CN116579894A - Teacher-student interaction detection method based on intelligent classroom of Internet of things - Google Patents
Teacher-student interaction detection method based on intelligent classroom of Internet of things Download PDFInfo
- Publication number
- CN116579894A CN116579894A CN202310364609.6A CN202310364609A CN116579894A CN 116579894 A CN116579894 A CN 116579894A CN 202310364609 A CN202310364609 A CN 202310364609A CN 116579894 A CN116579894 A CN 116579894A
- Authority
- CN
- China
- Prior art keywords
- interaction
- student
- students
- teacher
- degree
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 230000003993 interaction Effects 0.000 title claims abstract description 712
- 238000001514 detection method Methods 0.000 title claims abstract description 9
- 230000006399 behavior Effects 0.000 claims abstract description 66
- 238000011156 evaluation Methods 0.000 claims abstract description 32
- 230000008859 change Effects 0.000 claims abstract description 31
- 230000000007 visual effect Effects 0.000 claims abstract description 6
- 230000002452 interceptive effect Effects 0.000 claims description 80
- 238000000034 method Methods 0.000 claims description 58
- 230000008451 emotion Effects 0.000 claims description 29
- 230000009471 action Effects 0.000 claims description 18
- 230000000694 effects Effects 0.000 claims description 16
- 230000000977 initiatory effect Effects 0.000 claims description 16
- 238000013210 evaluation model Methods 0.000 claims description 14
- 238000012545 processing Methods 0.000 claims description 13
- 208000031636 Body Temperature Changes Diseases 0.000 claims description 9
- 230000036760 body temperature Effects 0.000 claims description 9
- 239000013598 vector Substances 0.000 claims description 8
- 230000011218 segmentation Effects 0.000 claims description 7
- 238000010586 diagram Methods 0.000 claims description 6
- 238000004891 communication Methods 0.000 claims description 5
- 238000010801 machine learning Methods 0.000 claims description 3
- 238000012549 training Methods 0.000 claims description 3
- 238000010276 construction Methods 0.000 claims description 2
- 238000012216 screening Methods 0.000 claims 1
- 238000013461 design Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000012821 model calculation Methods 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/20—Education
- G06Q50/205—Education administration or guidance
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/205—Parsing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/237—Lexical tools
- G06F40/242—Dictionaries
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/48—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use
- G10L25/51—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use for comparison or discrimination
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/48—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use
- G10L25/51—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use for comparison or discrimination
- G10L25/63—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use for comparison or discrimination for estimating an emotional state
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Tourism & Hospitality (AREA)
- Human Computer Interaction (AREA)
- Signal Processing (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Software Systems (AREA)
- Strategic Management (AREA)
- Educational Technology (AREA)
- Educational Administration (AREA)
- Human Resources & Organizations (AREA)
- Medical Informatics (AREA)
- Computing Systems (AREA)
- Evolutionary Computation (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Business, Economics & Management (AREA)
- Primary Health Care (AREA)
- Marketing (AREA)
- Economics (AREA)
- Child & Adolescent Psychology (AREA)
- Hospice & Palliative Care (AREA)
- Psychiatry (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Electrically Operated Instructional Devices (AREA)
Abstract
The application provides a teacher-student interaction detection method based on an intelligent classroom of the Internet of things, which comprises the following steps: acquiring interaction information of teachers and students according to visual and auditory information of interaction in a classroom; obtaining the learning ability of the students according to the interaction information of the students; judging the interaction behavior degree of the students based on the learning ability and learning evaluation of the students; judging the interaction tendency of a teacher according to the learning ability value of the students participating in the interaction and the interaction behavior degree of the students; predicting the student interaction type according to the physical sign state and the active interaction duty ratio of student interaction; judging the reasons that students do not love the interaction according to the interaction content of the students; based on interaction information of teachers and students, tendency degrees of the students to different interaction modes are obtained; and establishing a student interaction enthusiasm assessment model based on the interaction behaviors of the students, and obtaining the change of interaction tendency of the students in different interaction modes.
Description
Technical Field
The application relates to the technical field of information, in particular to a teacher-student interaction detection method based on an intelligent classroom of the Internet of things.
Background
The teacher-student interaction is a common mode for teachers to increase the atmosphere of a classroom in class, and is usually initiated by the teacher first, then the teacher invites the classmates to cooperate with the teacher to achieve a certain teaching purpose, make something, solve which topic, or assist the teacher in completing something. However, in the daily classroom teaching process, uneven interaction between teachers and students often occurs, for example, when a teacher invites students to interact, the students which interact are always selected, and other students which participate in the interaction and students which do not participate in the interaction are ignored; when a teacher initiates interaction, some students not participating in the interaction do not really want to participate in the interaction, and some other students do not participate in the interaction, so that the teacher can form interaction with the students not want to participate in the interaction, and the later interaction tendency of the students can be reduced; therefore, the design of the teacher-student interaction balancing method capable of improving the interaction tendency of students has important practical significance and theoretical research value;
Disclosure of Invention
The invention provides a teacher-student interaction detection method based on an intelligent classroom of the Internet of things, which mainly comprises the following steps:
acquiring interaction information of teachers and students according to visual and auditory information of interaction in a classroom; obtaining the learning ability of the students according to the interaction information of the students; judging the interactive behavior degree of the student based on the learning ability and learning evaluation of the student, specifically comprising: judging the interaction behavior degree of students; the interactive tendency of a teacher is judged according to the learning ability value of the student participating in the interaction and the interactive behavior degree of the student, and the interactive tendency of the teacher is judged according to the learning ability value of the student participating in the interaction and the interactive behavior degree of the student, and the method specifically comprises the following steps: judging the interaction tendency of a teacher, and judging whether the teacher needs to interact with students with different learning abilities according to the interaction tendency of the teacher; predicting the student interaction type according to the sign state and the active interaction duty ratio during student interaction, and specifically comprising the following steps: judging the interaction type of students; the reasons for the lovely interaction of the students are judged through the interaction content of the students, and the reasons for the lovely interaction of the students are judged through the interaction content of the students, specifically comprising: based on the student interaction type and the interaction content, acquiring student interaction audio content, calculating the text similarity of the interaction content of a teacher and a student and emotion analysis of the student interaction content according to the student interaction audio content, constructing a student interaction completion degree model, and judging the reasons that the student does not like interaction; based on interaction information of teachers and students, the tendency degree of the students to different interaction modes is obtained, and the method specifically comprises the following steps: judging whether the student is more prone to an interaction type based on the tendency degree of the student to different interaction modes; establishing a student interaction enthusiasm assessment model based on the interaction behaviors of students, and acquiring the change of interaction tendency of the students in different interaction modes; according to the different interaction tendency changes of students under different interactions, judging the interaction mode suitable for the students comprises the following steps: and acquiring the interaction tendency change of the student after the teacher performs the interaction mode suitable for the student.
Further optionally, the obtaining the interaction information of teachers and students according to the visual and audible information of the interaction in the classroom includes:
the teacher information and the student information are acquired through a school affair management system; the method comprises the steps of constructing a recognition model of teacher-student interaction based on a machine learning supervised mode recognition method, training the model by using a large amount of action information and audio information of interaction between teachers and students, and judging a mode of the teacher to initiate interaction according to the interaction information between the teachers and the students by the recognition model of the teacher-student interaction; after the approval of the school side and the teacher and student is obtained, real-time identity information of the teacher and the student in the classroom is acquired in real time through cameras at a plurality of angles in the classroom by adopting a face recognition method, and action information during interaction of the teacher and the student is obtained; after the approval of the school side and the teacher and student is obtained, audio information during the interaction of the teacher and the student is obtained through an audio pick-up device in the classroom; transmitting the action information and the audio information to a recognition model of the interaction between the teachers and the students, and judging the mode of the interaction by the recognition model of the interaction between the teachers and the students; and counting the mode of interaction initiation by a teacher in a classroom according to the real-time identity information and the interaction mode, and storing the times and interaction time of participation of students in the interaction into a teacher-student interaction information database.
