ratio between the number of the first student's completions of the first question and the total number of the first question as the first student's knowledge point learning progress parameter, and determining the ratio between the number of the first student's correct answers to the first question and the number of the first question completed as the first student's knowledge point accuracy parameter; performing weighted summation on the learning progress parameter and the knowledge point accuracy parameter to determine the first knowledge point proficiency parameter of the first student; when the modification time after the first student incorrectly answers the first question is less than the preset modification time, and / or, when the number of times the first student correctly answers the first question is greater than the preset correct number threshold, and the standard deviation of the time intervals for the first student to correctly answer the first question is greater than the preset standard deviation threshold, determining that the first student has a suspicious event; when the number of suspicious events of the first student is greater than or equal to the preset suspicious threshold, correcting the first knowledge point proficiency parameter to obtain the knowledge point proficiency label; when the number of suspicious events of the first student is less than the preset suspicious threshold, determining the first knowledge point proficiency parameter as the knowledge point proficiency label.

[0008] In some embodiments of the present disclosure, the first knowledge point proficiency parameter is corrected to obtain a knowledge point proficiency label, including: associating the first knowledge point proficiency parameter with the time when the first student answers the first question to obtain a second knowledge point proficiency parameter with a time label; based on the second knowledge point proficiency parameter, the number of suspicious events corresponding to the time label, a preset suspicious threshold, a preset suspicious event weight, and a preset steepness parameter, combined with the following formula 1, determining a proficiency correction parameter: Formula 1, Among them, FraudScore represents the proficiency correction parameter, Indicates the preset suspicious event weight. Indicates the number of suspicious events corresponding to the time tag. represents the preset suspicious threshold, k represents the preset steepness parameter; Based on the proficiency correction parameter, in combination with the following formula 2, the proficiency parameter of the second knowledge point is corrected to obtain the proficiency parameter of the third knowledge point; Formula 2, Among them, M final represents the proficiency parameter of the third knowledge point, M t represents the proficiency parameter of the second knowledge point, Based on the proficiency parameter of the third knowledge point, the preset difficulty of the first question, the preset memory retention rate and the preset interference parameter, the knowledge point proficiency label is determined in combination with the following formula 3: Formula 3, Among them, P represents the knowledge point proficiency label, Indicates the preset interference parameters, Indicates the preset memory retention rate, Indicates the preset difficulty of the first question of the tth question.

[0009] In some embodiments of the present disclosure, the method further includes: determining a calibration difficulty based on the correct answer rate of the first question, the average answer time of the first question, the total number of questions of the first question, and the actual answer time of the first question in combination with the following formula 4, so as to replace the preset difficulty with the calibration difficulty: Formula 4, in, represents the calibration difficulty of the tth first question, U represents the total number of questions in the first question, Indicates the actual answer time for the first question. represents the average answer time for the first question, Indicates the actual answer time of the first question of the tth question.

[0010] In some embodiments of the present disclosure, based on the answer data of the first student to different knowledge point questions in the learning data, determining the global ability evaluation parameters of the comprehensive evaluation of the first student includes: smoothing and normalizing the answer data to obtain standardized answer data; taking a weighted sum of the standardized basic knowledge point accuracy and the standardized basic knowledge point learning completion in the standardized answer data to determine the basic knowledge point score of the standardized answer data; taking a weighted sum of the standardized key knowledge point accuracy and the standardized key knowledge point learning completion in the standardized answer data to determine the key knowledge point score of the standardized answer data; based on the basic knowledge point score, the key knowledge point score, and the preset angle difference between the basic knowledge point and the key knowledge point, combined with the following formula 5, determining the global ability evaluation parameters, Formula 5, Among them, T represents the global capability evaluation parameter, S BP Indicates the basic knowledge point score, S IW Indicates the score of key knowledge points, Indicates the preset angle difference.

[0011] In some embodiments of the present disclosure, the learning type label of the first student is determined by comparing the global ability evaluation parameter with a preset threshold value, including: when the global ability evaluation parameter is less than the first evaluation preset value, the learning type label is determined to be an advanced challenge label; when the global ability evaluation parameter is less than the second evaluation preset value, and the global ability evaluation parameter is greater than the third evaluation preset value, the learning type label is determined to be a balanced development label; when the global ability evaluation parameter is greater than the fourth evaluation preset value, the learning type label is determined to be a basic cultivation label, wherein the first evaluation preset value is greater than the second evaluation preset value, the second evaluation preset value is greater than the third evaluation preset value, and the third evaluation preset value is greater than the fourth evaluation preset value.

[0012] In some embodiments of the present disclosure, student behavior labels are standardized and aggregated to obtain a learning engagement parameter corresponding to a first label, a learning effectiveness parameter corresponding to a second label, and at least one of a learning type parameter corresponding to a third label, including: dividing the first label and the second label into positive labels and negative labels, and binary encoding the third label to obtain a coded label; performing weighted summation of the positive labels and negative labels corresponding to the first label to obtain a learning engagement parameter; performing weighted summation of the positive labels and negative labels corresponding to the second label to obtain a learning effectiveness parameter; and mapping the coded labels according to preset coding mapping rules to obtain a learning type parameter.

[0013] In some embodiments of the present disclosure, recommending a first student as a friend based on at least one of a learning engagement parameter, a learning effectiveness parameter, and a learning type parameter includes: recommending a second student whose learning engagement parameters, learning effectiveness parameters, and learning type parameters meet the preset recommendation parameter requirements to the first student according to preset parameter requirements of the first student; and / or performing global matching processing on the learning engagement parameters, learning effectiveness parameters, and learning type parameters to determine the label score of the first student, and determining the label similarity between the first students based on the label score of the first student and the preset label weight, so as to recommend the first student as a friend according to the label similarity.

[0014] In some embodiments of the present disclosure, performing global matching processing on the learning engagement parameter, the learning effectiveness parameter, and the learning type parameter to determine the label score of the first student includes: determining the label score of the first student based on the learning engagement parameter, the learning effectiveness parameter, and the learning type parameter in combination with the following formula 6: Formula 6, Among them, Match_Score represents the label score, X_Score represents the learning engagement parameter, Y_Score represents the learning effectiveness parameter, Z_Coeff represents the learning type parameter, and μ1, μ2, and μ3 represent the preset parameter weights.

[0015] In some embodiments of the present disclosure, determining the tag similarity between the first students based on the tag scores and preset tag weights of the first students, and recommending the first students as friends based on the tag similarity includes: determining the tag similarity between the first students based on the tag scores and preset tag weights of the first students in combination with the following formula 7: Formula 7, Among them, Sim(u,v) represents the label similarity between student u and student v, u k represents the student's label score in vector form, v k represents the label score of the student v in vector form, w k Represents the preset tag weight; when the tag similarity is greater than the preset similarity threshold, the first student is recommended as a friend.

