CN116259004A - Student learning state detection method and system applied to online education - Google Patents

Abstract

The present invention provides a method and system for detecting a student's learning state applied to online education, wherein the method includes: when the online education class starts, obtaining the first state information of the teaching teacher of the online education class; based on the first state information and preset Abnormal learning status detection template generation library of the student learning status abnormality detection template; obtain the second status information of the students in the online education classroom; based on the abnormal learning status detection template of the students, according to the second status information, the student learning status of the students Learning state anomaly detection. The method and system for detecting students' learning status applied to online education of the present invention greatly improves the applicability of abnormal detection of students' learning status and overcomes limitations. Secondly, there is no need for online classroom teachers to pay attention to students' learning status , which improves the convenience, which is beneficial to the teaching teachers to devote their energy to teaching, and improves the teaching quality to a certain extent.

Description

A method and system for detecting students' learning status applied to online education

technical field

The invention relates to the technical field of computer data processing, in particular to a method and system for detecting students' learning status applied to online education.

Background technique

Some social personnel will attend lectures and study in the form of online classrooms. Secondly, some schools that are not convenient for offline teaching will also arrange for teachers and students to take classes and attend classes in the form of online classes.

However, online classes are mostly implemented in the mode of multi-person video. In order to ensure the quality of students' lectures, teachers need to pay attention to the students' learning status. At this time, they need to check the video screens of each student one by one to determine whether the students are listening carefully, which is cumbersome. , It is cumbersome and also takes up part of the teacher's energy that should be used for teaching.

The prior art also provides a solution to this problem. For example, the Chinese patent document publication number is CN115546861A, an online classroom concentration recognition method, system, device and medium, in which student expressions are recognized based on student videos, and student expressions are determined based on student expressions. The student's concentration can replace the teacher's attention to the student's learning status.

However, in the real class, the timing and method of judging whether the students are focused or not depends on the teacher's class status; for example, when the teacher is teaching, the teacher reminds the students to take a break for 5 minutes, and it is reasonable for the students to take any inattentive behavior at this time Yes, if it is still based on the student video to determine whether the student is focused, it will cause misjudgment; another example: the teacher gives interactive follow-up prompts, and at this time to determine whether the student is focused should determine whether the student is follow-up. Secondly, there may be situations where expressions cannot be recognized; for example, when the teacher reminds students to complete the exercises in the textbook, students will lower their heads/bend their heads to perform, and may not be able to recognize expressions.

Therefore, the solutions given above have limitations in specific applications, that is, low applicability, which needs to be solved urgently.

Contents of the invention

One of the purposes of the present invention is to provide a method and system for detecting students' learning status applied to online education, which greatly improves the applicability of abnormal detection of students' learning status and overcomes limitations. The teaching teachers pay attention to the learning status of the students, which improves the convenience, which is beneficial to the teaching teachers to devote their energy to teaching, and improves the teaching quality to a certain extent.

A method for detecting a student's learning status applied to online education provided by an embodiment of the present invention includes:

When the online education class starts, obtain the first status information of the teaching teacher of the online education class;

Based on the first state information and the preset student learning state abnormality detection template generation library, generate a student learning state abnormality detection template;

Acquiring the second status information of the students attending the online education class;

Based on the abnormality detection template of the student's learning state, and according to the second state information, the abnormality detection of the student's learning state is performed on the student attending the class.

Preferably, the first state information of the teaching teacher of the online education classroom is acquired, including:

Pushing the preset teacher state selection table to the first mobile terminal of the teaching teacher;

Obtain the teaching teacher status selected by the teaching teacher from the teaching teacher status selection table;

determining first status information based on the status of the teaching teacher;

and / or,

Obtaining the first speech information of the teaching teacher through the first mobile terminal;

determining first status information based on the first speaking information;

and / or,

Obtaining the class image of the teaching teacher through the first mobile terminal;

extracting the first action information of the teaching teacher from the class image;

Based on the first action information, first state information is determined.