Further optionally, the obtaining the learning ability of the student according to the interaction information of the student includes:
the student information includes: usual score, job completion degree; the ordinary score is recorded by a school educational administration system; the degree of completion of the homework is obtained through statistics of a teacher on the homework of the students; firstly, a student learning ability assessment model M=A (N-S) +B.times.H is built based on the usual achievements and the homework completion degrees of students, wherein A, B is the weight of the achievements and the homework completion degrees respectively, S is the ranking of the achievements of the students, N is the number of class persons, and H is the completion degree of the usual homework; the assessment is performed based on a student learning ability assessment model, and the larger the M value is, the stronger the learning ability of the student is.
Further optionally, the determining the degree of student interaction behavior based on the student learning ability and the learning evaluation includes:
a teacher-student communication meeting is set at regular intervals in a school; the teacher-student interaction refers to the information about deep interaction between the teacher and each student for learning and classroom interaction; the teacher fills in student evaluation of the students according to the communicated content; judging the interaction behavior degree of the students according to the learning ability and learning evaluation of the students; comprising the following steps: judging the interaction behavior degree of students;
the judging of the interactive behavior degree of the student specifically comprises the following steps:
Student evaluation includes classroom interaction behavior, interaction performance, and student performance. Student performance is a ranking of the performance of students at a class. The interaction behavior is three types, namely, unwilling interaction, general interaction and willing active interaction. The interactive performance is mainly aimed at students who are unwilling to interact in the degree of interaction behavior, and the reason why the students are unwilling to interact is analyzed.
Further optionally, the determining the interaction tendency of the teacher according to the learning ability value of the student participating in the interaction and the interaction behavior degree of the student includes:
counting the learning ability value of students interacted with teachers and the interaction behavior degree of the students; counting the sum of interaction times of a teacher and students; according to the learning ability value of the student and the interaction behavior degree of the student, constructing a model to calculate the interaction tendency degree of the teacher to different types of students; judging whether the teacher has interaction tendency for a certain type of student according to the interaction tendency degree of the teacher; comprising the following steps: judging the interaction tendency of a teacher; judging whether the teacher needs to interact with students with different learning abilities according to the interaction tendency of the teacher;
the judging of the interaction tendency of the teacher specifically comprises the following steps:
through the high definition digtal camera in the religion, statistics is at every turn teachers and students interactive object. And recording the interaction times of teachers and different students based on the interaction of teachers and students each time. And counting the sum of the interaction times of the teacher and the students. And classifying and ranking the evaluation of the students according to the teacher, and ranking the students with the same interaction behavior degree in the evaluation according to the learning ability of the students. Building teacher interaction tendency model X= (C) based on learning ability value of student interacting with teacher and interaction times of student and teacher 1 *M 1 +C 2 *M 2 +...+C n *M n )/(C 1 +C 2 +...+C n ) Wherein n represents n-bit classmates in class, C 1 Xi Nengli rank the number of interactions of the first student with the teacher, M, representing the learning ability value of the input student 1 Calculating the learning ability value of the student through a student learning ability evaluation model, C 2 Is the interaction times of the students and teachers, M, of which the learning ability values of the students are input and the students of which the learning ability values are ranked second Xi Nengli 2 Calculating the learning ability value of the student through a student learning ability evaluation model, and similarly calculating the average value of the learning ability of all students interacted with a teacher in the input; dividing students into three categories according to the interaction behavior degree of the students; respectively inputting the learning ability values of students with the same interaction behavior degree and the interaction times of the students and teachers into a teacher interaction tendency model X to calculate a Xi Nengli average value; and calculating the average value of the learning ability of the teacher-interacted students according to the teacher-interacted tendency model.
Judging whether the teacher needs to interact with students with different learning abilities according to the interaction tendency of the teacher, and specifically comprising the following steps:
students whose learning ability value is before a certain threshold value in class are defined as excellent students. And calculating the average value A of the learning ability values of the students, wherein the interaction behavior degree in the student evaluation is willing to actively interact and the class ranks before a certain threshold value. And comparing the average value S of the learning ability of the students with which the teacher is interacted with A, and if the average value S of the learning ability of the students with which the teacher is interacted is greater than the value A of the learning ability of the students with which the class is ranked before a certain threshold value, considering that the teacher is favored to interact with excellent students, and reminding the teacher to interact with other students with which the teacher is willing to interact but with the lower level in the performance by the system. If the interactive behavior degree of the teacher interaction is that the average value S of the learning abilities of the students willing to be active is smaller than the learning ability value A of the students ranked before a certain threshold value, the interactive tendency of the teacher is considered to be relatively average.
Further optionally, the predicting the student interaction type according to the sign state and the active interaction ratio when the student interacts includes:
the interaction types of students are classified into active type, normal passive type and passive type; the active mode is defined as that the student is willing to actively interact with the teacher, the normal passive mode means that the student does not feel against the interaction with the teacher, but does not actively interact with the teacher, and the passive mode is that the student does not wish to even fear to interact with the teacher; inputting the audio information and the action information of the students during interaction into an identification model to judge whether the students actively initiate the interaction, and judging the interaction type of the students according to the tension degree of the students and the proportion of the active initiation interaction to the total interaction times; comprising the following steps: judging the interaction type of students;
the judging of the student interaction type specifically comprises the following steps:
the real-time identity information of teachers and students in the classroom is acquired in real time through cameras with multiple angles in the classroom by adopting a face recognition method, and action information during interaction of the teachers and the students is acquired. And acquiring audio information during interaction of teachers and students through an audio pick-up device in the teaching room. And transmitting the action information and the audio information to an identification model of the interaction between teachers and students to analyze and judge whether the students actively initiate the interaction. During classroom interaction, video information containing students and a hot spot area diagram of the students are acquired through a camera and an infrared sensor respectively, and body temperature changes of the students during interaction are acquired. And identifying the collected video information and the collected hot spot area map to obtain the human body characteristic data of the detected student. And constructing a tension degree model N=T based on the student human body characteristic data, wherein T is the difference between the body temperature of the student during interaction and the body temperature before interaction. And calculating the tension degree of the student interaction by using the change value of the student body temperature. And establishing a preset student interaction type analysis model based on the number of active interactions of the students and the tension and tension degree of the interactions, and judging the interaction type of the students according to the ratio of the number of active interactions of the students to the total number of interactions and the tension and tension degree of the interactions of the students. Counting the number of times of the active initiation of the interaction and the total interaction number of the students, and setting the interaction type analysis model to judge the interaction type of the students as active if the ratio of the number of times of the active initiation of the interaction of the students to the total interaction number of the students is larger than a certain threshold value. And calculating the tension of the student based on the analysis model of whether the student is tension and the tension degree in each interaction type, if the tension degree is lower than a certain threshold value, and the ratio of the number of times the student actively initiates the interaction to the total number of times the student is smaller than a certain threshold value, determining that the student interaction type is normal passive type by the analysis model of the interaction type, otherwise, if the tension degree is higher than a certain threshold value, and the ratio of the number of times the student actively initiates the interaction to the total number of times the student is smaller than a certain threshold value, determining that the student interaction type is passive type by the analysis model of the interaction type.
Further optionally, the determining, by the interaction content of the student, a reason why the student does not love the interaction includes:
acquiring interactive audio information of teachers and students of passive students; respectively processing the audio information of the teacher and the student to generate interactive audio content; performing text similarity processing on the interactive audio content of the teacher and the student, and judging whether the interactive audio content of the teacher and the student is similar; carrying out emotion analysis on the interactive audio content of the student; finally, judging the reason why students do not like to interact according to the text similarity and emotion analysis results and the student learning score; comprising the following steps: based on the student interaction type and the interaction content, acquiring student interaction audio content; according to the student interaction audio content, calculating the text similarity of the interaction content of a teacher and a student and emotion analysis of the student interaction content; constructing a student interaction completion degree model, and judging the reasons of the student loving interaction;
based on the student interaction type and the interaction content, the method for acquiring the student interaction audio content specifically comprises the following steps:
based on student interaction behavior analysis, passive and passive types of student information are screened out. And the audio pickup device is used for acquiring interactive audio information of a teacher and the students during interaction of passive students, and the audio acquired by the audio pickup device is preprocessed. And extracting and identifying the interactive audio contents of the teacher and the students, and respectively generating the interactive audio contents of the teacher and the students.