[0016] In summary, the present disclosure proposes a learning friend recommendation method based on a smart education terminal, including: obtaining the learning data of the first student from the smart education terminal to determine the student behavior label of the first student, the student behavior label includes at least one of the following: a first label, a second label and a third label, wherein the first label is used to indicate the learning behavior of the first student, the second label is used to indicate the knowledge mastery of the first student, and the third label is used to indicate the learning style of the first student; standardizing and aggregating the student behavior labels to obtain at least one of the learning participation parameter corresponding to the first label, the learning effectiveness parameter corresponding to the second label, and the learning type parameter corresponding to the third label; based on the learning participation parameter, the learning effectiveness parameter, and at least one of the learning type parameters, recommending friends to the first student. The method disclosed in the present disclosure, by obtaining the learning data of the first student, generates the label of the first student from multiple aspects such as learning behavior, knowledge mastery and learning style according to the learning data of the first student, can objectively generate the learning portrait of the first student, and then perform similarity matching on the labels of different students, so as to recommend the first student as a friend according to the similarity matching result, thereby improving the objectivity and accuracy of the friend recommendation.

[0017] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure. Figure 1 A flowchart of a learning friend recommendation method based on a smart education terminal provided in an embodiment of the present disclosure; Figure 2 A flowchart of another learning friend recommendation method based on a smart education terminal provided in an embodiment of the present disclosure; Figure 3 A flowchart of another learning friend recommendation method based on a smart education terminal provided in an embodiment of the present disclosure; Figure 4 A flowchart of another learning friend recommendation method based on a smart education terminal provided in an embodiment of the present disclosure; Figure 5 A flowchart of another learning friend recommendation method based on a smart education terminal provided in an embodiment of the present disclosure; Figure 6 A flowchart of another learning friend recommendation method based on a smart education terminal provided in an embodiment of the present disclosure; Figure 7A flowchart of another learning friend recommendation method based on a smart education terminal provided in an embodiment of the present disclosure; Figure 8 A structural diagram of a learning friend recommendation device based on a smart education terminal provided in an embodiment of the present disclosure; Fig. 9 A schematic diagram of the hardware structure of an electronic device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0019] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0020] In recent years, with the continuous advancement of computer technology and educational informatization, computer and artificial intelligence technology have gradually been applied to daily educational and teaching activities. More and more intelligent devices are gradually being applied to teaching scenarios. Students can learn through intelligent devices, providing students with more convenient learning services.

[0021] When students use intelligent devices to study, they can usually only manually select their own learning tags, which is subjective. Therefore, when students are recommended friends based on the learning tags, the acceptance rate is not high, and it is not possible to accurately recommend learning partners to students. Therefore, a solution is needed to objectively recommend friends to students.

[0022] In order to solve the problems in the related technology, the present invention proposes a learning friend recommendation method based on a smart education terminal, which obtains the learning data of a first student to generate a label for the first student from multiple aspects such as learning behavior, knowledge mastery and learning style, and then performs similarity matching on the labels of different students, so as to recommend friends to the first student based on the similarity matching results, thereby improving the objectivity and accuracy of the friend recommendation.

[0023] A learning friend recommendation method based on a smart education terminal according to an embodiment of the present disclosure is described below with reference to the accompanying drawings.

[0024] Figure 1 The flowchart of a learning friend recommendation method based on a smart education terminal provided by an embodiment of the present disclosure is as follows. Figure 1 As shown, the method includes: Step 101: Acquire learning data of a first student from a smart education terminal to determine a student behavior label of the first student.

[0025] In some embodiments, the smart education terminal can be implemented based on hardware devices (such as tablet computers, smart whiteboards), or based on software applications and installed on ordinary computers or mobile devices. Usually, these smart education terminals can be connected to cloud services to access rich educational resources and perform data synchronization. In other words, the smart education terminal can record the first student's problem-solving, learning, reviewing and other processes while the first student is doing the exercises and learning, thereby storing the first student's learning data.

[0026] In some embodiments, the student behavior label includes at least one of the following: a first label, a second label and a third label, wherein the first label is used to indicate the learning behavior of the first student, the second label is used to indicate the first student's knowledge mastery, and the third label is used to indicate the first student's learning style.

[0027] It should be understood that the student behavior label may include at least one of the first label, the second label and the third label, and may also include other labels other than the first label, the second label and the third label.

[0028] In other words, any label that can be used to indicate the first student's learning progress, learning habits, etc. is used as a student behavior label.

[0029] In some embodiments, the first tag may include one or more sub-tags, the second tag may include one or more sub-tags, and the third tag may include one or more sub-tags.

[0030] In some embodiments, the learning data may include, but is not limited to, the chat content of the first student, the first student's understanding of the teaching material, the first student's answer data, etc.

[0031] In some embodiments, the behavior label of the first student can be determined by performing clustering, statistics, and other processing on the first student's learning data.

[0032] Step 102: Standardize and aggregate the student behavior labels to obtain at least one of a learning engagement parameter corresponding to the first label, a learning effectiveness parameter corresponding to the second label, and a learning type parameter corresponding to the third label.

[0033] In some embodiments, student behavior labels can be divided into categories such as positive labels, negative labels, and learning types, thereby achieving aggregate processing of student behavior data.

[0034] In some embodiments, the aggregated student behavior labels may be normalized to achieve standardized processing of the student behavior labels.

[0035] In some embodiments, the student behavior labels after standardization and aggregation processing are determined as at least one of a learning engagement parameter, a learning effectiveness parameter, and a learning type parameter, wherein the learning engagement parameter corresponds to the first label, the learning effectiveness parameter corresponds to the second label, and the learning type parameter corresponds to the third label.

[0036] Step 103: Recommending a friend to the first student based on at least one of the learning participation parameter, the learning effectiveness parameter, and the learning type parameter.

[0037] In some embodiments, based on the preset parameter requirements of the first student, a second student whose learning participation parameters, learning effectiveness parameters, and learning type parameters meet the preset recommendation parameter requirements can be recommended to the first student.

[0038] For example, the preset parameter requirements for the first student are: the student with the expected learning participation parameter A1, the learning effectiveness parameter A2, and the learning type parameter A3 can be recommended to the second student with the recommended learning participation parameter A1, the learning effectiveness parameter A2, and the learning type parameter A3.

[0039] In some instances, global matching processing is performed on learning engagement parameters, learning effectiveness parameters, and learning type parameters to determine the label scores of the first students, and based on the label scores of the first students and preset label weights, the label similarities between the first students are determined. When the label similarity between the first students is greater than a preset threshold, the first students whose label similarities are greater than the preset threshold can be recommended to each other.