Preferably, based on the first state information and the preset student learning state abnormality detection template generation library, generate a student learning state abnormality detection template, including:

Analyzing the number of information types of the first state information;

When the number of information types is unique, determine the student learning state abnormality detection rule corresponding to the first state information from the student learning state abnormality detection template generation library;

Based on the student learning status abnormality detection rule, generate a student learning status abnormality detection template;

When the number of information types is not unique, analyze the state information feature set of the first state information based on a preset feature analysis template;

Constructing a first feature description vector of the first state information based on the state information feature set;

Extract multiple groups of one-to-one corresponding second feature description vectors and student learning state abnormality detection rule sets from the student learning state abnormality detection template generation library;

calculating a vector similarity between the first feature description vector and any of the second feature description vectors;

An abnormal student learning state detection template is generated based on the student learning state abnormality detection rule set corresponding to the second feature description vector with the largest vector similarity.

Preferably, obtaining the second status information of the students attending the online education classroom includes:

Obtaining the second speech information of the student attending the class through the second mobile terminal of the student attending the class;

determining second status information based on the second speaking information;

and / or,

Obtaining the lecture image of the student attending the lecture through the second mobile terminal;

extracting second action information of the student attending the class from the class attending image;

Based on the second action information, second state information is determined.

Preferably, the student learning status detection method applied to online education also includes:

Obtain the attendance record of the student attending the class;

Based on the class attendance records, classifying the students attending classes into key students and non-key students;

Adaptive allocation of detection resources for the key students and the non-key students to detect the abnormality of student learning status.

Preferably, based on the class attendance records, classifying the students attending classes into key students includes:

Extracting the abnormal learning status records of the students who attended the lectures from the lectures;

Based on the abnormal learning status record, determine the abnormal learning status frequency of the students attending the class;

When the abnormal frequency of the learning state is greater than or equal to the preset abnormal frequency threshold of the learning state, the corresponding student attending the class will be regarded as a key student;

and / or,

Obtain the prediction basis for the upper limit of the student's serious state of listening to the class;

Input the upper limit prediction basis of the serious state of listening to a class into the preset upper limit prediction model of the serious state of listening to a class to determine the upper limit of the serious state of listening to a class;

Extracting the first continuous lecture situation within the latest preset time range of the student attending the lecture from the lecture attendance record;

Based on the first continuous lecture situation, determine whether the student has reached the upper limit of the seriousness of the lecture;

When yes, the corresponding student attending the class will be regarded as the key student.

Preferably, the basis for predicting the upper limit of the student's listening seriousness state is obtained, including:

Acquiring the recording time of the abnormal learning status record;

Based on the recording time, the abnormal learning state is recorded on a preset time axis;

Retrieving abnormal records of the target learning state that meet the preset first retrieval condition from the time axis;

Extract the second continuous lecture situation within the time range before the recording time of the abnormal record of the target learning state from the lecture record, and use it as the basis for predicting the upper limit of the state of seriousness in class;

Wherein, the first retrieval condition includes: the total number of abnormal records of the learning status whose abnormal record type is the preset type within a preset time distance before and after the abnormal learning status record on the time axis is greater than or equal to the preset number threshold;

and / or,

Obtain the student information of the student attending the class;

generating a template based on preset retrieval conditions, and generating a second retrieval condition according to the student information;

Other upper limits of serious state of listening to lectures satisfying the second search condition are retrieved from the preset upper limit collection library of serious state of listening to lectures, and used as a basis for predicting the upper limit of serious state of listening to lectures.

Preferably, the adaptive allocation of detection resources for the abnormal detection of student learning status to the key students and the non-key students includes:

Count the first student total number of the key students and the second student total number of the non-key students respectively;

calculating a numerical ratio of the first total number of students to the second total number of students;

determining a detection resource allocation strategy corresponding to the number ratio from a preset detection resource allocation strategy library;

Based on the detection resource allocation strategy, the detection resources are allocated to the key students and the non-key students respectively.