According to the student interaction audio content, calculating the text similarity of the interaction content of a teacher and a student and emotion analysis of the student interaction content, specifically comprising:
inputting the content of the teacher and student interaction audio into an N-gram model to perform Chinese word segmentation on the student interaction audio. And respectively carrying out text vectorization on the segmented text by using a word2vec model to obtain vectorized representation of each word. And weighting and adding each dimension of the vectors to form a text vector, and further calculating the similarity of the text by using the cosine distance. And (5) traversing the words in the segmented sentences one by one and comparing the words with the BosonNLP emotion dictionary. If the word hits the dictionary, processing of the corresponding weight is performed. The positive word weight is addition, the negative word weight is subtraction, the negative word weight is the opposite number, the degree adverb weight is multiplied by the word weight modified by the negative word weight, and the emotion weight value of the interactive content text is obtained through summation.
The construction of the student interaction completion degree model and the judgment of the reasons of the student loving interaction specifically comprise the following steps:
and constructing a student interaction completion degree model K=L.times.M.times.M (T-S) based on the similarity of the student interaction content text and the emotion analysis of the student interaction content and the student score ranking, wherein L is the student interaction content text similarity, if the similarity is higher than a certain threshold value, the L value is 1, otherwise, the L value is 0, M is the emotion analysis weight value of the student interaction content, if the weight value is a non-negative number, the M value is 1, otherwise, the M value is-1, T is the total number of classes +1, and S is the student score ranking. And judging the reason why the students do not love the interaction based on the student interaction completion degree model. If the value of K is 0, the student is considered to be one of the reasons why the student does not like to interact is not aware of the content of the interaction. If the value of K is smaller than 0, the student is considered to be one of the reasons for dislike of interaction as low enthusiasm for interaction. If the absolute value of K is smaller than a certain threshold, the student is considered to be one of the reasons for the dislike of interaction as weak learning ability. If the absolute value of K is larger than a certain threshold, the student is considered to have weak social ability as one of the reasons for loving interaction.
Further optionally, the obtaining the tendency degree of the student to different interaction modes based on the interaction information of the teachers and students includes:
acquiring teacher information from a teacher information database; taking the mode of initiating interaction by a teacher in each interaction of students and the time of the interaction as each interaction mark label; classifying different interactive mark labels according to an interactive mode; summing the interaction time in the interaction mark labels of different interaction modes to obtain the tendency degree of students to the interaction modes; comprising the following steps: judging whether the student is more prone to an interaction type based on the tendency degree of the student to different interaction modes;
based on the tendency degree of the students to different interaction modes, judging whether the students are more prone to an interaction type or not, specifically comprising:
and processing the trend degree data based on different trend degrees of different interaction modes of the students to obtain standard data. And analyzing based on the standard data and extracting the characteristics of the interaction tendency of the students by adopting a characteristic extractor. A classifier is employed to determine whether student interaction is too prone to an interaction type based on the extracted interaction-prone features.
Further optionally, the establishing a student interaction enthusiasm assessment model based on the student interaction behavior, and the obtaining the change of the student interaction tendency under different interaction modes includes:
Counting active sponsors of each teacher-student interaction through a high-definition camera in the teaching room; during classroom interaction, the camera and the infrared sensor are respectively used for carrying out interactionCollecting video information of students and a hot spot area diagram of the students, and acquiring the body temperature change of the students during interaction; calculating the tension degree of the student interaction based on the tension degree model and the body temperature change of the student during the interaction; establishing a student interaction enthusiasm assessment model, wherein E=Q 1 *C-Q 2 *T-Q 3 * N+m, where E is the student interaction aggressiveness,
Q 1 、Q 2 、Q 3 for different weight values, C is the number of times that the student actively initiates the interaction, T is whether the student is stressed or not during the interaction, the stress and relaxation values are respectively 1, 0, -1, N is the stress degree of the student for the current interaction, and M is a constant; counting the interaction information of students after each interaction, and updating the interaction activity degree of the students; if the interaction activity degree of the students changes after different interaction modes are carried out, the interaction tendency of the students is changed.
Further optionally, the determining the interaction mode suitable for the student according to the change of different interaction tendencies of the student under different interactions includes:
counting the information of each interaction according to an evaluation model of the interaction enthusiasm of students; based on the latest ten interactive enthusiasms of each classmate, counting the change value of the last enthusiasm of students in each interactive mode; adding the enthusiasm change values under each interaction mode to obtain the influence of different interaction modes on the interaction enthusiasm of students; comprising the following steps: the method comprises the steps of obtaining interaction tendency change of students after a teacher performs an interaction mode suitable for the students;
The method for obtaining the interaction tendency change of the student after the teacher performs the interaction mode suitable for the student specifically comprises the following steps:
the teacher takes an interaction mode suitable for students in class. And calculating the enthusiasm evaluation value of five interactions of the student after the teacher adopts a proper interaction mode according to the evaluation model of the enthusiasm of the student. And judging the interaction tendency change of the students according to the final five-time student enthusiasm evaluation values.
The technical scheme provided by the embodiment of the application can have the following beneficial effects:
the application designs a teacher-student interaction balancing method capable of improving the interaction tendency of students, which adjusts the interaction tendency of the teacher and prompts the teacher to select students outside the interaction tendency to form interaction, and for students not participating in the interaction, the interaction tendency of the students is changed by adjusting the interaction method, so that the interaction tendency of the students is improved.
Drawings
Fig. 1 is a flowchart of a teacher-student interaction detection method based on an intelligent classroom of the internet of things.
Fig. 2 is a flowchart of a teacher-student interaction detection method based on the intelligent classroom of the internet of things.
Detailed Description
For a further understanding of the present application, the present application will be described in detail with reference to the drawings and examples. The application is described in further detail below with reference to the drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the application and are not limiting of the application. It should be noted that, for convenience of description, only the portions related to the application are shown in the drawings.
The teacher-student interaction detection method based on the intelligent classroom of the Internet of things specifically comprises the following steps:
and step 101, acquiring interaction information of teachers and students according to visual and auditory information of interaction in a classroom.
The teacher information and the student information are acquired through a school affair management system. A supervised mode recognition method based on machine learning builds a recognition model of teacher-student interaction, a large amount of action information and audio information of interaction between teachers and students are used for training the model, and the recognition model of the teacher-student interaction judges a mode of the teacher to initiate interaction according to the interaction information between the teachers and the students. After the approval of the school side and the teacher and student is obtained, the real-time identity information of the teacher and the student in the classroom is acquired in real time through cameras at a plurality of angles in the classroom by adopting a face recognition method, and the action information of the teacher and the student in interaction is acquired. After the approval of the school side and the teacher and student is obtained, the audio information during the interaction of the teacher and the student is obtained through the audio pick-up device in the classroom. And transmitting the action information and the audio information to a recognition model of the interaction between the teachers and the students, and judging the mode of the interaction by the recognition model of the interaction between the teachers and the students. And counting the mode of interaction initiation by a teacher in a classroom according to the real-time identity information and the interaction mode, and storing the times and interaction time of participation of students in the interaction into a teacher-student interaction information database. For example, to count the interaction mode of a teacher and a student, identity information of the teacher and the student in a class is first identified through a camera. And then, the image information of the camera and the audio information recorded by the audio pick-up device are processed by an identification model of the interaction between teachers and students to judge the mode of the interaction initiated by the teachers and record according to the object information of the students. At the same time, the type and the times of the interaction initiated by the teacher are recorded in the interaction information record of the student.
Step 102, learning ability of the students is obtained according to the interaction information of the students.