[0040] In some embodiments, when making friend recommendations, the student who is closest to the first student in the preset learning grid and meets the friend recommendation requirements can be recommended to the first student according to the preset learning grid. The preset learning grid is used to be composed of the learning behavior paths of the students, and the coordinate points of the students in the preset learning grid can indicate the learning behaviors of the students (such as learning habits, learning progress, etc.). The closer the distance between the students in the preset learning grid, the more similar the learning behaviors between the students are.

[0041] In summary, the present disclosure proposes a learning friend recommendation method based on a smart education terminal, including: obtaining the learning data of the first student from the smart education terminal to determine the student behavior label of the first student; standardizing and aggregating the student behavior label to obtain the learning participation parameter corresponding to the first label, the learning effectiveness parameter corresponding to the second label, and at least one of the learning type parameters corresponding to the third label; based on the learning participation parameter, the learning effectiveness parameter, and at least one of the learning type parameters, recommending friends to the first student. The method disclosed herein obtains the learning data of the first student to generate a label for the first student from multiple aspects such as learning behavior, knowledge mastery, and learning style, and then performs similarity matching on the labels of different students to recommend friends to the first student based on the similarity matching results, thereby improving the objectivity and accuracy of friend recommendation.

[0042] As a possible implementation, Figure 2 The flowchart of another method for recommending learning friends based on a smart education terminal is shown in FIG. 1 , which further explains step 101 based on the above embodiment and includes the following steps: Step 201: Determine the chat data label of the first student based on the chat content of the first student in the learning data.

[0043] In some embodiments, the first student's online learning time label can be determined based on whether the first student's chat content in the learning data belongs to a preset knowledge base and the length of time the first student spends answering questions; in other words, the length of time corresponding to the first student's chat content that belongs to the preset knowledge base can be added to the length of time a student spends answering questions, and the sum result can be determined as the first student's online learning time label.

[0044] In some embodiments, the number of times students initiate learning help requests in the learning data may be counted, and the counted number may be determined as a help request number label for the first student.

[0045] In some embodiments, the number of times students solve learning help requests in the learning data can be counted, and the counted number can be determined as the effective solution number label of the first student.

[0046] In some embodiments, the response efficiency label of the first student can be determined based on the student's response time when receiving a request for learning assistance in the learning data, and the total time between initiating the learning assistance and resolving the learning assistance.

[0047] Among them, the above-mentioned number of help requests label, number of effective solutions label, response efficiency label and online learning time label can all belong to chat data labels, among which the chat data label can belong to the first label.

[0048] Step 202: Determine a learning comprehension label of the first student based on the first student's understanding of the teaching material in the learning data.

[0049] In some embodiments, the answer accuracy label of the first student can be determined based on the answer data of the first student to the same question in the learning data.

[0050] In some embodiments, the learning progress label of the first student may be determined based on the first student's learning progress of the teaching material content in the learning data.

[0051] In some embodiments, the first student's answering time label may be determined based on the time it takes the first student to answer the same question for the first time in the learning data.

[0052] In some embodiments, the knowledge point proficiency label of the first student may be determined based on the answer data of the first student to the first question corresponding to the same knowledge point in the learning data.

[0053] Among them, the answer accuracy label, the learning progress label, and the answering time label can all belong to the learning comprehension label, among which the learning comprehension label can belong to the second label.

[0054] Step 203: Determine the knowledge point proficiency label of the first student based on the answer data of the first student to the first question corresponding to the same knowledge point in the learning data.

[0055] In some embodiments, the knowledge point proficiency tag may also belong to the second tag.

[0056] Specifically, in some embodiments, the knowledge point accuracy parameter and the knowledge point learning progress parameter of the first student can be determined according to the answer data of the first question corresponding to the same knowledge point of the first student (for example, the number of correct answers of the first student, the total number of questions answered by the first student, etc.); and the learning progress parameter and the knowledge point accuracy parameter are parameter fused to determine the first knowledge point proficiency parameter of the first student, and it is judged whether the first student has plagiarized or cheated, so as to determine the knowledge point proficiency label according to the judgment result, thereby improving the accuracy of label determination, wherein the judgment rule can be as follows: When the modification time after the first student incorrectly answers the first question is less than the preset modification time, and / or when the number of times the first student correctly answers the first question is greater than a preset correct number threshold, and the standard deviation of the time intervals when the first student correctly answers the first question is greater than a preset standard deviation threshold, it is determined that the first student has a suspicious event; When the number of suspicious events of the first student is greater than or equal to a preset suspicious threshold, the first knowledge point proficiency parameter is modified to obtain a knowledge point proficiency label; When the number of suspicious events of the first student is less than a preset suspicious threshold, the first knowledge point proficiency parameter is determined as a knowledge point proficiency label.

[0057] Step 204: Based on the answer data of the first student to questions on different knowledge points in the learning data, determine the global ability evaluation parameters of the comprehensive evaluation of the first student, and use the global ability evaluation parameters to compare with the preset threshold to determine the learning type label of the first student.

[0058] In some embodiments, the learning type label includes at least one of the following: an advanced challenge label, a balanced development label, and a basic training label. Among them, the learning type label may belong to the third label.

[0059] In some embodiments, the learning type label of the first student may be determined based on the first student's answers to questions of different difficulty levels in the answer data.

[0060] For example, if the first student only answers questions with a relatively high difficulty and has a high accuracy rate (greater than a preset threshold), the learning type label of the first student can be determined as an advanced challenge label.

[0061] Specifically, in some embodiments, when the global capability evaluation parameter is less than the first evaluation preset value, the learning type label is determined to be an advanced challenge type label; When the global capability evaluation parameter is less than the second evaluation preset value, and the global capability evaluation parameter is greater than the third evaluation preset value, determining the learning type label as a balanced development label; When the global capability evaluation parameter is greater than the fourth evaluation preset value, the learning type label is determined as the basic cultivation label, wherein the first evaluation preset value is greater than the second evaluation preset value, the second evaluation preset value is greater than the third evaluation preset value, and the third evaluation preset value is greater than the fourth evaluation preset value.

[0062] In summary, the present disclosure proposes a learning friend recommendation method based on a smart education terminal, including: determining the chat data label of the first student based on the chat content of the first student in the learning data; determining the learning understanding label of the first student based on the first student's understanding of the teaching material in the learning data; determining the knowledge point proficiency label of the first student based on the answer data of the first student to the first question corresponding to the same knowledge point in the learning data; determining the global ability evaluation parameters of the comprehensive evaluation of the first student based on the answer data of the first student to different knowledge point questions in the learning data, and using the global ability evaluation parameters to compare with the preset threshold to determine the learning type label of the first student, wherein the learning type label includes at least one of the following: an advanced challenge label, a balanced development label, and a basic cultivation label. The method disclosed in the present disclosure determines the student's behavior labels from multiple angles through different data in the learning data, which can improve the diversity of student labels and lay the foundation for accurate friend recommendation.