A student learning status detection system applied to online education provided by an embodiment of the present invention is characterized in that it includes:

The teaching teacher state information acquisition module is used to obtain the first state information of the teaching teacher of the online education class when the online education class starts;

A student learning state abnormality detection template generation module is used to generate a student learning state abnormality detection template based on the first state information and a preset student learning state abnormality detection template generation library;

The state information acquisition module of students attending classes is used to obtain the second state information of students attending classes in the online education classroom;

The abnormality detection module of the student's learning state is configured to detect the abnormality of the student's learning state for the students attending the class based on the abnormality detection template of the student's learning state and according to the second state information.

Preferably, the teaching teacher state information acquisition module obtains the first state information of the teaching teacher in the online education classroom, and performs the following operations:

Pushing the preset teacher state selection table to the first mobile terminal of the teaching teacher;

Obtain the teaching teacher status selected by the teaching teacher from the teaching teacher status selection table;

determining first status information based on the status of the teaching teacher;

and / or,

Obtaining the first speech information of the teaching teacher through the first mobile terminal;

determining first status information based on the first speech information;

and / or,

Obtaining the class image of the teaching teacher through the first mobile terminal;

extracting the first action information of the teaching teacher from the class image;

Based on the first action information, first state information is determined.

Additional features and advantages of the invention will be set forth in the description which follows, and in part will be apparent from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention may be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.

The technical solutions of the present invention will be described in further detail below with reference to the accompanying drawings and embodiments.

Description of drawings

The accompanying drawings are used to provide a further understanding of the present invention, and constitute a part of the description, and are used together with the embodiments of the present invention to explain the present invention, and do not constitute a limitation to the present invention. In the attached picture:

1 is a schematic diagram of a method for detecting a student's learning status applied to online education in an embodiment of the present invention;

Fig. 2 is a schematic diagram of a student learning state detection system applied to online education in an embodiment of the present invention.

Detailed ways

The preferred embodiments of the present invention will be described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described here are only used to illustrate and explain the present invention, and are not intended to limit the present invention.

An embodiment of the present invention provides a method for detecting a student's learning status applied to online education, as shown in Figure 1, including:

Step S1: When the online education class starts, obtain the first status information of the teaching teacher of the online education class;

Step S2: Based on the first state information and the preset student learning state abnormality detection template generation library, generate a student learning state abnormality detection template;

Step S3: Acquiring the second state information of the students attending the online education classroom;

Step S4: Based on the abnormality detection template of the student's learning state, and according to the second state information, detect the abnormality of the student's learning state for the student attending the class.

The working principle and beneficial effects of the above-mentioned technical scheme are:

The first status information includes: in lecture, prompting students to take a break, actions made in class, prompting students to complete textbook exercises, and the like. The second state information includes: facial expressions and actions made during lectures, and the like. The abnormal detection template of the students' learning status is generated by the first status information of the teaching teacher; for example, when the teacher is giving a lecture, the abnormal detection template of the students' learning status is to detect whether the students' expressions and movements are inattentive, etc.; The learning status abnormality detection template is not to do any detection; another example: the teacher makes a follow-up action of clapping, and the student learning status abnormality detection template is to detect whether the student follows the clapping; second example: the teacher reminds the students to complete the textbook exercises, and the students' learning status is abnormal The detection template is to detect whether the students have completed the exercise with their heads down or buried in their heads.

In the specific application, it continuously obtains the status of the teaching teacher, and determines the timing and means of judging whether the students are attentive in class based on the status of the teaching teacher.

This application has greatly improved the applicability of the abnormal detection of students' learning status and overcomes the limitations. Secondly, it does not require the teaching teachers in the online classroom to pay attention to the students' learning status, which improves the convenience and is beneficial for the teaching teachers to devote their energy to In teaching, the quality of teaching has been improved to a certain extent.

In one embodiment, obtaining the first status information of the teaching teacher of the online education classroom includes:

Pushing the preset teacher state selection table to the first mobile terminal of the teaching teacher;

Obtain the teaching teacher status selected by the teaching teacher from the teaching teacher status selection table;

determining first status information based on the status of the teaching teacher;

and / or,

Obtaining the first speech information of the teaching teacher through the first mobile terminal;

determining first status information based on the first speech information;

and / or,

Obtaining the class image of the teaching teacher through the first mobile terminal;

extracting the first action information of the teaching teacher from the class image;

Based on the first action information, first state information is determined.