The student information includes: usual achievements and job completion degrees. The usual achievements are recorded by a school educational administration system. The degree of completion of the homework is obtained by the statistics of the teacher on the homework of the students. Firstly, a student learning ability assessment model M=A (N-S) +B.times.H is built based on the usual achievements and the homework completion degrees of students, wherein A, B is the weight of the achievements and the homework completion degrees respectively, S is the ranking of the achievements of the students, N is the number of class persons, and H is the completion degree of the usual homework. The assessment is performed based on a student learning ability assessment model, and the larger the M value is, the stronger the learning ability of the student is. For example, 50 persons are shared by a class, and the ranks of the students are 2, 10 and 38 respectively for A, B, C students, and the completion degrees of the ordinary works are 1, 0.95 and 0.7 respectively. The weight of A is 1, the weight of B is 10, and the learning ability of the A, B, C three students is obtained by model calculation and is 59, 49.5 and 19 respectively. It indicates that a has the strongest learning ability.
And step 103, judging the interactive behavior degree of the students based on the learning ability and learning evaluation of the students.
The school sets up teachers and students to communicate regularly. The teacher-student interaction refers to information about deep interaction of the teacher with each student. The teacher fills in the student evaluation of the student according to the content of the communication. And judging the interaction behavior degree of the students according to the learning ability and learning evaluation of the students.
And judging the interaction behavior degree of the students.
Student evaluation includes classroom interaction behavior, interaction performance, and student performance. Student performance is a ranking of the performance of students at a class. The interaction behavior is three types, namely, unwilling interaction, general interaction and willing active interaction. The interactive performance is mainly aimed at students who are unwilling to interact in the degree of interaction behavior, and the reason why the students are unwilling to interact is analyzed. For example, in a teacher-student communication of a school organization, after a teacher communicates with a student, the student is analyzed to have a tendency to be reluctant to interact, and the reason why he is reluctant to interact is to speak before the person through further communication. The teacher's assessment of the student in the student assessment may be: degree of interaction behavior: unwilling to interact, interactive performance: and (5) inward and inputting the student achievement ranking.
And 104, judging the interaction tendency of the teacher according to the learning ability value of the students participating in the interaction and the interaction behavior degree of the students.
And counting the learning ability value of the student interacted with the teacher and the interaction behavior degree of the student. And counting the sum of the interaction times of the teacher and the students. And a model is constructed according to the learning ability value of the student and the interaction behavior degree of the student to calculate the interaction tendency degree of the teacher to different types of students. Judging whether the teacher has interaction tendency for a certain type of student according to the interaction tendency degree of the teacher.
And judging the interaction tendency of the teacher.
Counting the object interacted by teachers and students each time through a high-definition camera in the teaching room; recording the interaction times of teachers and different students based on the interaction of teachers and students each time; counting the sum of interaction times of a teacher and students; classifying and ranking the evaluation of the students according to teacher, and ranking the students with the same interactive behavior degree in the evaluation according to the learning ability of the students; building teacher interaction tendency model X= (C) based on learning ability value of student interacting with teacher and interaction times of student and teacher 1 *M 1 +C 2 *M 2 +...+C n *M n )/(C 1 +C 2 +...+C n ) Wherein n represents a shiftIn the stage there are n-bit classmates, C 1 Xi Nengli rank the number of interactions of the first student with the teacher, M, representing the learning ability value of the input student 1 Calculating the learning ability value of the student through a student learning ability evaluation model, C 2 Is the interaction times of the students and teachers, M, of which the learning ability values of the students are input and the students of which the learning ability values are ranked second Xi Nengli 2 Calculating the learning ability value of the student through a student learning ability evaluation model, and similarly calculating the average value of the learning ability of all students interacted with a teacher in the input; dividing students into three categories according to the interaction behavior degree of the students; respectively inputting the learning ability values of students with the same interaction behavior degree and the interaction times of the students and teachers into a teacher interaction tendency model X to calculate a Xi Nengli average value; and calculating the average value of the learning ability of the teacher-interacted students according to the teacher-interacted tendency model. For example, the teacher has a total number of interactions of 10, 2 with students a, 3 with students B, 4 with students C, and 1 with students D, wherein A, B, C, D have learning ability values of 50, 55, 35, and 20, respectively. The interaction behavior of A, B, C, D is respectively willing to actively interact, general, unwilling to interact and willing to actively interact. The average values of learning ability of different interactive behavior degrees of the teacher interactive students are calculated as follows: willing to actively interact: (2×50+1×20)/(2+1) =40, general: (3 x 55)/3=55, unwilling to interact actively: 4 x 35/4=35. The comparison of the average value with the average value of the learning ability of the students with different interaction trends in the class can judge whether the interaction trend of the teacher is related to the learning ability of the students.
Judging whether the teacher needs to interact with students with different learning abilities according to the interaction tendency of the teacher.
Students whose learning ability value is before a certain threshold value in class are defined as excellent students. And calculating the average value A of the learning ability values of the students, wherein the interaction behavior degree in the student evaluation is willing to actively interact and the class ranks before a certain threshold value. And comparing the average value S of the learning ability of the students with which the teacher is interacted with A, and if the average value S of the learning ability of the students with which the teacher is interacted is greater than the value A of the learning ability of the students with which the class is ranked before a certain threshold value, considering that the teacher is favored to interact with excellent students, and reminding the teacher to interact with other students with which the teacher is willing to interact but with the lower level in the performance by the system. If the interactive behavior degree of the teacher interaction is that the average value S of the learning abilities of the students willing to be active is smaller than the learning ability value A of the students ranked before a certain threshold value, the interactive tendency of the teacher is considered to be relatively average. For example, the total number of interactions by the teacher is 10, wherein 2 times with the a-student, 1 time with the B-student, 4 times with the C-student, and 3 times with the D-student, wherein the learning ability values of the A, B, C, D-student are 50, 20, 35, and 55, respectively, and A, B, C, D are all students willing to actively interact. The average learning ability of the teacher interacted students is calculated to be (2×50+1×20+35×4+55×3)/10 to be 42.5. Students in a class willing to actively interact and ranked the top 25% have a learning ability value of 38. At this point, the average student learning ability value for teacher interaction is greater than the student learning ability value before class ranking is 25% before, and the system will schedule the teacher to tend to interact with the superior students, at which point the teacher will be reminded to interact with other students willing to interact but at the lower level in performance.
Step 105, predicting the student interaction type according to the sign state and the active interaction ratio of the student during interaction.
The types of interaction of students are classified into active, normal passive, passive. The active mode is defined as that the student is willing to actively interact with the teacher, the normal passive mode means that the student does not feel the reaction to interact with the teacher, but does not actively interact with the teacher, and the passive mode is that the student does not wish to even fear to interact with the teacher. And inputting the audio information and the action information of the student during interaction into the recognition model to judge whether the student actively initiates the interaction, and judging the interaction type of the student according to the tension degree of the student and the proportion of the active initiation interaction to the total interaction times.
And judging the interaction type of the students.