[0063] As a possible implementation, Figure 3 The flowchart of another method for recommending learning friends based on a smart education terminal is shown. On the basis of the above embodiment, the knowledge point proficiency label of the first student is further determined based on the answer data of the first question corresponding to the same knowledge point in the learning data, including the following steps: Step 301: Determine the ratio between the number of first questions completed by the first student and the total number of first questions as the first student's knowledge point learning progress parameter, and determine the ratio between the number of correct answers to the first question and the number of first questions completed by the first student as the first student's knowledge point accuracy parameter.

[0064] In some embodiments, the knowledge point learning progress parameter may be determined by the following formula:

[0065] Among them, Q represents the knowledge point learning progress parameter, q k represents the total number of questions in the first question, q done Indicates the number of completed questions for the first question.

[0066] Step 302: Perform a weighted summation of the learning progress parameter and the knowledge point accuracy parameter to determine the first knowledge point proficiency parameter of the first student.

[0067] In some embodiments, the first knowledge point proficiency parameter may be determined by the following formula:

[0068] Among them, C coverage represents the proficiency parameter of the first knowledge point, P represents the learning progress parameter, Indicates the preset weight.

[0069] Step 303: When the modification time after the first student incorrectly answers the first question is less than the preset modification time, and / or when the number of times the first student correctly answers the first question is greater than a preset correct number threshold, and the standard deviation of the time intervals when the first student correctly answers the first question is greater than a preset standard deviation threshold, it is determined that a suspicious event has occurred with the first student.

[0070] In some embodiments, by defining suspicious events, it is determined whether the first student has cheated, plagiarized, or other abnormal behaviors when studying or practicing questions, so as to determine whether the proficiency parameter of the first knowledge point needs to be corrected.

[0071] Step 304: When the number of suspicious events of the first student is greater than or equal to the preset suspicious threshold, the first knowledge point proficiency parameter is modified to obtain a knowledge point proficiency label.

[0072] In some embodiments, when the number of suspicious events of the first student is greater than or equal to a preset suspicious threshold, it is determined that the first student has abnormal behavior. At this time, the first knowledge point proficiency parameter is corrected, and the knowledge point proficiency label is determined according to the corrected first knowledge point proficiency parameter. Figure 4 The embodiment shown.

[0073] Step 305: When the number of suspicious events of the first student is less than a preset suspicious threshold, the first knowledge point proficiency parameter is determined as a knowledge point proficiency label.

[0074] In some embodiments, when the number of suspicious events of the first student is less than a preset suspicious threshold, it is determined that the first student does not have abnormal behavior. At this time, the first knowledge point proficiency parameter can be determined as a knowledge point proficiency label.

[0075] In summary, according to a learning friend recommendation method based on a smart education terminal proposed in the present invention, by determining the number of suspicious events of the first student, the knowledge point proficiency label is determined in different ways, thereby improving the accuracy of the knowledge point proficiency label and laying the foundation for accurate friend recommendation.

[0076] As a possible implementation, Figure 4 The flowchart of another method for recommending learning friends based on a smart education terminal is shown in FIG. 3 , which further explains step 304 based on the above embodiment and includes the following steps: Step 401 , associating the first knowledge point proficiency parameter with the time when the first student answers the first question, to obtain a second knowledge point proficiency parameter with a time tag.

[0077] In some embodiments, when the number of suspicious events of the first student is greater than or equal to a preset suspicious threshold, it indicates that the first student has studied the first knowledge point for a long time. Therefore, in order to determine whether the first student will forget the first knowledge point over time, it is necessary to associate the proficiency parameter of the first knowledge point with the time when the first student answers the first question.

[0078] For example, Represents the proficiency parameter of the first knowledge point of the first student at time t.

[0079] Step 402, based on the proficiency parameter of the second knowledge point, the number of suspicious events corresponding to the time tag, the preset suspicious threshold, the preset suspicious event weight, and the preset steepness parameter, determine the proficiency correction parameter.

[0080] In some embodiments, the proficiency correction parameter may be determined based on the proficiency parameter of the second knowledge point, the number of suspicious events corresponding to the time tag, a preset suspicious threshold, a preset suspicious event weight, and a preset steepness parameter in combination with the following formula 1: Formula 1, Among them, FraudScore represents the proficiency correction parameter, Indicates the preset suspicious event weight. Indicates the number of suspicious events corresponding to the time tag. represents a preset suspicious threshold, and k represents a preset steepness parameter.

[0081] Step 403: based on the proficiency correction parameter, correct the second knowledge point proficiency parameter to obtain a third knowledge point proficiency parameter.

[0082] In some embodiments, the second knowledge point proficiency parameter may be corrected based on the proficiency correction parameter in combination with the following formula 2 to obtain a third knowledge point proficiency parameter: Formula 2, Among them, Mfinal represents the proficiency parameter of the third knowledge point, and Mt represents the proficiency parameter of the second knowledge point.

[0083] Step 404, determining a knowledge point proficiency label based on the third knowledge point proficiency parameter, the preset difficulty of the first question, the preset memory retention rate and the preset interference parameter.

[0084] In some examples, the knowledge point proficiency label may be determined based on the third knowledge point proficiency parameter, the preset difficulty of the first question, the preset memory retention rate, and the preset interference parameter, in combination with the following formula 3: Formula 3, Among them, P represents the knowledge point proficiency label, Indicates the preset interference parameters, Indicates the preset memory retention rate, Indicates the preset difficulty of the first question of the tth question.

[0085] In some embodiments, the calibration difficulty may be determined based on the correct answer rate of the first question, the average answer time of the first question, the total number of questions of the first question, and the actual answer time of the first question in combination with the following formula 4, so as to replace the preset difficulty with the calibration difficulty, thereby dynamically determining the difficulty of the first question by using the actual question-taking data of the first student, so as to improve the accuracy and objectivity of determining the knowledge point proficiency label: Formula 4, in, represents the calibration difficulty of the tth first question, U represents the total number of questions in the first question, Indicates the actual answer time for the first question. represents the average answer time for the first question, Indicates the actual answer time of the first question of the tth question.

[0086] As a possible implementation, Figure 5 The flowchart of another method for recommending learning friends based on a smart education terminal is shown. On the basis of the above embodiment, the global ability evaluation parameter for determining the comprehensive evaluation of the first student based on the answer data of the first student to different knowledge points in the learning data is further explained, including the following steps: Step 501, smoothing and normalizing the answer data to obtain standardized answer data.