The working principle and beneficial effects of the above-mentioned technical scheme are:

"And/or" means that there are three ways to determine the first status information of the teaching teacher: the first one is to push the teacher's status selection table for the teacher to choose; Options such as actions and prompting students to complete textbook exercises. The second type is determined based on the teacher's speech; for example, if the teacher says "take a break for 5 minutes", the first status information is to remind students to take a break, and if the teacher says "continue to class", then the first status information is mid-class. The third type is determined based on the teacher's actions; for example, if the teacher performs a yoga demonstration action, the first status information is to prompt the students to perform yoga follow-up.

The embodiment of the present invention introduces three ways to determine the first state information of the teaching teacher, which improves the applicability of the system.

In one embodiment, based on the first state information and the preset student learning state abnormality detection template generation library, generating a student learning state abnormality detection template includes:

Analyzing the number of information types of the first state information;

When the number of information types is unique, determine the student learning state abnormality detection rule corresponding to the first state information from the student learning state abnormality detection template generation library;

Based on the student learning status abnormality detection rule, generate a student learning status abnormality detection template;

When the number of information types is not unique, analyze the state information feature set of the first state information based on a preset feature analysis template;

Constructing a first feature description vector of the first state information based on the state information feature set;

Extract multiple groups of one-to-one corresponding second feature description vectors and student learning state abnormality detection rule sets from the student learning state abnormality detection template generation library;

calculating a vector similarity between the first feature description vector and any of the second feature description vectors;

An abnormal student learning state detection template is generated based on the student learning state abnormality detection rule set corresponding to the second feature description vector with the largest vector similarity.

The working principle and beneficial effects of the above-mentioned technical scheme are:

Generally, the number of information types of the first state information is unique, at this time, determine the student learning state abnormality detection rule corresponding to the first state information from the student learning state abnormality detection template generation library, and generate the student learning state abnormality detection template; for example : The teacher only reminds the students of the rest state, and the abnormal detection rule of the student’s learning status should be to generate a template without any detection; another example: the teacher is only in the state of lecturing, and the abnormal detection rule of the student’s learning status should be to detect the student’s expression and movement If there is inattention, generate templates.

However, there are also situations where the number of information types is not unique; for example: the teacher is teaching dance and makes a leg-raising movement, but verbally says "don't follow me, let me demonstrate", at this time the teacher has two states. At this time, based on the feature analysis template, the state information feature set of the first state information is analyzed; for example, the state information feature set is extracted, including the teacher's speech, including action demonstration instructions, the teacher also performs actions, and the teacher is teaching dance. Based on the state information feature set, the first feature description vector of the first state information is constructed. The greater the vector similarity between the first feature description vector and the second feature description vector, the more suitable the corresponding student learning state abnormality detection rule set is to execute; for example: When teaching dance, the teacher’s speech should prevail, and the abnormal detection rule set of the student’s learning status corresponding to the second feature description vector of the maximum vector similarity includes detecting whether the student has watched the demonstration, detecting whether the student has not learned to do the action, and so on. An anomaly detection template for students' learning status is generated based on the anomaly detection rule set of students' learning status.

In the embodiment of the present invention, considering the unique and non-unique states of the teaching teacher, the abnormality detection templates of the students' learning states are respectively generated, which further improves the applicability of the system.

In one embodiment, obtaining the second state information of the students attending the online education classroom includes:

Obtaining the second speech information of the student attending the class through the second mobile terminal of the student attending the class;

determining second status information based on the second speaking information;

and / or,

Obtaining the lecture image of the student attending the lecture through the second mobile terminal;

extracting second action information of the student attending the class from the class attending image;

Based on the second action information, second state information is determined.

The working principle and beneficial effects of the above-mentioned technical scheme are:

"And/or" means that there are two ways to determine the second status information of the students attending the class: the first one is determined based on the student's speech; for example, the student reads out the content of the text, indicating that the student's status is reading the text, etc. The second is determined based on the student's actions; for example, if the student makes a leg-raising action, it means that the student's state is leg-raising, etc.; another example: if the student makes an expression of laughing, the student's state is laughing.