The real-time identity information of teachers and students in the classroom is acquired in real time through cameras with multiple angles in the classroom by adopting a face recognition method, and action information during interaction of the teachers and the students is acquired. And acquiring audio information during interaction of teachers and students through an audio pick-up device in the teaching room. And transmitting the action information and the audio information to an identification model of the interaction between teachers and students to analyze and judge whether the students actively initiate the interaction. During classroom interaction, video information containing students and a hot spot area diagram of the students are acquired through a camera and an infrared sensor respectively, and body temperature changes of the students during interaction are acquired. And identifying the collected video information and the collected hot spot area map to obtain the human body characteristic data of the detected student. And constructing a tension degree model N=T based on the student human body characteristic data, wherein T is the difference between the body temperature of the student during interaction and the body temperature before interaction. And calculating the tension degree of the student interaction by using the change value of the student body temperature. And establishing a preset student interaction type analysis model based on the number of active interactions of the students and the tension and tension degree of the interactions, and judging the interaction type of the students according to the ratio of the number of active interactions of the students to the total number of interactions and the tension and tension degree of the interactions of the students. Counting the number of times of the active initiation of the interaction and the total interaction number of the students, and setting the interaction type analysis model to judge the interaction type of the students as active if the ratio of the number of times of the active initiation of the interaction of the students to the total interaction number of the students is larger than a certain threshold value. And calculating the tension of the student based on the analysis model of whether the student is tension and the tension degree in each interaction type, if the tension degree is lower than a certain threshold value, and the ratio of the number of times the student actively initiates the interaction to the total number of times the student is smaller than a certain threshold value, determining that the student interaction type is normal passive type by the analysis model of the interaction type, otherwise, if the tension degree is higher than a certain threshold value, and the ratio of the number of times the student actively initiates the interaction to the total number of times the student is smaller than a certain threshold value, determining that the student interaction type is passive type by the analysis model of the interaction type. The number of times the student actively initiates the interaction and the total number of interactions of the student are counted. If the ratio of the number of times of the active initiation of the student to the total number of times of the student is greater than a certain threshold (e.g., 30%), the student's type of interaction is considered to be active. The information of each interaction is collected through the camera and the infrared sensor. The infrared sensor collects the temperature change before and after the interaction of the student, the study shows that the temperature of a person can rise when the student is stressed, and if the temperature change of the student exceeds 0.2 degree, the student is considered to be stressed in the interaction process. The collected information is used for judging the initiative and the tension degree of the interactive student. If the interaction is actively initiated by the student and the emotion of the student is relaxed and pleasant when the student interacts, the tension of the student in the student interaction type analysis model is reduced. The reliability of the judgment of the model is increased by continuously recording data. And finally judging the interaction type of each student according to the model.
And 106, judging the reasons that the students are loved to interact according to the interaction content of the students.
And acquiring interactive audio information of teachers and students of passive students. And respectively processing the audio information of the teacher and the student to generate interactive audio content. And carrying out text similarity processing on the interactive audio content of the teacher and the students, and judging whether the interactive audio content of the teacher and the students are similar. And carrying out emotion analysis on the interactive audio content of the student. And finally judging the reason why students do not like to interact according to the text similarity and emotion analysis results and the student learning score.
Based on the student interaction type and the interaction content, acquiring student interaction audio content.
Based on student interaction behavior analysis, passive and passive types of student information are screened out. And the audio pickup device is used for acquiring interactive audio information of a teacher and the students during interaction of passive students, and the audio acquired by the audio pickup device is preprocessed. And extracting and identifying interactive audio contents of the teacher and the students, respectively generating interactive audio contents of the teacher and the students, and recording interactive audio of the teacher and the students by using audio pick-up equipment when the passive type students interact. And processing the interactive audio and extracting interactive audio contents of teachers and students.
And calculating the text similarity of the interactive content of the teacher and the student and emotion analysis of the interactive content of the student according to the interactive audio content of the student.
Inputting the content of the teacher and student interaction audio into an N-gram model to perform Chinese word segmentation on the student interaction audio. And respectively carrying out text vectorization on the segmented text by using a word2vec model to obtain vectorized representation of each word. And weighting and adding each dimension of the vectors to form a text vector, and further calculating the similarity of the text by using the cosine distance. And (5) traversing the words in the segmented sentences one by one and comparing the words with the BosonNLP emotion dictionary. If the word hits the dictionary, processing of the corresponding weight is performed. The positive word weight is addition, the negative word weight is subtraction, the negative word weight is the opposite number, the degree adverb weight is multiplied by the word weight modified by the negative word weight, and the emotion weight value of the interactive content text is obtained through summation. Inputting the audio content of teachers and students into an N-gram model to perform Chinese word segmentation, performing text vectorization and weighted addition on the text after word segmentation to obtain text vectors, and performing cosine distance calculation on the text vectors to obtain text similarity. And carrying out emotion analysis on the text after word segmentation of the interactive audio content to obtain an emotion weight value.
And constructing a student interaction completion degree model, and judging the reasons that students do not love interaction.
And constructing a student interaction completion degree model K=L.times.M.times.M (T-S) based on the similarity of the student interaction content text and the emotion analysis of the student interaction content and the student score ranking, wherein L is the student interaction content text similarity, if the similarity is higher than a certain threshold value, the L value is 1, otherwise, the L value is 0, M is the emotion analysis weight value of the student interaction content, if the weight value is a non-negative number, the M value is 1, otherwise, the M value is-1, T is the total number of classes +1, and S is the student score ranking. And judging the reason why the students do not love the interaction based on the student interaction completion degree model. If the value of K is 0, the student is considered to be one of the reasons why the student does not like to interact is not aware of the content of the interaction. If the value of K is smaller than 0, the student is considered to be one of the reasons for dislike of interaction as low enthusiasm for interaction. If the absolute value of K is smaller than a certain threshold, the student is considered to be one of the reasons for the dislike of interaction as weak learning ability. If the absolute value of K is larger than a certain threshold, the student is considered to have weak social ability as one of the reasons for loving interaction. And calculating the student interaction completion K through the student interaction completion model according to the student interaction content text similarity and the emotion analysis and student score ranking of the student interaction content. And judging the reasons of the student loving interaction according to the value of K.
Step 107, based on the interaction information of teachers and students, the tendency degree of students to different interaction modes is obtained.
And obtaining the teacher information from the teacher information database. And taking the mode of initiating interaction by a teacher in each interaction of the student and the time of the interaction as each interaction mark label. And classifying different interactive mark labels according to the interactive mode. And summing the interaction time in the interaction mark labels of different interaction modes to obtain the tendency degree of students to the interaction modes. For example, the modes of lesson interaction are: a question method, an activity method, a game method. The interaction record of the teacher and each student and the time of the interaction are recorded and marked. And respectively calculating the tendency degree of the student for different interaction modes according to the marks of the student in different modes. The interaction type tendency value is the tendency of the student to different degrees of interaction.
Based on the tendency degree of the students to different interaction modes, whether the students tend to be more one interaction type is judged.
And processing the trend degree data based on different trend degrees of different interaction modes of the students to obtain standard data. And analyzing based on the standard data and extracting the characteristics of the interaction tendency of the students by adopting a characteristic extractor. A classifier is employed to determine whether student interaction is too prone to an interaction type based on the extracted interaction-prone features. For example, the modes of lesson interaction are: a question method, an activity method, a game method. The interaction tendency values of the students are respectively 60, 80 and 70 through model statistics. And predicting the interaction tendency of the students by adopting a classifier. In this example, the student's trend values are averaged, so the student is not determined to be too prone to a student of a certain interaction type. If the student has an interaction tendency value of 100, 5 or 10. Since the interaction tendency value of the question asking method of the student is far higher than that of other interaction types. After the data passes through the classifier, the system determines that the student is too inclined to ask questions.
And step 108, establishing a student interaction enthusiasm assessment model based on the interaction behaviors of the students, and obtaining the change of interaction tendency of the students in different interaction modes.
Counting active sponsors of each teacher-student interaction through a high-definition camera in the teaching room; in classDuring hall interaction, video information containing students and a hot spot area diagram of the students are acquired through a camera and an infrared sensor respectively, and body temperature changes of the students during interaction are acquired; calculating the tension degree of the student interaction based on the tension degree model and the body temperature change of the student during the interaction; establishing a student interaction enthusiasm assessment model, wherein E=Q 1 *C-Q 2 *T-Q 3 * N+m, where E is the student interaction aggressiveness,
Q 1 、Q 2 、Q 3 for different weight values, C is the number of times that the student actively initiates the interaction, T is whether the student is stressed or not during the interaction, the stress and relaxation values are respectively 1, 0, -1, N is the stress degree of the student for the current interaction, and M is a constant; counting the interaction information of students after each interaction, and updating the interaction activity degree of the students; if the interaction activity degree of the students changes after different interaction modes are carried out, the interaction tendency of the students is changed. For example: the interaction activity of a student at the last time is 1*0-0.5x1-0.5x0.9+1.1 equal to 0.15. In one interaction, a teacher uses an interaction mode of an activity method to perform activities with the students. The student is still stressed but the stress level is obviously reduced when the student interacts. The student's interaction activity level is 1*0-0.5 x 1-0.5 x 0.3+1.1 equal to 0.45, and the activity level is improved compared with the last time. The student is considered to have a changed tendency to interact in the interactive mode of the activity method.