[0087] In some embodiments, the answer data may be smoothed by introducing Bayesian smoothing, but the present invention is not limited thereto. Any processing method that can avoid the existence of extreme values ​​in the answer data may be the above-mentioned smoothing process.

[0088] In some embodiments, the smoothed answer data may be mapped to the interval [0, 1] to achieve normalization, and the smoothed and normalized answer data may be determined as standardized answer data.

[0089] In some embodiments, since the difficulty levels of the knowledge points answered by the first student are different, the standardized answer data can be divided into standardized basic knowledge point data and standardized key knowledge point data, wherein the standardized basic knowledge point data can include: standardized basic knowledge point accuracy rate and standardized basic knowledge point learning completion rate; the standardized key knowledge point data can include: standardized key knowledge point accuracy rate and standardized key knowledge point learning completion rate.

[0090] Step 502 , weighted sum of the standardized basic knowledge point accuracy and the standardized basic knowledge point learning completion in the standardized answer data to determine the basic knowledge point score of the standardized answer data.

[0091] In some embodiments, the standardized basic knowledge point accuracy rate may be determined based on the basic knowledge point questions answered by the first student and the basic knowledge point questions answered correctly by the first student.

[0092] Step 503 , weighted sum of the standardized key knowledge point accuracy and the standardized key knowledge point learning completion in the standardized answer data to determine the key knowledge point score of the standardized answer data.

[0093] In some embodiments, the standardized key knowledge point accuracy rate may be determined based on the key knowledge point questions answered by the first student and the key knowledge point questions answered correctly by the first student.

[0094] Step 504, determining global capability evaluation parameters based on the basic knowledge point scores, the key knowledge point scores, and the preset angle differences between the basic knowledge points and the key knowledge points.

[0095] In some embodiments, the global capability evaluation parameter may be determined based on the basic knowledge point score, the key knowledge point score, and the preset angle difference between the basic knowledge point and the key knowledge point, in combination with the following formula 5: Formula 5, Among them, T represents the global capability evaluation parameter, S BP Indicates the basic knowledge point score, S IW Indicates the score of key knowledge points, Indicates the preset angle difference.

[0096] As a possible implementation, Figure 6 The flowchart of another method for recommending learning friends based on a smart education terminal is shown in FIG. 1 , which further explains step 102 based on the above embodiment and includes the following steps: Step 601: divide the first label and the second label into a positive label and a negative label, and perform binary encoding on the third label to obtain an encoded label.

[0097] In some embodiments, the first label and the second label may be divided into positive labels and negative labels as shown in Table 1, and the third label may be binary encoded:

[0098] Table 1 Among them, the classification indicators shown in the table are encoded using one-hot encoding, that is, the Z1 balanced development type is taken as 1, and the basic cultivation type and advanced challenge type are both taken as 0.

[0099] In some embodiments, the positive labels and negative labels in the table may also be standardized by a normalization formula to facilitate the determination of the learning engagement parameter.

[0100] Step 602: Perform a weighted summation on the positive label and the negative label corresponding to the first label to obtain a learning participation parameter.

[0101] In some embodiments, the positive label and the negative label corresponding to the first label may be weighted and summed by the following formula to obtain the learning engagement parameter:

[0102] Among them, X_Score represents the learning participation parameter, X norm represents the positive label and negative label corresponding to the first label, and α represents the preset weight.

[0103] Step 603: Perform weighted summation on the positive label and the negative label corresponding to the second label to obtain a learning effectiveness parameter.

[0104] In some embodiments, the positive label and the negative label corresponding to the second label may be weighted and summed using the following formula to obtain the learning performance parameter:

[0105] Among them, Y_Score represents the learning effectiveness parameter, Y norm represents the positive label and negative label corresponding to the second label, and β represents the preset weight.

[0106] Step 604: Map the coding labels according to preset coding mapping rules to obtain learning type parameters.

[0107] In some embodiments, the encoding labels may be mapped according to the following preset encoding mapping rules to obtain learning type parameters: , Among them, Z_Coeff represents the learning type parameter.

[0108] As a possible implementation, Figure 7 The flowchart of another method for recommending learning friends based on a smart education terminal is shown. On the basis of the above embodiment, the method for recommending friends to the first student according to the tag similarity is further explained, and includes the following steps: Step 701, based on the learning participation parameter, the learning effectiveness parameter and the learning type parameter, combined with the following formula 6, determine the label score of the first student.

[0109] In some embodiments, the label score of the first student may be determined based on the learning participation parameter, the learning effectiveness parameter, and the learning type parameter in combination with the following formula 6: Formula 6, Among them, Match_Score represents the label score, X_Score represents the learning engagement parameter, Y_Score represents the learning effectiveness parameter, Z_Coeff represents the learning type parameter, and μ1, μ2, and μ3 represent the preset parameter weights.

[0110] Step 702, based on the tag scores of the first students and the preset tag weights, combined with the following formula 7, determine the tag similarity between the first students.

[0111] In some embodiments, the label scores of the first students may be vectorized to obtain label scores in vector form, and then the label similarity between the first students may be determined based on the label scores in vector form and preset label weights in combination with the following formula 7: Formula 7, Among them, Sim(u,v) represents the label similarity between student u and student v, u k represents the student's label score in vector form, v k represents the label score of the student v in vector form, w k Indicates the preset label weight.

[0112] In some embodiments, the preset tag weights may also be dynamically updated according to the following formula to improve the accuracy of tag similarity determination:

[0113] in, represents the updated preset label weight, AcceptRatek is the preset learning rate n It represents the acceptance rate of the first user for the nth type of label.

[0114] Step 703: When the tag similarity is greater than a preset similarity threshold, a friend recommendation is performed for the first student.

[0115] In some embodiments, when the tag similarity is greater than a preset similarity threshold, it indicates that the learning behaviors of the first students are relatively similar. At this time, the first students whose tag similarity is greater than the preset similarity threshold are recommended to each other.

[0116] In summary, a learning friend recommendation method based on a smart education terminal is disclosed, which generates a label for a first student from multiple aspects such as learning behavior, knowledge mastery and learning style, and then performs similarity matching on labels of different students, so as to recommend friends to the first student based on the similarity matching results, thereby improving the objectivity and accuracy of friend recommendations.

[0117] Corresponding to the above-mentioned learning friend recommendation method based on a smart education terminal, the present invention also proposes a learning friend recommendation device based on a smart education terminal. Since the device embodiment of the present invention corresponds to the above-mentioned method embodiment, the details not disclosed in the device embodiment can be referred to the above-mentioned method embodiment, and will not be repeated in the present invention.

[0118] Figure 8 A structural diagram of a learning friend recommendation device based on a smart education terminal provided in an embodiment of the present disclosure is shown in FIG. Figure 8 As shown, the device comprises: The determining unit 810 is used to obtain the learning data of the first student from the smart education terminal to determine the student behavior label of the first student, where the student behavior label includes at least one of the following: a first label, a second label, and a third label, wherein the first label is used to indicate the learning behavior of the first student, the second label is used to indicate the knowledge mastery of the first student, and the third label is used to indicate the learning style of the first student; A processing unit 820 is used to perform standardization and aggregation processing on the student behavior label to obtain at least one of a learning participation parameter corresponding to the first label, a learning effectiveness parameter corresponding to the second label, and a learning type parameter corresponding to the third label; The recommendation unit 830 is used to recommend a friend to the first student based on at least one of the learning participation parameter, the learning effectiveness parameter, and the learning type parameter.

[0119] In some embodiments of the present disclosure, the determination unit 810 is further used to: determine the chat data label of the first student based on the chat content of the first student in the learning data; determine the learning understanding label of the first student based on the degree of understanding of the teaching material by the first student in the learning data; determine the knowledge point proficiency label of the first student based on the answer data of the first student to the first question corresponding to the same knowledge point in the learning data; determine the global ability evaluation parameters of the comprehensive evaluation of the first student based on the answer data of the first student to questions on different knowledge points in the learning data, so as to utilize the global ability evaluation parameters to compare with a preset threshold value to determine the learning type label of the first student, wherein the learning type label includes at least one of the following: an advanced challenge label, a balanced development label, and a basic cultivation label.

[0120] In some embodiments of the present disclosure, the determination unit 810 is further used to: determine the ratio between the number of first questions completed by the first student and the total number of first questions as the knowledge point learning progress parameter of the first student, and determine the ratio between the number of correct answers to the first question by the first student and the number of completed first questions as the knowledge point accuracy parameter of the first student; perform weighted summation on the learning progress parameter and the knowledge point accuracy parameter to determine the first knowledge point proficiency parameter of the first student; when the modification time after the first student incorrectly answers the first question is less than the preset modification time, and / or when the number of times the first student correctly answers the first question is greater than the preset correct number threshold, and the standard deviation of the time intervals for the first student to correctly answer the first question is greater than the preset standard deviation threshold, determine that the first student has a suspicious event; when the number of suspicious events of the first student is greater than or equal to the preset suspicious threshold, perform correction processing on the first knowledge point proficiency parameter to obtain a knowledge point proficiency label; when the number of suspicious events of the first student is less than the preset suspicious threshold, determine the first knowledge point proficiency parameter as the knowledge point proficiency label.

[0121] In some embodiments of the present disclosure, the determining unit 810 is further configured to: modify the first knowledge point proficiency parameter to obtain the knowledge point proficiency label, including: Associating the first knowledge point proficiency parameter with the time when the first student answers the first question to obtain a second knowledge point proficiency parameter with a time tag; Based on the proficiency parameter of the second knowledge point, the number of suspicious events corresponding to the time tag, the preset suspicious threshold, the preset suspicious event weight, and the preset steepness parameter, the proficiency correction parameter is determined in combination with the following formula 1: Formula 1, Among them, FraudScore represents the proficiency correction parameter, Indicates the preset suspicious event weight. Indicates the number of suspicious events corresponding to the time tag. represents the preset suspicious threshold, k represents the preset steepness parameter; Based on the proficiency correction parameter, in combination with the following formula 2, the proficiency parameter of the second knowledge point is corrected to obtain the proficiency parameter of the third knowledge point; Formula 2, Among them, Mfinal represents the proficiency parameter of the third knowledge point, Mt represents the proficiency parameter of the second knowledge point, Based on the proficiency parameter of the third knowledge point, the preset difficulty of the first question, the preset memory retention rate and the preset interference parameter, the knowledge point proficiency label is determined in combination with the following formula 3: Formula 3, Among them, P represents the knowledge point proficiency label, Indicates the preset interference parameters, Indicates the preset memory retention rate, Indicates the preset difficulty of the first question of the tth question.

[0122] In some embodiments of the present disclosure, the determining unit 810 is further configured to determine the calibration difficulty based on the correct answer rate of the first question, the average answer time of the first question, the total number of questions of the first question, and the actual answer time of the first question, in combination with the following formula 4, so as to replace the preset difficulty with the calibration difficulty: Formula 4, in, represents the calibration difficulty of the tth first question, U represents the total number of questions in the first question, Indicates the actual answer time for the first question. represents the average answer time for the first question, Indicates the actual answer time of the first question of the tth question.

[0123] In some embodiments of the present disclosure, the determining unit 810 is further configured to: perform smoothing and normalization processing on the answer data to obtain standardized answer data; The accuracy rate of the standardized basic knowledge points and the learning completion degree of the standardized basic knowledge points in the standardized answer data are weighted and summed to determine the basic knowledge point score of the standardized answer data; The accuracy rate of standardized key knowledge points and the degree of learning completion of standardized key knowledge points in the standardized answer data are weighted and summed to determine the key knowledge point score of the standardized answer data; Based on the scores of basic knowledge points, key knowledge points, and the preset angle difference between basic knowledge points and key knowledge points, combined with the following formula 5, the global ability evaluation parameters are determined: Formula 5, Among them, T represents the global capability evaluation parameter, S BP Indicates the basic knowledge point score, S IW Indicates the score of key knowledge points, Indicates the preset angle difference.

[0124] In some embodiments of the present disclosure, the determining unit 810 is further configured to: when the global capability evaluation parameter is less than the first evaluation preset value, determine that the learning type label is an advanced challenge label; When the global capability evaluation parameter is less than the second evaluation preset value, and the global capability evaluation parameter is greater than the third evaluation preset value, determining the learning type label as a balanced development label; When the global capability evaluation parameter is greater than the fourth evaluation preset value, the learning type label is determined as the basic cultivation label, wherein the first evaluation preset value is greater than the second evaluation preset value, the second evaluation preset value is greater than the third evaluation preset value, and the third evaluation preset value is greater than the fourth evaluation preset value.

[0125] In some embodiments of the present disclosure, the processing unit 820 is further used to: divide the first label and the second label into positive labels and negative labels, and binary encode the third label to obtain a coded label; perform weighted summation of the positive label and the negative label corresponding to the first label to obtain a learning engagement parameter; perform weighted summation of the positive label and the negative label corresponding to the second label to obtain a learning effectiveness parameter; map the coded labels according to preset coding mapping rules to obtain a learning type parameter.

[0126] In some embodiments of the present disclosure, the recommendation unit 830 is further configured to: recommend, according to the preset parameter requirements of the first student, a second student whose learning participation parameter, learning effectiveness parameter, and learning type parameter meet the preset recommendation parameter requirements to the first student; and / or The learning participation parameters, learning effectiveness parameters and learning type parameters are globally matched to determine the label score of the first student, and based on the label score of the first student and the preset label weight, the label similarity between the first students is determined to recommend the first student as a friend based on the label similarity.

[0127] In some embodiments of the present disclosure, the recommendation unit 830 is further configured to determine the label score of the first student based on the learning participation parameter, the learning effectiveness parameter, and the learning type parameter in combination with the following formula 6: Formula 6, Among them, Match_Score represents the label score, X_Score represents the learning engagement parameter, Y_Score represents the learning effectiveness parameter, Z_Coeff represents the learning type parameter, and μ1, μ2, and μ3 represent the preset parameter weights.

[0128] In some embodiments of the present disclosure, the recommendation unit 830 is further configured to: determine the tag similarity between the first students based on the tag score and the preset tag weight of the first student in combination with the following formula 7: Formula 7, Among them, Sim(u,v) represents the label similarity between student u and student v, u k represents the student's label score in vector form, v k represents the label score of the student v in vector form, w kRepresents the preset tag weight; when the tag similarity is greater than the preset similarity threshold, the first student is recommended as a friend.

[0129] It should be noted that the above explanation of the method embodiment is also applicable to the device of this embodiment, and the principle is the same, which is not limited in this embodiment.

[0130] Based on the above Figures 1 to 7 The method shown in the figure, accordingly, this embodiment also provides a computer program product, including a computer program, which implements the above-mentioned Figures 1 to 5 The method shown.

[0131] Based on the above Figures 1 to 7 The method shown in the embodiment also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the above-mentioned Figures 1 to 5 The method shown.

[0132] Based on this understanding, the technical solution of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, USB flash drive, mobile hard disk, etc.), including a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods of various implementation scenarios of the present application.

[0133] like Fig. 9 The figure shows a hardware structure diagram of an electronic device of the present invention, including: at least one processor 901; and, A memory 902 that is communicatively connected to at least one of the processors 901; wherein: The memory 902 stores instructions that can be executed by at least one of the processors, and the instructions are executed by at least one of the processors so that at least one of the processors can execute the learning friend recommendation method based on the smart education terminal as described above.

[0134] Fig. 9 A processor 901 is taken as an example.

[0135] The electronic device may further include: an input device 903 and a display device 904 .

[0136] The processor 901, the memory 902, the input device 903 and the display device 904 may be connected via a bus or other means, and the figure takes the connection via a bus as an example.

[0137] The memory 902 is a non-volatile computer-readable storage medium that can be used to store non-volatile software programs, non-volatile computer executable programs and modules, such as program instructions / modules corresponding to the review content generation method in the embodiment of the present application, for example, Figure 1 , Figure 2 as well as Figure 3 The processor 901 executes various functional applications and data processing by running the non-volatile software programs, instructions and modules stored in the memory 902, that is, a learning friend recommendation method based on a smart education terminal in the above embodiment is implemented.

[0138] The memory 902 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created according to the use of the review content generation method, etc. In addition, the memory 902 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage devices. In some embodiments, the memory 902 may optionally include a memory remotely arranged relative to the processor 901, and these remote memories may be connected to the device for executing the review content generation method via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0139] The input device 903 can receive user clicks and generate signal inputs related to user settings and function controls of the review content generation method. The display device 904 can include display devices such as a display screen.

[0140] The one or more modules are stored in the memory 902, and when executed by the one or more processors 901, a learning friend recommendation method based on a smart education terminal in any of the above method embodiments is executed.

[0141] Optionally, the above-mentioned physical device may also include a user interface, a network interface, a camera, a radio frequency (RF) circuit, a sensor, an audio circuit, a WI-FI module, etc. The user interface may include a display, an input unit such as a keyboard, etc., and the optional user interface may also include a USB interface, a card reader interface, etc. The network interface may optionally include a standard wired interface, a wireless interface (such as a WI-FI interface), etc.

[0142] Those skilled in the art will appreciate that the above-mentioned physical device structure provided in this embodiment does not constitute a limitation on the physical device, and may include more or fewer components, or a combination of certain components, or different arrangements of components.

[0143] The storage medium may also include an operating system and a network communication module. The operating system is a program that manages the hardware and software resources of the above-mentioned physical device, and supports the operation of the information processing program and other software and / or programs. The network communication module is used to realize the communication between the components inside the storage medium, and the communication with other hardware and software in the information processing physical device.

[0144] Through the description of the above implementation methods, those skilled in the art can clearly understand that the present application can be implemented by means of software plus a necessary general hardware platform, or by hardware.

[0145] It should be noted that, in this article, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0146] The above is only a specific implementation of the present application, so that those skilled in the art can understand or implement the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments described herein, but will conform to the widest scope consistent with the principles and novel features applied for herein.

Claims

1. A learning friend recommendation method based on a smart education terminal, characterized in that: The method comprises: Acquire learning data of a first student from the smart education terminal to determine a student behavior label of the first student, wherein the student behavior label includes at least one of the following: a first label, a second label, and a third label, wherein the first label is used to indicate the learning behavior of the first student, the second label is used to indicate the knowledge mastery of the first student, and the third label is used to indicate the learning style of the first student; Performing standardization and aggregation processing on the student behavior labels to obtain at least one of a learning engagement parameter corresponding to the first label, a learning effectiveness parameter corresponding to the second label, and a learning type parameter corresponding to the third label; Based on at least one of the learning participation parameter, the learning effectiveness parameter, and the learning type parameter, a friend recommendation is made to the first student.

2. The method according to claim 1, characterized in that The acquiring the learning data of the first student to determine the student behavior label includes at least one of the following: Determining a chat data label of the first student based on the chat content of the first student in the learning data; Determining a learning comprehension label of the first student based on the first student's comprehension level of the teaching material in the learning data; Determining a knowledge point proficiency label of the first student based on answer data of the first student to a first question corresponding to the same knowledge point in the learning data; Based on the answer data of the first student to questions on different knowledge points in the learning data, a global ability evaluation parameter of the comprehensive evaluation of the first student is determined, and the learning type label of the first student is determined by comparing the global ability evaluation parameter with a preset threshold value, The learning type label includes at least one of the following: an advanced challenge label, a balanced development label, and a basic training label.

3. The method according to claim 2, characterized in that The determining of the knowledge point proficiency label of the first student based on the answer data of the first student to the first question corresponding to the same knowledge point in the learning data comprises: Determine the ratio between the number of first questions completed by the first student and the total number of the first questions as the knowledge point learning progress parameter of the first student, and determine the ratio between the number of correct answers to the first question by the first student and the number of first questions completed as the knowledge point accuracy parameter of the first student; Taking a weighted sum of the learning progress parameter and the knowledge point accuracy parameter to determine a first knowledge point proficiency parameter of the first student; When the modification time after the first student incorrectly answers the first question is less than the preset modification time, and / or when the number of times the first student correctly answers the first question is greater than a preset correct number threshold, and the standard deviation of the time intervals when the first student correctly answers the first question is greater than a preset standard deviation threshold, it is determined that the first student has a suspicious event; When the number of suspicious events of the first student is greater than or equal to a preset suspicious threshold, the first knowledge point proficiency parameter is modified to obtain the knowledge point proficiency label; When the number of suspicious events of the first student is less than the preset suspicious threshold, the first knowledge point proficiency parameter is determined as the knowledge point proficiency label.

4. The method according to claim 3, characterized in that The modifying process of the first knowledge point proficiency parameter to obtain the knowledge point proficiency label includes: Associating the first knowledge point proficiency parameter with the time when the first student answers the first question to obtain a second knowledge point proficiency parameter with a time tag; Based on the proficiency parameter of the second knowledge point, the number of suspicious events corresponding to the time tag, the preset suspicious threshold, the preset suspicious event weight, and the preset steepness parameter, the proficiency correction parameter is determined in combination with the following formula 1: Formula 1, Among them, FraudScore represents the proficiency correction parameter, Indicates the preset suspicious event weight. Indicates the number of suspicious events corresponding to the time tag, represents the preset suspicious threshold, k represents a preset steepness parameter; Based on the proficiency correction parameter, in combination with the following formula 2, the second knowledge point proficiency parameter is corrected to obtain a third knowledge point proficiency parameter; Formula 2, Among them, M final represents the proficiency parameter of the third knowledge point, M t represents the proficiency parameter of the second knowledge point, Based on the third knowledge point proficiency parameter, the preset difficulty of the first question, the preset memory retention rate and the preset interference parameter, the knowledge point proficiency label is determined in combination with the following formula 3: Formula 3, Wherein, P represents the proficiency label of the knowledge point, represents the preset interference parameter, represents the preset memory retention rate, Indicates the preset difficulty of the first question of the tth question.

5. The method according to claim 4, characterized in that The method further comprises: Based on the correct answer rate of the first question, the average answer time of the first question, the total number of the first question, and the actual answer time of the first question, the calibration difficulty is determined in combination with the following formula 4, so as to replace the preset difficulty with the calibration difficulty: Formula 4, in, represents the calibration difficulty of the tth first question, U represents the total number of questions in the first question, Indicates the actual answer time for the first question, represents the average answer time for the first question, Indicates the actual answer time of the first question of the tth question.

6. The method according to claim 2, characterized in that Determining the global ability evaluation parameters of the comprehensive evaluation of the first student based on the answer data of the first student to questions on different knowledge points in the learning data includes: Performing smoothing and normalization processing on the answer data to obtain standardized answer data, the standardized answer data; Taking a weighted sum of the accuracy rate of standardized basic knowledge points and the degree of learning completion of standardized basic knowledge points in the standardized answer data, to determine a basic knowledge point score of the standardized answer data; The standardized key knowledge point accuracy and the standardized key knowledge point learning completion in the standardized answer data are weighted and summed to determine the key knowledge point score of the standardized answer data; Based on the basic knowledge point score, the key knowledge point score, and the preset angle difference between the basic knowledge point and the key knowledge point, combined with the following formula 5, the global ability evaluation parameter is determined: Formula 5, Wherein, T represents the global capability evaluation parameter, S BP represents the score of the basic knowledge point, S IW represents the score of the key knowledge point, Indicates the preset angle difference.

7. The method according to claim 2, characterized in that: The determining the learning type label of the first student by comparing the global ability evaluation parameter with a preset threshold value comprises: When the global capability evaluation parameter is less than a first evaluation preset value, determining that the learning type label is the advanced challenge label; When the global capability evaluation parameter is less than the second evaluation preset value, and the global capability evaluation parameter is greater than the third evaluation preset value, determining that the learning type label is the balanced development label; When the global capability evaluation parameter is greater than the fourth evaluation preset value, the learning type label is determined to be the basic cultivation type label, wherein the first evaluation preset value is greater than the second evaluation preset value, the second evaluation preset value is greater than the third evaluation preset value, and the third evaluation preset value is greater than the fourth evaluation preset value.

8. The method according to claim 1, characterized in that The step of performing standardization and aggregation processing on the student behavior labels to obtain at least one of a learning engagement parameter corresponding to the first label, a learning effectiveness parameter corresponding to the second label, and a learning type parameter corresponding to the third label includes: Dividing the first label and the second label into a positive label and a negative label, and binary encoding the third label to obtain an encoded label; Performing a weighted summation on the positive label and the negative label corresponding to the first label to obtain the learning participation parameter; Performing a weighted summation on the positive label and the negative label corresponding to the second label to obtain the learning effectiveness parameter; According to a preset encoding mapping rule, the encoding label is mapped to obtain the learning type parameter.

9. The method according to claim 1, characterized in that: The recommending a friend to the first student based on at least one of the learning participation parameter, the learning effectiveness parameter, and the learning type parameter includes: According to the preset parameter requirements of the first student, recommending a second student whose learning participation parameter, learning effectiveness parameter, and learning type parameter meet the preset recommendation parameter requirements to the first student; and / or The learning engagement parameter, the learning effectiveness parameter and the learning type parameter are globally matched to determine the label score of the first student, and based on the label score of the first student and the preset label weight, the label similarity between the first students is determined to recommend the first student as a friend based on the label similarity.

10. The method according to claim 9, characterized in that The performing global matching processing on the learning participation parameter, the learning effectiveness parameter, and the learning type parameter to determine the label score of the first student includes: Based on the learning participation parameter, the learning effectiveness parameter, and the learning type parameter, in combination with the following formula 6, the label score of the first student is determined: Formula 6, Among them, the Match_Score represents the label score, X_Score represents the learning participation parameter, Y_Score represents the learning effectiveness parameter, Z_Coeff represents the learning type parameter, and μ1, μ2 and μ3 represent preset parameter weights.

11. The method according to claim 9, characterized in that The determining the tag similarity between the first students based on the tag score and the preset tag weight of the first students, and recommending the first students as friends according to the tag similarity includes: Based on the label scores and preset label weights of the first students, the label similarity between the first students is determined in combination with the following formula 7: Formula 7, Among them, Sim(u,v) represents the label similarity between student u and student v, u k represents the student's label score in vector form, v k represents the label score of the student v in vector form, w k represents the preset tag weight; When the tag similarity is greater than a preset similarity threshold, the first student is recommended as a friend.

Images