The embodiment of the present invention introduces two ways to determine the second state information of the students attending the class, which improves the applicability of the system.

In one embodiment, the student learning state detection method applied to online education further includes:

Obtain the attendance record of the student attending the class;

Based on the class attendance records, classifying the students attending classes into key students and non-key students;

Adaptive allocation of detection resources for the key students and the non-key students to detect the abnormality of student learning status.

The working principle and beneficial effects of the above-mentioned technical scheme are:

Normally, the performance of the students attending the lectures varies. Some students are prone to inattention, while others are not. If every student is tested all the time, it may cause unreasonable allocation of testing resources. Therefore, based on the listening records of the students attending the class, the students attending the class are divided into key students and non-key students, and then the adaptive allocation of detection resources is carried out.

In the embodiment of the present invention, students attending a class are divided into key points and non-key points, and then the adaptive allocation of detection resources for abnormal detection of students' learning status is performed, which improves the rationality of the allocation of detection resources.

In one embodiment, based on the class attendance records, classifying the students attending classes into key students includes:

Extracting the abnormal learning status records of the students who attended the lectures from the lectures;

Based on the abnormal learning status record, determine the abnormal learning status frequency of the students attending the class;

When the abnormal frequency of the learning state is greater than or equal to the preset abnormal frequency threshold of the learning state, the corresponding student attending the class will be regarded as a key student;

and / or,

Obtain the prediction basis for the upper limit of the student's serious state of listening to the class;

Input the upper limit prediction basis of the serious state of listening to a class into the preset upper limit prediction model of the serious state of listening to a class to determine the upper limit of the serious state of listening to a class;

Extracting the first continuous lecture situation within the latest preset time range of the student attending the lecture from the lecture attendance record;

Based on the first continuous lecture situation, determine whether the student has reached the upper limit of the seriousness of the lecture;

When yes, the corresponding student attending the class will be regarded as the key student.

The working principle and beneficial effects of the above-mentioned technical scheme are:

"And/or" means that there are two ways to divide the students attending the lectures into key students: the first one is to determine the abnormal frequency of the learning status of the students attending the lectures based on the lecture records. The possibility of concentration is high, and they are regarded as key students; when the students who attend the lectures have abnormal learning status records such as inattentive expressions in the lectures and not bowing/putting their heads down to complete the exercises, they will be recorded in the lecture records. The second is to determine whether the student has reached the upper limit of the serious state of listening to the class based on the class listening records. If it is, it means that the student is more likely to be inattentive in the learning state after that, as a key student; determine the upper limit of the serious state of listening to the class, and predict the upper limit of the serious state of listening to the class According to the input to the upper limit prediction model of the serious state of listening to lectures, the upper limit prediction model of the serious state of listening to lectures is an artificial intelligence model that uses a large number of artificial intelligence to train the neural network model to converge based on the logical records of the upper limit prediction of the serious state of listening to lectures , can replace the manual prediction of the upper limit of the serious state of listening to the class based on the prediction basis of the upper limit of the serious state of listening to the class; the logical record can be, for example: the upper limit of the serious state of listening to the class is predicted based on the fact that the student's history of continuous listening to the class reaches 30, 33, etc. Minutes after the frequent occurrence of unfocused expressions , determine the upper limit of the serious state of listening to the class is the upper limit of the serious state of listening to the class is the total duration of the class is 30 minutes; the first continuous listening situation includes: the length of listening to the class, the type of listening (dance class, math class, etc.), the type of follow-up action and the duration of follow-up, etc.

The embodiment of the present invention introduces two ways to classify students who attend lectures into key students, which improves the comprehensiveness and rationality of the division of key students. Especially the second method, when students attend lectures continuously for a long time, there will be an upper limit of the state of seriousness in lectures, based on The continuous lecture status of students can be used to judge whether they have reached the upper limit of serious state of lectures. If so, as key students, the accuracy and applicability of the key student classification are improved.