Step 109, judging the interaction mode suitable for the students according to the different interaction tendency changes of the students under different interactions.
And counting the information of each interaction according to the evaluation model of the interaction enthusiasm of the students. Based on the latest ten interactive enthusiasm of each classmate, the change value of the last enthusiasm of the students in each interactive mode is counted. And adding the enthusiasm change values under each interaction mode to obtain the influence of different interaction modes on the interaction enthusiasm of students. For example: the enthusiasm of the near ten interactions of a student are respectively
0.3, 0.5, 0.1, 0.8, 1.2, 0.6, 0.55, 0.35, 0.45, 0.75. The near ten interaction modes of the student are respectively as follows: a question method, an activity method, a question method, a game method, a question method, an activity method, a question method, and a game method. The enhancement values of each interaction method are respectively: the question asking method comprises the following steps: -0.4-0.6+0.1 equals-0.9, activity method: 0.3-0.05-0.2 is equal to 0.05, game method: 0.7+0.3 equals 1. And summarizing the influence of all the methods on the enthusiasm of the students, and judging that the interaction mode most suitable for the students is a game method.
And acquiring the interaction tendency change of the student after the teacher performs the interaction mode suitable for the student.
The teacher takes an interaction mode suitable for students in class. And calculating the enthusiasm evaluation value of five interactions of the student after the teacher adopts a proper interaction mode according to the evaluation model of the enthusiasm of the student. And judging the interaction tendency change of the students according to the final five-time student enthusiasm evaluation values. For example, after a teacher deliberately adopts an interaction mode suitable for students, the teacher continuously records five interaction information of the students after the students, and calculates each interaction enthusiasm evaluation value of the students by using the student interaction enthusiasm evaluation model. And calculating the average value of the five interactive enthusiasm evaluation values, calculating the average value of the interactive enthusiasm evaluation values of the students before the teacher adopts a proper interactive mode, and comparing the front average value with the rear average value. If the average value of the enthusiasm after the change is higher than the average value of the enthusiasm before the change, the interaction tendency of the students is judged to be improved. Otherwise, the interaction enthusiasm is reduced.
In addition, the specific features described in the above embodiments may be combined in any suitable manner, and in order to avoid unnecessary repetition, various possible combinations are not described further. Moreover, any combination of the various embodiments of the invention can be made without departing from the spirit of the invention, which should also be considered as disclosed herein.
Claims (10)
1. A teacher-student interaction detection method based on an intelligent classroom of the Internet of things is characterized by comprising the following steps:
acquiring interaction information of teachers and students according to visual and auditory information of interaction in a classroom; obtaining the learning ability of the students according to the interaction information of the students; judging the interactive behavior degree of the student based on the learning ability and learning evaluation of the student, specifically comprising: judging the interaction behavior degree of students; the interactive tendency of a teacher is judged according to the learning ability value of the student participating in the interaction and the interactive behavior degree of the student, and the interactive tendency of the teacher is judged according to the learning ability value of the student participating in the interaction and the interactive behavior degree of the student, and the method specifically comprises the following steps: judging the interaction tendency of a teacher, and judging whether the teacher needs to interact with students with different learning abilities according to the interaction tendency of the teacher; predicting the student interaction type according to the sign state and the active interaction duty ratio during student interaction, and specifically comprising the following steps: judging the interaction type of students; the reasons for the lovely interaction of the students are judged through the interaction content of the students, and the reasons for the lovely interaction of the students are judged through the interaction content of the students, specifically comprising: based on the student interaction type and the interaction content, acquiring student interaction audio content, calculating the text similarity of the interaction content of a teacher and a student and emotion analysis of the student interaction content according to the student interaction audio content, constructing a student interaction completion degree model, and judging the reasons that the student does not like interaction; based on interaction information of teachers and students, the tendency degree of the students to different interaction modes is obtained, and the method specifically comprises the following steps: judging whether the student is more prone to an interaction type based on the tendency degree of the student to different interaction modes; establishing a student interaction enthusiasm assessment model based on the interaction behaviors of students, and acquiring the change of interaction tendency of the students in different interaction modes; according to the different interaction tendency changes of students under different interactions, judging the interaction mode suitable for the students comprises the following steps: and acquiring the interaction tendency change of the student after the teacher performs the interaction mode suitable for the student.
2. The method of claim 1, wherein the obtaining the teacher-student interaction information from visual and auditory information of the interaction in the class comprises:
the teacher information and the student information are acquired through a school affair management system; the method comprises the steps of constructing a recognition model of teacher-student interaction based on a machine learning supervised mode recognition method, training the model by using a large amount of action information and audio information of interaction between teachers and students, and judging a mode of the teacher to initiate interaction according to the interaction information between the teachers and the students by the recognition model of the teacher-student interaction; after the approval of the school side and the teacher and student is obtained, real-time identity information of the teacher and the student in the classroom is acquired in real time through cameras at a plurality of angles in the classroom by adopting a face recognition method, and action information during interaction of the teacher and the student is obtained; after the approval of the school side and the teacher and student is obtained, audio information during the interaction of the teacher and the student is obtained through an audio pick-up device in the classroom; transmitting the action information and the audio information to a recognition model of the interaction between the teachers and the students, and judging the mode of the interaction by the recognition model of the interaction between the teachers and the students; and counting the mode of interaction initiation by a teacher in a classroom according to the real-time identity information and the interaction mode, and storing the times and interaction time of participation of students in the interaction into a teacher-student interaction information database.
3. The method of claim 1, wherein the obtaining learning ability of the student according to the interaction information of the student comprises:
the student information includes: usual score, job completion degree; the ordinary score is recorded by a school educational administration system; the degree of completion of the homework is obtained through statistics of a teacher on the homework of the students; firstly, a student learning ability assessment model M=A (N-S) +B.times.H is built based on the usual achievements and the homework completion degrees of students, wherein A, B is the weight of the achievements and the homework completion degrees respectively, S is the ranking of the achievements of the students, N is the number of class persons, and H is the completion degree of the usual homework; the assessment is performed based on a student learning ability assessment model, and the larger the M value is, the stronger the learning ability of the student is.
4. The method of claim 1, wherein the determining the degree of student interaction behavior based on the student learning ability and the learning assessment comprises:
a teacher-student communication meeting is set at regular intervals in a school; the teacher-student interaction refers to the information about deep interaction between the teacher and each student for learning and classroom interaction; the teacher fills in student evaluation of the students according to the communicated content; judging the interaction behavior degree of the students according to the learning ability and learning evaluation of the students; comprising the following steps: judging the interaction behavior degree of students;
The judging of the interactive behavior degree of the student specifically comprises the following steps:
student evaluation comprises classroom interaction behavior, interaction performance and student score; student performance is a ranking of the student's performance at a class; the interaction behavior degree is three, namely, unwilling interaction, general and willing to actively interact; the interactive performance is mainly aimed at students who are unwilling to interact in the degree of interaction behavior, and the reason why the students are unwilling to interact is analyzed.