In one embodiment, the basis for predicting the upper limit of the student's seriousness in class is obtained, including:

Acquiring the recording time of the abnormal learning status record;

Based on the recording time, the abnormal learning state is recorded on a preset time axis;

Retrieving abnormal records of the target learning state that meet the preset first retrieval condition from the time axis;

Extract the second continuous lecture situation within the time range before the recording time of the abnormal record of the target learning state from the lecture record, and use it as the basis for predicting the upper limit of the state of seriousness in class;

Wherein, the first retrieval condition includes: the total number of abnormal records of the learning status whose abnormal record type is the preset type within a preset time distance before and after the abnormal learning status record on the time axis is greater than or equal to the preset number threshold;

and / or,

Obtain the student information of the student attending the class;

generating a template based on preset retrieval conditions, and generating a second retrieval condition according to the student information;

Other upper limits of serious state of listening to lectures satisfying the second search condition are retrieved from the preset upper limit collection library of serious state of listening to lectures, and used as a basis for predicting the upper limit of serious state of listening to lectures.

The working principle and beneficial effects of the above-mentioned technical scheme are:

"And/or" means that there are two ways to obtain the basis for predicting the upper limit of the student's serious state of listening to the class: the first is to determine based on the continuous listening situation before the student's inattention in the history of the class; record the abnormal learning status on the time axis Expand on top, set the record time and the time point on the time axis in one-to-one correspondence, set the first search condition, and the default type is the abnormal record type that represents the student's fatigue, desertion, etc., for example: tired expression, nodding multiple times, etc., learning status When there are many abnormal records of the preset type of learning status within the time distance before and after the abnormal records, it means that the students in the class are not focused, and the previous second continuous class attendance is extracted as the basis for predicting the upper limit of the seriousness of the class. The second method is based on the upper limit of other students' seriousness in class; student information includes grade information, age information, gender information, subject information, historical subject performance information, etc., and a template is generated based on the search conditions to generate a second search based on student information Conditions: The grade of the retrieved student should be x (same as the student attending the class), the age of the retrieved student should be xx (same as the student attending the class), the gender of the retrieved student should be x (same as the student attending the class), and the subject information of the retrieved student is good The similarity with the students' good subject information is ≥90%, and the similarity between the historical subject performance information of the retrieved students and the historical subject performance information of the students is ≥85%. The upper limit comes from the following sources: 1. Obtained when predicting the upper limit of the seriousness status of other students attending lectures; 2. It is given by relevant experts who study students' lectures.

The embodiment of the present invention introduces two ways to obtain the upper limit prediction basis of the student's serious state of listening to the class, which improves the comprehensiveness of obtaining the upper limit prediction basis of the serious state of listening to the class, and indirectly improves the comprehensiveness of the upper limit prediction, especially the second acquisition method, retrieval The upper limit of other serious state of listening to lectures needs to meet the second retrieval condition. If not, the value as the basis for prediction of the upper limit of serious state of listening to lectures is not high, and the accuracy of the selection of prediction basis is improved.

In one embodiment, the adaptive allocation of detection resources for the key students and the non-key students to detect the abnormality of student learning status includes:

Count the first student total number of the key students and the second student total number of the non-key students respectively;

calculating a numerical ratio of the first total number of students to the second total number of students;

determining a detection resource allocation strategy corresponding to the number ratio from a preset detection resource allocation strategy library;

Based on the detection resource allocation strategy, the detection resources are allocated to the key students and the non-key students respectively.

The working principle and beneficial effects of the above-mentioned technical scheme are:

In the allocation of appropriate detection resources, based on the ratio of the number of key students and non-key students, the detection resource allocation strategy is determined from the detection resource allocation strategy library and implemented. The larger the number ratio, it means that it is necessary to focus on the detection of students' learning status. The more students there are, the more emphasis should be placed on the allocation of testing resources to each key student in the testing resource allocation strategy. The detection resource is the detection frequency, and the more detection resources are allocated, the higher the frequency of detection of students' learning status will be.