5. The method of claim 1, wherein the determining the interactive tendency of the teacher based on the learning ability value of the student participating in the interaction and the degree of the student interaction behavior comprises:
counting the learning ability value of students interacted with teachers and the interaction behavior degree of the students; counting the sum of interaction times of a teacher and students; according to the learning ability value of the student and the interaction behavior degree of the student, constructing a model to calculate the interaction tendency degree of the teacher to different types of students; judging whether the teacher has interaction tendency for a certain type of student according to the interaction tendency degree of the teacher; comprising the following steps: judging the interaction tendency of a teacher; judging whether the teacher needs to interact with students with different learning abilities according to the interaction tendency of the teacher;
the judging of the interaction tendency of the teacher specifically comprises the following steps:
Counting the object interacted by teachers and students each time through a high-definition camera in the teaching room; recording the interaction times of teachers and different students based on the interaction of teachers and students each time; counting the sum of interaction times of a teacher and students; classifying and ranking the evaluation of the students according to teacher, and ranking the students with the same interactive behavior degree in the evaluation according to the learning ability of the students; building teacher interaction tendency model X= (C) based on learning ability value of student interacting with teacher and interaction times of student and teacher 1 *M 1 +C 2 *M 2 +...+C n *M n )/(C 1 +C 2 +...+C n ) Wherein n represents n-bit classmates in class, C 1 Representative of student inputLearning ability rank the number of interactions of the first student with teacher in the learning ability value, M 1 Calculating the learning ability value of the student through a student learning ability evaluation model, C 2 Is the interaction times of the students and teachers, M, of which the learning ability values of the students are input and the students of which the learning ability values are ranked second Xi Nengli 2 Calculating the learning ability value of the student through a student learning ability evaluation model, and similarly calculating the average value of the learning ability of all students interacted with a teacher in the input; dividing students into three categories according to the interaction behavior degree of the students; respectively inputting the learning ability values of students with the same interaction behavior degree and the interaction times of the students and teachers into a teacher interaction tendency model X to calculate a Xi Nengli average value; calculating the average value of learning ability of students interacted by teachers according to the teacher interaction tendency model;
Judging whether the teacher needs to interact with students with different learning abilities according to the interaction tendency of the teacher, and specifically comprising the following steps:
defining students whose learning ability values are before a certain threshold value in class as excellent students; calculating an average value A of learning ability values of students, wherein the interaction behavior degree in the student evaluation is willing to actively interact and the class rank is before a certain threshold value; comparing the average value S of the learning ability of the students with which the teacher is interacted with A, if the average value S of the learning ability of the students with which the teacher is interacted is greater than the value A of the learning ability of the students with which the class is ranked before a certain threshold value, considering that the teacher is favored to interact with excellent students, and reminding the teacher to interact with other students with which the teacher is willing to interact but the lower level in the performance by the system; if the interactive behavior degree of the teacher interaction is that the average value S of the learning abilities of the students willing to be active is smaller than the learning ability value A of the students ranked before a certain threshold value, the interactive tendency of the teacher is considered to be relatively average.
6. The method of claim 1, wherein the predicting the student interaction type based on the sign status and the active interaction duty cycle at the student interaction comprises:
the interaction types of students are classified into active type, normal passive type and passive type; the active mode is defined as that the student is willing to actively interact with the teacher, the normal passive mode means that the student does not feel against the interaction with the teacher, but does not actively interact with the teacher, and the passive mode is that the student does not wish to even fear to interact with the teacher; inputting the audio information and the action information of the students during interaction into an identification model to judge whether the students actively initiate the interaction, and judging the interaction type of the students according to the tension degree of the students and the proportion of the active initiation interaction to the total interaction times; comprising the following steps: judging the interaction type of students;
The judging of the student interaction type specifically comprises the following steps:
real-time identity information of teachers and students in the classroom is acquired in real time through cameras with multiple angles in the classroom by adopting a face recognition method, and action information during interaction of the teachers and the students is acquired; acquiring audio information during interaction of teachers and students through an audio pick-up device in the teaching room; transmitting the action information and the audio information to an identification model of the interaction between teachers and students to analyze and judge whether the students actively initiate the interaction; during classroom interaction, video information containing students and a hot spot area diagram of the students are acquired through a camera and an infrared sensor respectively, and body temperature changes of the students during interaction are acquired; identifying the collected video information and the collected hot spot area map to obtain the human body characteristic data of the detected student; constructing a tension degree model N=T based on student human body characteristic data, wherein T is the difference between the body temperature of the student during interaction and the body temperature before interaction; calculating the tension degree of the student interaction by using the change value of the student body temperature; establishing a preset student interaction type analysis model based on the number of active interactions of the students and the tension and tension degree of the interactions, and judging the interaction type of the students according to the ratio of the number of active interactions of the students to the total number of interactions and the tension and tension degree of the interactions of the students; counting the number of times of the active initiation of the interaction and the total interaction number of the students, and setting the interaction type analysis model to judge the interaction type of the students as active if the ratio of the number of times of the active initiation of the interaction of the students to the total interaction number of the students is greater than a certain threshold; and calculating the tension of the student based on the analysis model of whether the student is tension and the tension degree in each interaction type, if the tension degree is lower than a certain threshold value, and the ratio of the number of times the student actively initiates the interaction to the total number of times the student is smaller than a certain threshold value, determining that the student interaction type is normal passive type by the analysis model of the interaction type, otherwise, if the tension degree is higher than a certain threshold value, and the ratio of the number of times the student actively initiates the interaction to the total number of times the student is smaller than a certain threshold value, determining that the student interaction type is passive type by the analysis model of the interaction type.
7. The method of claim 1, wherein the determining the cause of the student's dislike of the interaction by the student's interaction content comprises:
acquiring interactive audio information of teachers and students of passive students; respectively processing the audio information of the teacher and the student to generate interactive audio content; performing text similarity processing on the interactive audio content of the teacher and the student, and judging whether the interactive audio content of the teacher and the student is similar; carrying out emotion analysis on the interactive audio content of the student; finally, judging the reason why students do not like to interact according to the text similarity and emotion analysis results and the student learning score; comprising the following steps: based on the student interaction type and the interaction content, acquiring student interaction audio content; according to the student interaction audio content, calculating the text similarity of the interaction content of a teacher and a student and emotion analysis of the student interaction content; constructing a student interaction completion degree model, and judging the reasons of the student loving interaction;
based on the student interaction type and the interaction content, the method for acquiring the student interaction audio content specifically comprises the following steps:
based on student interaction behavior analysis, screening out passive and passive types of student information; the audio pickup device is used for acquiring interactive audio information of a teacher and students during interaction of passive students, and audio acquired by the audio pickup device is preprocessed; extracting and identifying interactive audio contents of a teacher and a student, and respectively generating the interactive audio contents of the teacher and the student;
According to the student interaction audio content, calculating the text similarity of the interaction content of a teacher and a student and emotion analysis of the student interaction content, specifically comprising:
inputting the content of the teacher and student interaction audio into an N-gram model to perform Chinese word segmentation on the student interaction audio; respectively carrying out text vectorization on the segmented text by using a word2vec model to obtain vectorized representation of each word; each dimension of the vectors is weighted and added to form a text vector, and the similarity of the text can be calculated by using cosine distances; the words in the sentences after the word segmentation are traversed one by one and compared with a BosonNLP emotion dictionary; if the word hits the dictionary, processing corresponding weight is carried out; the positive word weight is addition, the negative word weight is subtraction, the negative word weight is the opposite number, the degree adverb weight is multiplied by the word weight modified by the negative word weight, and the emotion weight value of the interactive content text is obtained by summation;
the construction of the student interaction completion degree model and the judgment of the reasons of the student loving interaction specifically comprise the following steps:
constructing a student interaction completion model K=L.times.M.times.M (T-S) based on the similarity of the student interaction content text and the emotion analysis of the student interaction content and the student score ranking, wherein L is the student interaction content text similarity, if the similarity is higher than a certain threshold value, the L value is 1, otherwise, the L value is 0, M is the emotion analysis weight value of the student interaction content, if the weight value is a non-negative number, the M value is 1, otherwise, the M value is-1, T is the total number of classes +1, and S is the student score ranking; judging the reason why students do not love the interaction based on the student interaction completion degree model; if the value of K is 0, the student is considered to be one of the reasons why the student does not like to interact is not aware of the content of the interaction; if the value of K is smaller than 0, the student is considered to be dislike the reason of interaction is that the enthusiasm for interaction is low; if the absolute value of K is smaller than a certain threshold, the student is considered to be one of the reasons for dislike of interaction as weak learning ability; if the absolute value of K is larger than a certain threshold, the student is considered to have weak social ability as one of the reasons for loving interaction.