The embodiment of the present invention introduces the detection resource allocation strategy library, and determines and executes the detection resource allocation strategy based on the ratio of the number of key students and non-key students, which improves the rationality and allocation of suitable detection resources for key students and non-key students efficiency.

An embodiment of the present invention provides a student learning status detection system applied to online education, as shown in Figure 2, including:

The teaching teacher state information acquisition module 1 is used to obtain the first state information of the teaching teacher of the online education class when the online education class starts;

Abnormal student learning state detection template generation module 2, for generating an abnormal student learning state detection template based on the first state information and a preset student learning state abnormal detection template generation library;

The state information acquisition module 3 of the students attending the class is used to obtain the second state information of the students attending the class in the online education classroom;

The abnormality detection module 4 of the student's learning state is configured to detect the abnormality of the student's learning state for the students attending the class based on the abnormality detection template of the student's learning state and according to the second state information.

In one embodiment, the teaching teacher state information acquisition module 1 obtains the first state information of the teaching teacher in the online education classroom, and performs the following operations:

Pushing the preset teacher state selection table to the first mobile terminal of the teaching teacher;

Obtain the teaching teacher status selected by the teaching teacher from the teaching teacher status selection table;

determining first status information based on the status of the teaching teacher;

and / or,

Obtaining the first speech information of the teaching teacher through the first mobile terminal;

determining first status information based on the first speaking information;

and / or,

Obtaining the class image of the teaching teacher through the first mobile terminal;

extracting the first action information of the teaching teacher from the class image;

Based on the first action information, first state information is determined.

Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention also intends to include these modifications and variations.

Claims (10)

1. A student learning state detection method applied to online education, characterized in that, comprising: When the online education class starts, obtain the first status information of the teaching teacher of the online education class; Based on the first state information and the preset student learning state abnormality detection template generation library, generate a student learning state abnormality detection template; Acquiring the second status information of the students attending the online education class; Based on the abnormality detection template of the student's learning state, and according to the second state information, the abnormality detection of the student's learning state is performed on the student attending the class. 2. A method for detecting a student's learning status applied to online education as claimed in claim 1, wherein obtaining the first status information of the teaching teacher of the online education classroom includes: Pushing the preset teacher state selection table to the first mobile terminal of the teaching teacher; Obtain the teaching teacher status selected by the teaching teacher from the teaching teacher status selection table; determining first status information based on the status of the teaching teacher; and / or, Obtaining the first speech information of the teaching teacher through the first mobile terminal; determining first status information based on the first speaking information; and / or, Obtaining the class image of the teaching teacher through the first mobile terminal; extracting the first action information of the teaching teacher from the class image; Based on the first action information, first state information is determined. 3. A method for detecting student learning status applied to online education as claimed in claim 1, wherein the student learning status is generated based on the first status information and a preset student learning status abnormality detection template generation library Anomaly detection templates, including: Analyzing the number of information types of the first state information; When the number of information types is unique, determine the student learning state abnormality detection rule corresponding to the first state information from the student learning state abnormality detection template generation library; Based on the student learning status abnormality detection rule, generate a student learning status abnormality detection template; When the number of information types is not unique, analyze the state information feature set of the first state information based on a preset feature analysis template; Constructing a first feature description vector of the first state information based on the state information feature set; Extract multiple groups of one-to-one corresponding second feature description vectors and student learning state abnormality detection rule sets from the student learning state abnormality detection template generation library; calculating a vector similarity between the first feature description vector and any of the second feature description vectors; An abnormal student learning state detection template is generated based on the student learning state abnormality detection rule set corresponding to the second feature description vector with the largest vector similarity. 4. A method for detecting student learning status applied to online education as claimed in claim 1, wherein obtaining the second status information of students attending a class in the online education classroom includes: Obtaining the second speech information of the student attending the class through the second mobile terminal of the student attending the class; determining second status information based on the second speaking information; and / or, Obtaining the lecture image of the student attending the lecture through the second mobile terminal; extracting second action information of the student attending the class from the class attending image; Based on the second action information, second state information is determined. 5. a kind of student learning state detection method that is applied to online education as claimed in claim 1, is characterized in that, also comprises: Obtain the attendance record of the student attending the class; Based on the class attendance records, classifying the students attending classes into key students and non-key students; Adaptive allocation of detection resources for the key students and the non-key students to detect the abnormality of student learning status. 6. A method for detecting a student's learning status applied to online education as claimed in claim 5, wherein, based on the class attendance records, dividing the class-attending students into key students includes: Extracting the abnormal learning status records of the students who attended the lectures from the lectures; Based on the abnormal learning status record, determine the abnormal learning status frequency of the students attending the class; When the abnormal frequency of the learning state is greater than or equal to the preset abnormal frequency threshold of the learning state, the corresponding student attending the class will be regarded as a key student; and / or, Obtain the prediction basis for the upper limit of the student's serious state of listening to the class; Input the upper limit prediction basis of the serious state of listening to a class into the preset upper limit prediction model of the serious state of listening to a class to determine the upper limit of the serious state of listening to a class; Extracting the first continuous lecture situation within the latest preset time range of the student attending the lecture from the lecture attendance record; Based on the first continuous lecture situation, determine whether the student has reached the upper limit of the seriousness of the lecture; When yes, the corresponding student attending the class will be regarded as the key student. 7. A method for detecting a student's learning state applied to online education as claimed in claim 6, wherein obtaining the basis for predicting the upper limit of the student's serious state of listening to a class includes: Acquiring the recording time of the abnormal learning status record; Based on the recording time, the abnormal learning state is recorded on a preset time axis; Retrieving abnormal records of the target learning state that meet the preset first retrieval condition from the time axis; Extract the second continuous lecture situation within the time range before the recording time of the abnormal record of the target learning state from the lecture record, and use it as the basis for predicting the upper limit of the state of seriousness in class; Wherein, the first retrieval condition includes: the total number of abnormal records of the learning status whose abnormal record type is the preset type within a preset time distance before and after the abnormal learning status record on the time axis is greater than or equal to the preset number threshold; and / or, Obtain the student information of the student attending the class; generating a template based on preset retrieval conditions, and generating a second retrieval condition according to the student information; Other upper limits of serious state of listening to lectures satisfying the second retrieval condition are retrieved from the preset upper limit collection library of serious state of listening to lectures, and used as a basis for predicting the upper limit of serious state of listening to lectures. 8. A method for detecting students' learning status applied to online education as claimed in claim 5, characterized in that, carrying out adaptive allocation of detection resources for abnormal detection of students' learning status to said key students and said non-key students, include: Count the first student total number of the key students and the second student total number of the non-key students respectively; calculating a numerical ratio of the first total number of students to the second total number of students; determining a detection resource allocation strategy corresponding to the number ratio from a preset detection resource allocation strategy library; Based on the detection resource allocation strategy, the detection resources are allocated to the key students and the non-key students respectively. 9. A student learning status detection system applied to online education, characterized in that it includes: The teaching teacher state information acquisition module is used to obtain the first state information of the teaching teacher of the online education class when the online education class starts; A student learning state abnormality detection template generation module is used to generate a student learning state abnormality detection template based on the first state information and a preset student learning state abnormality detection template generation library; The state information acquisition module of students attending classes is used to obtain the second state information of students attending classes in the online education classroom; The abnormality detection module of the student's learning state is configured to detect the abnormality of the student's learning state for the students attending the class based on the abnormality detection template of the student's learning state and according to the second state information. 10. A student learning state detection system applied to online education as claimed in claim 9, wherein the teaching teacher state information acquisition module obtains the first state information of the teaching teacher in the online education classroom, and executes Do as follows: Pushing the preset teacher state selection table to the first mobile terminal of the teaching teacher; Obtain the teaching teacher status selected by the teaching teacher from the teaching teacher status selection table; determining first status information based on the status of the teaching teacher; and / or, Obtaining the first speech information of the teaching teacher through the first mobile terminal; determining first status information based on the first speaking information; and / or, Obtaining the class image of the teaching teacher through the first mobile terminal; extracting the first action information of the teaching teacher from the class image; Based on the first action information, first state information is determined.

Images