8. The method of claim 1, wherein the obtaining the degree of tendency of the student to different interaction modes based on the interaction information of the teacher and the student comprises:
acquiring teacher information from a teacher information database; taking the mode of initiating interaction by a teacher in each interaction of students and the time of the interaction as each interaction mark label; classifying different interactive mark labels according to an interactive mode; summing the interaction time in the interaction mark labels of different interaction modes to obtain the tendency degree of students to the interaction modes; comprising the following steps: judging whether the student is more prone to an interaction type based on the tendency degree of the student to different interaction modes;
based on the tendency degree of the students to different interaction modes, judging whether the students are more prone to an interaction type or not, specifically comprising:
based on different tendency degrees of different interaction modes of students, processing tendency degree data to obtain standard data; analyzing based on standard data and extracting characteristics of student interaction tendency by adopting a characteristic extractor; a classifier is employed to determine whether student interaction is too prone to an interaction type based on the extracted interaction-prone features.
9. The method of claim 1, wherein the establishing a student interaction aggressiveness evaluation model based on the student interaction behavior, obtaining the change of the student interaction tendency under different interaction modes, comprises:
counting active sponsors of each teacher-student interaction through a high-definition camera in the teaching room; during classroom interaction, video information containing students and a hot spot area diagram of the students are acquired through a camera and an infrared sensor respectively, and body temperature changes of the students during interaction are acquired; calculating the tension degree of the student interaction based on the tension degree model and the body temperature change of the student during the interaction; establishing a student interaction enthusiasm assessment model, wherein E=Q 1 *C-Q 2 *T-Q 3 * N+M, wherein E is student interaction aggressiveness, Q 1 、Q 2 、Q 3 For different weight values, C is the number of times that the student actively initiates the interaction, T is whether the student is stressed or not during the interaction, the stress and relaxation values are respectively 1, 0, -1, N is the stress degree of the student for the current interaction, and M is a constant; counting the interaction information of students after each interaction, and updating the interaction activity degree of the students; if the interaction of students takes place actively after different interaction modes are performedAnd if the interaction changes, the interaction tendency of the student is changed.
10. The method of claim 1, wherein the determining the interaction mode suitable for the student according to the change of the different interaction tendencies of the student under different interactions comprises:
counting the information of each interaction according to an evaluation model of the interaction enthusiasm of students; based on the latest ten interactive enthusiasms of each classmate, counting the change value of the last enthusiasm of students in each interactive mode; adding the enthusiasm change values under each interaction mode to obtain the influence of different interaction modes on the interaction enthusiasm of students; comprising the following steps: the method comprises the steps of obtaining interaction tendency change of students after a teacher performs an interaction mode suitable for the students;
the method for obtaining the interaction tendency change of the student after the teacher performs the interaction mode suitable for the student specifically comprises the following steps:
the teacher adopts an interaction mode suitable for students in class; according to the evaluation model of the interaction enthusiasm of the students, calculating the enthusiasm evaluation value of five interactions of the students after the teacher adopts a proper interaction mode; and judging the interaction tendency change of the students according to the final five-time student enthusiasm evaluation values.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310364609.6A CN116579894B (en) | 2023-04-06 | 2023-04-06 | Teacher-student interaction detection method based on intelligent classroom of Internet of things |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310364609.6A CN116579894B (en) | 2023-04-06 | 2023-04-06 | Teacher-student interaction detection method based on intelligent classroom of Internet of things |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116579894A true CN116579894A (en) | 2023-08-11 |
| CN116579894B CN116579894B (en) | 2023-10-24 |
Family
ID=87542159
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310364609.6A Active CN116579894B (en) | 2023-04-06 | 2023-04-06 | Teacher-student interaction detection method based on intelligent classroom of Internet of things |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116579894B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107895244A (en) * | 2017-12-26 | 2018-04-10 | 重庆大争科技有限公司 | Classroom teaching quality assessment method |
| CN108108903A (en) * | 2017-12-26 | 2018-06-01 | 重庆大争科技有限公司 | Classroom teaching quality assessment system |
| CN110782185A (en) * | 2019-11-09 | 2020-02-11 | 上海光数信息科技有限公司 | Classroom behavior recognition and analysis method |
| CN110991381A (en) * | 2019-12-12 | 2020-04-10 | 山东大学 | Real-time classroom student state analysis and indication reminding system and method based on behavior and voice intelligent recognition |
| CN111709358A (en) * | 2020-06-14 | 2020-09-25 | 东南大学 | Teacher-student behavior analysis system based on classroom video |
| CN112702416A (en) * | 2020-12-21 | 2021-04-23 | 泰康保险集团股份有限公司 | Interactive method and device for online teaching |
| CN115545587A (en) * | 2022-12-05 | 2022-12-30 | 成都大学 | Classroom teaching scheme evaluation method and system based on multiple agents |
-
2023
- 2023-04-06 CN CN202310364609.6A patent/CN116579894B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107895244A (en) * | 2017-12-26 | 2018-04-10 | 重庆大争科技有限公司 | Classroom teaching quality assessment method |
| CN108108903A (en) * | 2017-12-26 | 2018-06-01 | 重庆大争科技有限公司 | Classroom teaching quality assessment system |
| CN110782185A (en) * | 2019-11-09 | 2020-02-11 | 上海光数信息科技有限公司 | Classroom behavior recognition and analysis method |
| CN110991381A (en) * | 2019-12-12 | 2020-04-10 | 山东大学 | Real-time classroom student state analysis and indication reminding system and method based on behavior and voice intelligent recognition |
| CN111709358A (en) * | 2020-06-14 | 2020-09-25 | 东南大学 | Teacher-student behavior analysis system based on classroom video |
| CN112702416A (en) * | 2020-12-21 | 2021-04-23 | 泰康保险集团股份有限公司 | Interactive method and device for online teaching |
| CN115545587A (en) * | 2022-12-05 | 2022-12-30 | 成都大学 | Classroom teaching scheme evaluation method and system based on multiple agents |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116579894B (en) | 2023-10-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111291613B (en) | Classroom performance evaluation method and system | |
| Inventado et al. | Predicting student emotions resulting from appraisal of ITS feedback | |
| CN112085392A (en) | Learning participation degree determining method and device and computer equipment | |
| Setiyowati et al. | Analyzing classroom diversity and its contribution to multicultural education in Indonesia | |
| Rzepka et al. | Fairness of In-session Dropout Prediction. | |
| Ryumina et al. | OCEAN-AI framework with EmoFormer cross-hemiface attention approach for personality traits assessment | |
| Kumano et al. | Multitask item response models for response bias removal from affective ratings | |
| Mastorodimos et al. | Studying affective tutoring systems for mathematical concepts | |
| Arroyo et al. | Bayesian networks and linear regression models of students’ goals, moods, and emotions | |
| CN116579894B (en) | Teacher-student interaction detection method based on intelligent classroom of Internet of things | |
| Suparwito | Factors influencing the difficulty level of the subject: Machine learning technique approaches | |
| CN115795015A (en) | Comprehensive knowledge tracking method for enhancing test question difficulty | |
| Gambo et al. | A conceptual framework for detection of learning style from facial expressions using convolutional neural network | |
| Lee et al. | Prescribing deep attentive score prediction attracts improved student engagement | |
| CN113378076A (en) | Online education-oriented learner collaborative learning social relationship construction method | |
| Abdellaoui et al. | Emotion Detection and Student Engagement in Distance Learning During Containment Due to the COVID-19 | |
| Musso et al. | New approaches for improved quality in educational assessments: Using automated predictive systems in reading and mathematics | |
| Shenoy et al. | A Study on various Applications of Computer Vision for Teaching Learning in Classroom | |
| Šošević et al. | Support platform for learning about multimodal biometrics | |
| CN111209817A (en) | Assessment method, device and equipment based on artificial intelligence and readable storage medium | |
| Hocine | Feedback and scenario adaptation of serious games for job interview training | |
| Haglind | Coping with success and failure–A qualitative study on athletes and coaches in track and field | |
| Qadir et al. | Personality Prediction Using Sklearn | |
| Kaur et al. | Nuanced Emotion State Identification: Insights from a CNN-RF Framework | |
| Karunasekara et al. | A mobile gamified application to improve mathematics skills of students from age 9 to 11 years |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |