A student engagement calculation method, system and readable storage medium

Abstract

The application discloses a student participation calculation method and system and a readable storage medium, and relates to the technical field of education. The method comprises the following steps: determining a first sampling time according to a preset first sampling strategy; analyzing teaching process data, identifying a preset student behavior, and statistically obtaining a student behavior distribution for each first sampling time; determining a second sampling time according to a preset second sampling strategy; analyzing teaching process data for each second sampling time; determining a classroom scene category according to the distribution of various classroom activity parameters in a preset time period before and after the second sampling time; generating a classroom scene sequence of the entire teaching process according to the classroom scene category; determining a participation weight table according to the consistency of the preset student behavior and the classroom scene; and calculating a student participation value at each classroom time according to the participation weight table, the student behavior distribution and the classroom scene sequence. The application can reduce the misjudgment of student participation and improve the accuracy of the student participation result.

Description

Technical Field

[0001] This application relates to the field of educational technology, and in particular to a method, system, and readable storage medium for calculating student engagement. Background Technology

[0002] In modern educational settings, student classroom participation is widely considered an important indicator of classroom teaching effectiveness. With the development of educational technology, teachers and researchers have begun to explore various tools and methods to quantitatively measure and evaluate student classroom participation.

[0003] In related technologies, for example, Chinese invention patent with application number CN202210482804.4, entitled "A Method for Analyzing Classroom Student Participation," discloses a "Method for Calculating Student Participation Based on Classroom Behavior Analysis," which includes the following steps: acquiring classroom teaching video data; obtaining recognition results: identifying student behavior based on the classroom teaching video to obtain recognition results; classifying behavior: classifying the recognition results to obtain classification results, wherein the categories of student behavior include: listening, reading and writing, student interaction, raising hands, and responding; calculating weight ratio: assigning weight values ​​to student behavior according to the degree of influence of student behavior on student participation, and calculating the weight ratio of student behavior in a single recognition result; calculating participation: calculating participation based on the classification results.

[0004] However, the methods for analyzing classroom student engagement provided by related technologies have a significant problem: they do not explicitly consider the type of classroom scenario. In reality, student behavior and engagement are directly related to the type of classroom activity. For example, actively speaking in a teacher-student discussion might be considered active participation, while the same speaking behavior might be considered disruptive during a teacher lecture. Therefore, these technologies do not take into account the impact of the classroom scenario on student engagement, and their methods for analyzing classroom student engagement may lead to misjudgments of student engagement, thus affecting the accuracy of the results. Summary of the Invention

[0005] This application provides a method, system, and readable storage medium for calculating student participation, which reduces misjudgments of student participation and improves the accuracy of student participation results.

[0006] Firstly, this application provides a method for calculating student participation, including: The first sampling time is determined according to the preset first sampling strategy. For each first sampling time, the teaching process data is analyzed, the preset student behaviors are identified, and the distribution of student behaviors is statistically obtained. The second sampling time is determined according to the preset second sampling strategy. For each second sampling time, the teaching process data is analyzed, and the classroom scene category is determined according to the distribution of various classroom activity parameters within the preset time period before and after the second sampling time. The classroom scene sequence of the entire teaching process is generated according to the classroom scene category. The participation weighting table is determined based on the conformity between the pre-set student behavior and the classroom scenario; The student participation value for each classroom moment is calculated based on the participation weight table, student behavior distribution, and classroom scene sequence.

[0007] In the above embodiments, by analyzing teaching process data and classroom activity parameters, the correlation between classroom scenario categories and student participation is explicitly considered. By generating a sequence of classroom scenarios for the entire teaching process and determining a participation weight table based on the conformity of preset student behavior with classroom scenarios, the actual student participation in different classroom scenarios can be reflected more accurately. Furthermore, since this invention can calculate student participation values ​​at each classroom moment, it reduces misjudgments of student participation and improves the accuracy of student participation results.

[0008] In conjunction with some embodiments of the first aspect, in some embodiments, the first sampling strategy and the second sampling strategy are the same.

[0009] In the above embodiments, using the same first sampling strategy and second sampling strategy can reduce the complexity of data processing and improve efficiency.

[0010] In conjunction with some embodiments of the first aspect, in some embodiments, the steps of determining the first sampling time according to a preset first sampling strategy, and for each first sampling time, analyzing teaching process data, identifying preset student behaviors, and statistically obtaining the distribution of student behaviors specifically include: The first sampling time is determined according to the preset first sampling strategy; For each first sampling moment, analyze the student area screen in the teaching process data and identify the preset student behaviors in the student area screen; The credibility of each student behavior is calculated based on a preset credibility algorithm. Extract preset student behaviors with a credibility level greater than a preset credibility threshold; The distribution of student behavior at each first sampling time is statistically analyzed.

[0011] In the above embodiments, a preset first sampling strategy is used to determine each first sampling time, and the data of the teaching process is analyzed at each sampling time to identify preset student behaviors in the student area screen during the teaching process. A preset credibility algorithm is used to calculate the credibility of each preset student behavior, and only preset student behaviors with a credibility greater than a preset credibility threshold are extracted. This not only improves the accuracy of student behavior identification but also reduces the impact of misidentification on student participation calculation.

[0012] In conjunction with some embodiments of the first aspect, in some embodiments, classroom activity parameters include teacher behavior, student behavior, courseware behavior, and classroom sound; classroom sound includes one or more of the following: teacher sound, individual student sound, student discussion sound, student chorus sound, and silence.

[0013] In the above embodiments, this comprehensive and detailed parameter definition can capture the characteristics of classroom activities more deeply and comprehensively, thereby improving the accuracy of student participation calculation.

[0014] In conjunction with some embodiments of the first aspect, in some embodiments, the classroom scenario categories include two or more of the following: teacher lecturing, courseware presentation, teacher-student Q&A, student practice, student-student interaction, student presentations, and classroom organization.

[0015] In the above embodiments, this comprehensive and detailed parameter definition can capture the characteristics of classroom scene categories more deeply and comprehensively, thereby improving the accuracy of student participation calculation.

[0016] In conjunction with some embodiments of the first aspect, in some embodiments, the step of determining the second sampling time according to a preset second sampling strategy, and for each second sampling time, analyzing teaching process data, and determining the classroom scenario category based on the distribution of various classroom activity parameters within a preset time period before and after the second sampling time, specifically includes: The second sampling time is determined according to the preset second sampling strategy; For each second sampling moment, analyze the teaching process data, identify classroom activity parameters, and generate a classroom activity parameter dataset; for the current second sampling moment, determine several time windows containing the current second sampling moment according to the preset time window selection strategy; the current second sampling moment can be any second sampling moment. For the current time window, the distribution of classroom activity parameters is statistically analyzed based on the classroom activity parameter dataset, and the characteristic values ​​of classroom activities are calculated. The current time window can be any time window. The classroom scene category at the current second sampling moment is determined based on the characteristic values ​​of classroom activities.

[0017] In the above embodiments, by using a preset second sampling strategy and a time window selection strategy, a large amount of teaching process data can be collected and analyzed, including courseware images, podium area images, student area images, and classroom audio, thereby generating a more detailed and accurate dataset of classroom activity parameters. Furthermore, it analyzes not only the distribution of classroom activity parameters at a single second sampling moment, but also the distribution of classroom activity parameters within a preset time period before and after it. This avoids errors caused by single moments and enhances the comprehensiveness of classroom scene categories.

[0018] In conjunction with some embodiments of the first aspect, in some embodiments, the step of determining the second sampling time according to a preset second sampling strategy, and for each second sampling time, analyzing teaching process data, and determining the classroom scenario category based on the distribution of various classroom activity parameters within a preset time period before and after the second sampling time, specifically includes: The second sampling time is determined according to the preset second sampling strategy; For each second sampling moment, the teaching process data within a preset time period before and after the second sampling moment is input into a preset classroom scene classification model to obtain the classroom scene category. The training samples of the classroom scene classification model are historical teaching process data. The classroom scene category of each classroom moment in the training samples is determined according to the distribution of classroom activity parameters within a certain period before and after that classroom moment.

[0019] In the above embodiments, this model-based classroom scenario classification method can greatly improve the processing efficiency and accuracy of teaching process data.

[0020] Secondly, embodiments of this application provide a student participation calculation system, including a computer, the computer comprising: The student behavior distribution module is used to determine the first sampling time according to the preset first sampling strategy. For each first sampling time, it analyzes the teaching process data, identifies the preset student behaviors, and statistically obtains the student behavior distribution. The classroom scenario category module is used to determine the second sampling time according to the preset second sampling strategy. For each second sampling time, the teaching process data is analyzed, and the classroom scenario category is determined according to the distribution of various classroom activity parameters within the preset time period before and after the second sampling time. The classroom scene sequence module is used to generate a sequence of classroom scenes for the entire teaching process based on the classroom scene category. The participation weighting table module is used to determine the participation weighting table based on the conformity between preset student behavior and classroom scenario; The student participation value module is used to calculate the student participation value for each classroom moment based on the participation weight table, student behavior distribution, and classroom scene sequence.

[0021] In conjunction with some embodiments of the second aspect, in some embodiments, the first sampling strategy and the second sampling strategy are the same.

[0022] In conjunction with some embodiments of the second aspect, in some embodiments, the student behavior distribution module specifically includes: The first sampling strategy submodule is used to determine the first sampling time according to the preset first sampling strategy; The preset student behavior submodule is used to analyze the student area screen in the teaching process data for each first sampling time and identify the preset student behavior in the student area screen; The credibility submodule is used to calculate the credibility of each preset student behavior according to a preset credibility algorithm; The credibility threshold submodule is used to extract preset student behaviors with credibility values ​​greater than a preset credibility threshold. The student behavior distribution submodule is used to statistically analyze the student behavior distribution at each first sampling time.

[0023] In conjunction with some embodiments of the second aspect, in some embodiments, classroom activity parameters include teacher behavior, student behavior, courseware behavior, and classroom sound; classroom sound includes one or more of the following: teacher sound, individual student sound, student discussion sound, student chorus sound, and silence.

[0024] In conjunction with some embodiments of the second aspect, in some embodiments, the classroom scenario categories include two or more of the following: teacher lecturing, courseware presentation, teacher-student Q&A, student practice, student-student interaction, student presentations, and classroom organization.

[0025] In conjunction with some embodiments of the second aspect, in some embodiments, the classroom scenario category module specifically includes: The second sampling time submodule is used to determine the second sampling time according to the preset second sampling strategy; The teaching process data submodule is used to analyze teaching process data, identify classroom activity parameters, and generate a classroom activity parameter dataset for each second sampling time. The time window submodule is used to determine, based on a preset time window selection strategy, several time windows that include the current second sampling moment; the current second sampling moment can be any second sampling moment. The Classroom Activity Feature Value Submodule is used to calculate the feature value of classroom activities for the current time window by statistically analyzing the distribution of classroom activity parameters based on the classroom activity parameter dataset. The current time window can be any time window. The classroom scene category submodule is used to determine the classroom scene category at the current second sampling moment based on the feature values ​​of classroom activities.

[0026] In conjunction with some embodiments of the second aspect, in some embodiments, the classroom scenario category module specifically includes: The second sampling time submodule is used to determine the second sampling time according to the preset second sampling strategy; The classroom scene classification model submodule is used to input the teaching process data within a preset time period before and after the second sampling time into the preset classroom scene classification model for each second sampling time to obtain the classroom scene category. The training samples of the classroom scene classification model are historical teaching process data, and the classroom scene category of each classroom moment in the training samples is determined according to the distribution of classroom activity parameters within a certain period before and after that classroom moment.

[0027] Thirdly, embodiments of this application provide a student engagement calculation system, which includes: one or more processors and a memory; The memory is coupled to the one or more processors and is used to store computer program code, which includes computer instructions that the one or more processors call to cause the student engagement calculation system to perform the methods described in the first aspect and any possible implementation thereof.

[0028] Fourthly, embodiments of this application provide a computer program product containing instructions that, when the computer program product is run on a system, cause the system to perform the method described in the first aspect and any possible implementation thereof.

[0029] Fifthly, embodiments of this application provide a computer-readable storage medium including instructions that, when executed on a student engagement calculation system, cause the student engagement calculation system to perform the method described in the first aspect and any possible implementation thereof.

[0030] Understandably, the student participation calculation system provided in the second aspect, the third aspect, the fourth aspect, and the fifth aspect are all used to execute the student participation calculation method provided in the embodiments of this application. Therefore, the beneficial effects they can achieve can be referred to the beneficial effects in the corresponding methods, and will not be repeated here.

[0031] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages: 1. The student participation calculation method provided in this application, through analysis of teaching process data and classroom activity parameters, explicitly considers the correlation between classroom scenario categories and student participation. By generating a sequence of classroom scenarios for the entire teaching process and determining a participation weight table based on the pre-set conformity of student behavior with classroom scenarios, it can more accurately reflect the actual student participation in different classroom scenarios. Furthermore, since this invention can calculate the student participation value at each classroom moment, it reduces misjudgments of student participation and improves the accuracy of student participation results.

[0032] 2. The student engagement calculation method provided in this application determines each first sampling time using a preset first sampling strategy, analyzes the teaching process data at each sampling time, and identifies preset student behaviors in the student area screen during the teaching process. A preset credibility algorithm is used to calculate the credibility of each preset student behavior, and only preset student behaviors with a credibility greater than a preset credibility threshold are extracted. This not only improves the accuracy of student behavior identification but also reduces the impact of misidentification on student engagement calculation.

[0033] 3. The student participation calculation method provided in this application, through the use of a preset second sampling strategy and time window selection strategy, can collect and analyze a large amount of teaching process data, including courseware images, podium area images, student area images, and classroom audio, thereby generating a more detailed and accurate dataset of classroom activity parameters. Furthermore, it analyzes not only the distribution of classroom activity parameters at a single second sampling moment but also the distribution of classroom activity parameters within preset time periods before and after it, thus avoiding errors caused by single moments and enhancing the accuracy of classroom scene category identification. Attached Figure Description

[0034] Figure 1 A flowchart illustrating the method for displaying classroom learning information provided in this application.

[0035] Figure 2 This is a schematic diagram illustrating the method for displaying classroom learning information provided in this application.

[0036] Figure 3 for Figure 2 A magnified view of a portion of the image.

[0037] Figure 4 for Figure 2 Another enlarged view of a specific area.

[0038] Figure 5 Another schematic diagram illustrating the method for displaying classroom learning provided in this application.

[0039] Figure 6 Another schematic diagram illustrating the method for displaying classroom learning provided in this application.

[0040] Figure 7 Another schematic diagram illustrating the method for displaying classroom learning provided in this application.

[0041] Figure 8 A flowchart illustrating the method for generating the student behavior distribution map provided in this application.

[0042] Figure 9 A flowchart illustrating the method for calculating student participation provided in this application.

[0043] Figure 10 A schematic diagram of a modular virtual device for calculating student engagement provided in this application.

[0044] Figure 11 A schematic diagram of the physical device of the student participation calculation system provided in this application. Detailed Implementation

[0045] The terminology used in the following embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. As used in the specification and appended claims of this application, the singular expressions “a,” “an,” “the,” “the,” “the,” and “this” are intended to include the plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in this application refers to any or all possible combinations including one or more of the listed items.

[0046] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying or suggesting relative importance or implicitly indicating the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature, and in the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more.

[0047] Since the embodiments of this application involve the application of classroom learning, for ease of understanding, the relevant terms and concepts involved in the embodiments of this application will be introduced below.

[0048] (1) Classroom learning situation is generally the situation of teaching and learning between teachers and students in the classroom teaching process. Correspondingly, classroom learning situation data is data about teaching characteristics. In this application, classroom learning situation data includes learning situation indicators, classroom scenarios and student behavior distribution. The above examples are only for adaptive explanation and are not limited here.

[0049] (2) Learning indicators are a comprehensive overview of learning in the teaching process. There may be one or more learning indicators. Each learning indicator can describe one aspect of the learning characteristics of the teaching process. For example, it may be one or more of the following: student participation, student activity, and consistency of student behavior. The above examples are just for illustrative purposes and are not limited here.

[0050] (3) Classroom scene is an important concept in the classroom teaching process. During the teaching process, similar teaching activities will appear within a certain period of time. For example, the teacher speaks continuously within a period of time, or there are multiple exchanges of words between teachers and students within a period of time, or the students are very noisy within a period of time. Dividing continuous teaching activities with similar characteristics into segments allows for large-scale, meaningful slices of the teaching process. These segments are called classroom scenes. In short, a classroom scene is a slice of the teaching process based on the characteristics of teaching activities. It is a large-scale, meaningful slice of the teaching process. Correspondingly, the entire teaching process can be divided into a sequence of classroom scenes. Typical classroom scenes include teacher lecturing, presentation slides, teacher-student Q&A, student-student interaction, student presentations, and student practice. Teacher lecturing refers to teaching activities in which the teacher mainly explains knowledge points through words. Presentation slides refer to teaching activities in which knowledge points are mainly explained through presentation slides. Teacher-student Q&A refers to teaching activities in which the teacher and one or more students ask and answer questions. Student-student interaction refers to teaching activities in which students discuss problems. Student presentations refer to teaching activities in which students demonstrate to other students on the podium. Student practice refers to teaching activities that involve students' independent learning. In practical applications, there can be more or fewer types of classroom scenarios; the examples above are merely illustrative and are not intended to limit the scope.

[0051] (4) Student behavior refers to the various activities or reactions exhibited by students in the learning environment. Examples include raising hands, standing, listening, reading, writing, turning one's back (towards the podium), etc. The above examples are only for illustrative purposes and are not limited here. The distribution of student behavior refers to the number of students exhibiting each type of behavior and their proportion, which is usually derived by collecting and analyzing a large amount of behavioral data.

[0052] The subject of this application can be a terminal device that runs software for analyzing classroom learning, software for displaying classroom learning, or opens a webpage displaying classroom learning, such as a mobile phone, tablet, desktop computer, smart TV, etc., without limitation.

[0053] The method for displaying classroom learning presented in this application does not limit the specific technical means of the UI (user interface). It can be any implementation method. Common UI implementation methods include mobile terminal APP, web page, client software, etc., which are not limited here.

[0054] The following describes the method for displaying classroom learning in this embodiment: like Figure 1 As shown, Figure 1 A flowchart illustrating the method for displaying classroom learning information provided in this application.

[0055] S101. Display the classroom learning situation display screen based on learning indicators, classroom scene and student behavior distribution. The classroom learning situation display screen includes the first area, the second area, the third area and the indicator line.

[0056] From another perspective, classroom learning data includes learning indicators, classroom scenarios, and student behavior distribution, and therefore, classroom learning data can also be used to display classroom learning information.

[0057] As an example, the basic processing procedure for classroom learning data involves collecting classroom teaching process data, with classroom audio and video being the most commonly used teaching process data; analyzing the teaching process data to extract classroom metadata; and calculating the final classroom learning data based on the classroom metadata using a comprehensive algorithm. It should be noted that this application does not limit the technical solutions for processing and calculating classroom learning data; any technical solution for processing and calculating classroom learning data can be used.

[0058] In order to collect teaching process data, this application provides a teaching process data collection system, the system including: a teacher camera, used to collect images of the podium area; Student cameras are used to capture footage of the student area. Sound pickup equipment, used to collect classroom audio; Courseware capture equipment, used to capture courseware images; Video recording equipment is used to record and collect classroom audio and video data.

[0059] It should be noted that the above-mentioned devices can be independent or integrated. For example, student cameras, courseware acquisition devices, and recording devices can be integrated into one device.

[0060] It should be noted that the classroom audio and video data includes the screen of the podium area, the screen of the student area, the classroom sound, and the screen of the courseware.

[0061] Teaching process data is defined as including the screen display of the lecture area, the screen display of the student area, classroom audio, and courseware visuals. However, this does not mean that teaching process data must include or is limited to these elements. These examples are merely to illustrate possible data types and do not represent the only or complete definition. This embodiment only provides an exemplary description of possible data types and does not impose specific limitations on them.

[0062] The hardware and usage of the teaching process data acquisition system are already very mature in the field, so they will not be elaborated here.

[0063] refer to Figure 2 , Figure 2 This is a schematic diagram illustrating the method for displaying classroom learning information provided in this application.

[0064] The classroom learning display screen includes a first area, a second area, a third area, and indicator lines. Although the screen may contain other areas, the first, second, and third areas, along with the indicator lines, are essential components. These three areas, as well as any other areas that may exist but are not mentioned, will be explained in detail in the following description; they are not limited here.

[0065] S102. Display the learning indicator chart in the first area. The learning indicator chart is a graphical representation of the changes in learning indicators over time. The important time points and periods in the teaching process can be located based on the high points, low points, or changes in the learning indicators.

[0066] Specifically, the learning performance indicator chart includes one or more learning performance indicator curves. Based on the high points, low points, or changes in the learning performance indicator curves, important time points or periods in the teaching process can be located. The learning performance indicator curves represent the changes in learning performance indicators over time in the form of curves.

[0067] As can be seen, the learning performance indicator curve visually illustrates the changes in learning performance indicators over time, allowing teachers to clearly see the trends in teaching effectiveness and improving the visibility of classroom learning. Furthermore, based on the highs, lows, or variations in the learning performance indicator curve, important moments or periods in the teaching process can be identified, thereby improving the efficiency of teaching research activities.

[0068] Specifically, the learning performance indicator curves are one or more of the following: student participation curve, student activity curve, and student behavior consistency curve. The student participation curve reflects the degree of student participation in teaching activities at each class moment. The student participation value for each class moment is calculated based on the consistency between student behavior and the classroom scenario, and then the student participation curve is generated based on the student participation value. The student activity curve reflects the level of student activity at each class moment. The student behavior consistency curve reflects the degree of consistency of student behavior at each class moment.

[0069] This embodiment calculates student engagement scores for each classroom moment by analyzing the consistency between the behavior of all or some students and the classroom scenario. The consistency between student behavior and the classroom scenario refers to the degree of participation of various student behaviors relative to the current classroom environment. For example, if the current classroom scenario is a teacher lecturing, then "listening" is consistent, while "turning one's back" (students facing away from the teacher) is not. Therefore, a score can be calculated for each student at the current classroom moment based on the consistency of student behavior and the classroom scenario. The student engagement score for the current classroom moment is calculated by summing the scores of all or some students. Then, a student engagement curve is generated based on the student engagement scores for all classroom moments. The student activity curve reflects the activity level of students at each classroom moment. This embodiment calculates student activity by analyzing parameters such as student volume, changes in student behavior, and the amount of head movement at each classroom moment. The student behavior consistency curve is plotted based on the number of students exhibiting the most common type of behavior at each classroom moment, reflecting the overall consistency of student actions in the classroom.

[0070] Specifically, classroom sound analysis techniques, such as speech recognition or sound detection, can be used to quantify student activity. For example, the frequency, volume, and duration of student speech can all serve as indicators of activity. Machine learning or artificial intelligence techniques, such as computer vision, can be used to analyze changes in student behavior. This may include recognizing and analyzing student gestures, facial expressions, and posture to assess their engagement and activity. Cameras or other motion capture devices can be used to record student head movements. By analyzing the speed, direction, and frequency of these movements, student activity can be determined. The above-described solutions are merely illustrative examples illustrating possible methods for calculating student activity by analyzing classroom behavior data. In reality, these technologies are already quite mature and will not be elaborated upon further here.

[0071] Specifically, the number of students exhibiting various behaviors is counted at each class moment. The student behavior with the most participants is identified, and the number of participants exhibiting that behavior is divided by the total number of students to obtain a consistency ratio. Over time, these ratios are connected to form a curve, resulting in a student behavior consistency curve. Of course, student behavior recognition technology has reached a fairly mature level in related technologies, so it will not be elaborated on here.

[0072] It is evident that by using student participation curves, student activity curves, and student behavior consistency curves, teachers can intuitively see the changes in students' participation, activity, and behavior consistency at various points in the teaching process. This allows for a better understanding and analysis of students' behavioral patterns and characteristics in the classroom, and helps teachers efficiently review classroom videos, thereby improving the efficiency of teaching and research activities.

[0073] An exemplary embodiment, referenced Figure 2 , Figure 2 This is a schematic diagram illustrating the classroom learning situation display method provided in this application. The learning situation indicator graph includes two learning situation indicator curves—a student participation curve and a student activity curve. Based on these two learning situation curves, the changes and highs and lows of student participation and activity during the teaching process can be observed. Among them, the student participation curve reflects the degree of student participation in teaching activities at each classroom moment.

[0074] It should be noted that the learning indicator graph is defined as having two learning indicator curves—a student participation curve and a student activity curve. However, this does not mean that the learning indicator graph is limited to these elements. These examples are only to illustrate possible learning indicators and do not represent the only or complete definition. In other words, the scope of learning indicators can be broader; this embodiment only provides an exemplary description of possible learning indicators and does not make specific limitations on them. Learning indicator curves can be fewer or more; in another embodiment, the learning indicator graph contains only one student participation curve, while in another embodiment, reference... Figure 5 , Figure 5 This is another schematic diagram illustrating the classroom learning situation display method provided in this application. The learning situation indicator graph includes three curves: student participation curve, student activity curve, and student behavior consistency curve.

[0075] Figure 2 , Figure 5 The learning performance indicator chart in the illustrated embodiment uses a curve presentation method, which can effectively show the high points, low points, and changes in the learning performance indicators. In another embodiment, the learning performance indicator chart is composed of discrete points and / or discrete line segments. Therefore, this application does not limit the presentation method of the learning performance indicator chart.

[0076] S103. In the second area, a classroom scene diagram is displayed. The classroom scene diagram is a graphical representation of the sequence of classroom scenes changing over time during the teaching process. A classroom scene is a slice of the teaching process based on the characteristics of the teaching activities.

[0077] Specifically, the presentation format of the classroom scene diagram is as follows: a total segment composed of one or more sub-segments pieced together in sequence, each sub-segment corresponding to a classroom scene, the length of the sub-segment being proportional to the duration of the corresponding classroom scene, and sub-segments of different types of classroom scenes being drawn in different colors.

[0078] It is evident that this visualization method allows teachers to clearly see the distribution and duration of various classroom scenarios, which helps guide teachers to efficiently review classroom videos and improve the efficiency of teaching and research activities.

[0079] An exemplary embodiment, referenced Figure 3 , Figure 3 for Figure 2 The image shows a magnified view of a classroom scene and color descriptions for each scene. In this embodiment, there are six classroom scenes (located at the bottom of the image): teacher lecturing, presentation slides, teacher-student Q&A, student-student interaction, student presentations, and student practice. The teaching process is divided into a sequence of classroom scenes. Each classroom scene image is composed of one or more segments pieced together sequentially. Each segment corresponds to a classroom scene, and the length of the segment is proportional to the duration of the corresponding classroom scene. Different types of classroom scene segments are drawn in different colors.

[0080] The classroom scene diagram visually presents the sequence of classroom scenes throughout the entire teaching process. This application does not limit the technical solutions for identifying classroom scenes, the types of classroom scenes, or the graphical form of the classroom scene sequence. For example, in another implementation example, there are seven classroom scenes. In addition to the six classroom scenes mentioned above, there is also classroom organization, characterized by the teacher organizing and guiding students in learning activities, rather than explaining knowledge points.

[0081] It should be noted that the number and types of classroom scenes presented in the classroom scene illustrations are determined by the specific actual situation and are not limited here.

[0082] S104. Display a student behavior distribution map in the third area. The student behavior distribution map is a graphical representation of how student behavior distribution changes over time.

[0083] An exemplary embodiment, referenced Figure 4 , Figure 4 for Figure 2 A magnified view of a portion of the image; exist Figure 2 , Figure 4 In the embodiment shown, the student behavior distribution map is presented in a two-dimensional format. Each vertical line in the two-dimensional map corresponds to a classroom moment. Each vertical line contains one or more first sub-segments. Each first sub-segment corresponds to a student behavior in the classroom moment. The first sub-segments of different student behaviors are drawn in different colors. The length of the first sub-segment is proportional to the number of students whose corresponding behavior is shown. In another exemplary embodiment, each horizontal line of the two-dimensional graph corresponds to each class moment, and the horizontal line contains one or more second segments. Each second segment corresponds to a student behavior in the class moment, and the second segments of different student behaviors are drawn with different colors. The length of the second segment is proportional to the number of students whose corresponding student behaviors are represented.

[0084] Student behavior recognition technology has reached a fairly mature level, and will not be elaborated upon here. (Reference) Figure 4Student behavior includes six categories: raising hands, standing, listening, reading and writing (a combination of reading and writing), turning one's back, and other behaviors (behaviors other than raising hands, standing, listening, reading and writing, and turning one's back). Figure 4 The color corresponding to the name of a student's behavior is the color used to draw that behavior. Throughout the entire classroom teaching process, the number of students engaging in various behaviors reflects a detailed characteristic of the teaching process. The amount of data is large, but by displaying this large-scale student behavior data graphically, teachers can intuitively view student behavior in the classroom and pinpoint points or time periods of interest based on the length of different colored line segments in each column (or row) and the area of ​​various colored blocks.

[0085] It is evident that this visualization method can help teachers see the changing trends and distribution of student behavior, enabling them to better understand the characteristics and patterns of student behavior in the classroom. This also helps guide teachers to efficiently review classroom videos and improve the efficiency of teaching and research activities.

[0086] S105. The indicator lines are movable and are used to indicate information in the first, second, or third area to interpret the teaching process.

[0087] Preferably, the indicator lines simultaneously indicate the information of the first, second, and third regions at the same time, which is used to interpret the teaching process by combining the information of the first, second, and third regions.

[0088] Preferably, the indicator lines synchronously indicate information from the first, second, and third areas corresponding to the same class time.

[0089] It is evident that the synchronized indication of the indicator lines allows for the rapid acquisition of information from different areas at the same time, significantly improving the efficiency of integrating and processing student behavior distribution, classroom scene sequences, and learning indicators. Furthermore, the combined information from the first, second, and third areas provides a comprehensive and in-depth perspective on the teaching process, enabling simultaneous interpretation from multiple angles. This effectively guides teachers to efficiently review classroom videos, thereby enhancing the efficiency of teaching research activities.

[0090] The classroom learning display method of this application includes at least one indicator line, which can be moved to indicate the information in the first area, the second area, and the third area for interpreting the teaching process.

[0091] In other embodiments, there is more than one indicator line that can move synchronously. That is, if one indicator line moves, the other indicator lines will follow, thus providing synchronous indication of information in the first, second, and third regions, allowing users to interpret the teaching process by combining the information from the three regions.

[0092] In other embodiments, the first region, the second region, and the third region are arranged vertically in an upper, middle, and lower layout. The indicator line passes through the first, second, and third areas simultaneously, enabling synchronous indication of information in the first, second, and third areas.

[0093] It is evident that the vertical layout and synchronized indicator lines provide an effective method for information exchange. By moving the indicator lines, information from the three areas at different times can be quickly obtained and compared, thereby effectively guiding teachers to efficiently review classroom videos and improving the efficiency of teaching and research activities.

[0094] In other embodiments, the first, second, and third regions are arranged side by side on the left, center, and right. This embodiment has three indicator lines, with each region containing one indicator line. Moving the indicator line of any one region will cause the indicator lines of the other two regions to move synchronously, thus achieving synchronous indication of information in the three regions.

[0095] It should be emphasized that the two embodiments described above are merely exemplary descriptions of the layout of the first, second, and third areas. These examples do not impose specific limitations on the layout of the areas. The actual layout can be adjusted according to specific needs to most effectively meet the requirements of teaching and data display, and no limitations are imposed here.

[0096] In other embodiments, reference is made to Figure 2 The first area also includes an information display area, which summarizes and displays classroom learning data for the three areas indicated by the indicator lines. In other embodiments, reference is made to Figure 2 The first area also contains a timeline. The numbers above the classroom scene diagram represent classroom moments, which together form the timeline of the entire teaching process, allowing you to locate the classroom time corresponding to the location of interest.

[0097] In other embodiments, reference is made to Figure 6 , Figure 6This is another schematic diagram illustrating the classroom learning display method provided in this application. After the indicator line moves to the target position, a preset first command controls the magnification of the information within a preset size interval indicated by the indicator line, allowing for viewing of the interval's details. The first command is used to trigger the magnification of information within a time range centered on the indicator line. It should be noted that the form of the first command is not limited; it can be any form of input operation supported by the computer device. Similarly, no specific requirements are set regarding the position or size of the magnified display time interval relative to the indicator line. In another embodiment, the first command is used to trigger the magnification of information within a time range following the indicator line. Therefore, adjustments can be made according to actual needs to most effectively meet the user's requirements.

[0098] As can be seen, information within a specified range can be magnified, making details clearer and thus improving the visibility of the magnified information. At the same time, it allows for more accurate interpretation and understanding of the data, thereby improving the precision of data interpretation.

[0099] In other embodiments, reference is made to Figure 7 , Figure 7 This is another schematic diagram illustrating the classroom learning display method provided in this application. The method, used in conjunction with a video player, can interpret the teaching process by combining classroom learning data and classroom videos. There are various specific combinations, and this application does not limit them. In one optional embodiment, the classroom learning data and classroom videos are vertically stacked. The classroom videos include videos of the podium area and the student area. During video playback, the indicator line moves along with the playback progress.

[0100] As can be seen, this approach allows users to intuitively see the correspondence between classroom learning data and classroom videos, thereby gaining a deeper understanding of the teaching process.

[0101] In other embodiments, a second command is set, which can be triggered to send a command to an external system to control the video player to jump the playback time of the classroom video to the classroom time indicated by the indicator line, which is more conducive to interpreting the teaching process by combining classroom learning data and classroom video.

[0102] In other embodiments, a third command is provided, which, when triggered, can magnify the first, second, and third regions as a whole, and the magnification factor can be controlled.

[0103] In summary, the method presented in this application enhances the analytical capabilities of classroom learning data (i.e., learning indicators, classroom scenarios, and student behavior distribution), providing richer and more intuitive data presentation formats. It graphically displays large-scale student behavior data through student behavior distribution maps, visualizes the entire teaching process using classroom scenario maps with large-granularity, pedagogically meaningful slices, and visualizes highly comprehensive classroom learning indicators through learning indicator maps. These three presentation methods organically integrate student behavior distribution, the teaching process, and highly comprehensive learning indicators, providing a novel and efficient way to present learning information.

[0104] Based on the student behavior distribution map, teachers can quickly locate specific time points or periods of student behavior distribution that interest them; based on the classroom scene map, teachers can clearly locate classroom scenes that interest them; and based on the highs, lows, or fluctuations in the learning indicators, teachers can quickly locate important moments and key times in the classroom. Indicator lines are also provided to indicate these three types of data, allowing teachers to combine them for classroom observation and accurately observe the learning situation at the indicated moments. This improved data type and presentation method not only enhances AI's ability to analyze classroom learning data but also significantly strengthens its role in guiding teachers to interpret the teaching process. Teachers can quickly and comprehensively identify areas of interest, efficiently review the teaching process, and improve the efficiency of teaching research activities.

[0105] This application also provides a method for generating a student behavior distribution map. The method for generating a student behavior distribution map in this embodiment is described below: like Figure 8 As shown, Figure 8 A flowchart illustrating the method for generating the student behavior distribution map provided in this application.

[0106] Step S801: Determine the first student behavior table and the second student behavior table. The first student behavior table includes the student behaviors to be identified by the behavior recognition algorithm, and the second student behavior table includes the student behaviors to be displayed by the student behavior distribution map.

[0107] The first student behavior table includes student behaviors to be identified by the behavior recognition algorithm, and the second student behavior table includes student behaviors to be displayed in the student behavior distribution map.

[0108] The types of student behaviors listed in the first and second student behavior tables are not specified here.

[0109] It should be noted that the first and second student behavior tables can contain various types of student behaviors. However, each behavior in the first student behavior table must be able to map to a certain behavior in the second student behavior table. This mapping can be one-to-one or many-to-one. For example, if the first student behavior table contains the behaviors "student reading" and "student writing," while the second student behavior table contains the behavior "student reading and writing," the mapping relationship is as follows: the "student reading" and "student writing" behaviors in the first student behavior table are both mapped to the "student reading and writing" behavior in the second student behavior table. Typically, the second student behavior table contains a behavior named "Other," and behaviors in the first behavior table that do not require special attention are mapped to "Other" behaviors.

[0110] The technology related to student behavior recognition has reached a fairly mature level, and will not be elaborated on here.

[0111] Step S802: At each sampling time, the behavior recognition algorithm is used to analyze the classroom video to obtain several student behaviors. Then, the first student behavior table is used to extract all or part of the student behaviors at the current sampling time, and the student behavior distribution data at the current sampling time is statistically analyzed. The student behavior distribution data at all sampling times are summarized into the first dataset. The sampling time is determined by the preset sampling strategy.

[0112] The sampling strategy can determine the sampling time at a fixed time interval, such as sampling once every 1 second or 2 seconds. In other embodiments, sampling intervals of varying lengths are also possible, and this is not limited here.

[0113] To put it simply: at each preset sampling time, a behavior recognition algorithm is used to analyze the classroom video to identify and record student behavior. The identified student behaviors are then mapped to a first student behavior table. The distribution of student behavior data at the current sampling time is then statistically analyzed.

[0114] Once the above steps are triggered, the student behavior distribution data at each moment will be collected and aggregated into a first dataset.

[0115] The specific recognition algorithm is not limited here; it can be a deep learning AI algorithm that analyzes classroom videos to classify the behavior of all or part of the students at the current sampling moment.

[0116] In actual use, some students' actions cannot be identified due to obstruction or other reasons. In such cases, this part of the data that cannot be accurately identified may interfere with or mislead the final analysis results, so this part of the data is discarded.

[0117] It is evident that this processing method helps to ensure the quality of the data and the accuracy of the analysis results.

[0118] Step S803: Generate a second dataset from the first dataset according to a preset mapping algorithm. The student behaviors corresponding to the elements of the second dataset are contained in the second student behavior table, and the time intervals between adjacent elements in the second dataset are the same.

[0119] As can be seen from step S801, the types of student behaviors included in the first student behavior table and the second student behavior table may be different. Also, as can be seen from step S801, depending on the sampling strategy used, the sampling times may be sparse or not evenly spaced. Therefore, time mapping and student behavior mapping may be necessary to map the first data into the second dataset. The second dataset has two characteristics: the first characteristic is that the student behaviors corresponding to the elements in the second dataset are included in the second student behavior table; the second characteristic is that the time intervals between adjacent elements are the same. This application does not limit the specific interval value; in this embodiment, the time interval is set to 1 second. If the class duration is 40 minutes, then the second dataset has 2400 elements.

[0120] As for the specific mapping methods, the relevant technologies have reached a fairly mature level, so they will not be elaborated here.

[0121] Step S804: Set different drawing colors for each student behavior in the second student behavior table, and draw a two-dimensional student behavior distribution map based on the second dataset. Each element of the second dataset corresponds to a vertical line in the student behavior distribution map. The vertical line includes one or more sub-segments, each sub-segment corresponds to a student behavior, and the length of the sub-segment is proportional to the number of students corresponding to the behavior. Sub-segments of different student behaviors are drawn with different colors.

[0122] A predefined drawing color is assigned to each student behavior in the second student behavior table. Using this drawing color, a vertical line can be drawn for a student behavior distribution map based on an element of the second dataset. The following example illustrates how to draw the corresponding vertical line for element E of the second dataset. Assume the total number of students is 48, the height of the student behavior distribution map is 200 pixels (i.e., the length of the vertical line is 200 pixels), and the student behavior distribution data for element E is: (Raised hand = 3, Standing = 2, Listening = 30, Reading and writing = 10, Turning away = 0, Other = 3). The vertical line is drawn from top to bottom in the order of raised hand, standing, listening, reading and writing, turning away, and others: First, using the "raised hand color," a line of length [missing information] is drawn from the starting point downwards. A line segment of length L1 (L1 = 3 * 200 / 48 = 12) is drawn, and then a line segment of length L2 (L1 = 2 * 200 / 48 = 8) is drawn next using the "standing color". Using the same calculation method, a line segment of length L3 (125) corresponding to "listening" and a line segment of length L4 (41) corresponding to "reading and writing" are drawn. There are 0 people with their backs turned, so no line segment is drawn for them. Finally, a line segment of length L5 (L5 = 200 - L1 - L2 - L3 - L4 - L5 = 14) is drawn using the color corresponding to "others". It should be noted that the above is just an example, not a limitation. It can be drawn in other ways. For example, in the above example, vertical lines are drawn from top to bottom in the order of raising hands, standing, listening, reading and writing, turning back, and others. In fact, vertical lines can be drawn in other orders. The vertical lines drawn for all elements of the second dataset constitute the student behavior distribution map. If the second dataset has 2400 elements, then the resolution of the generated student behavior distribution map will be 2400*200.

[0123] This application also provides a method for calculating student participation, such as... Figure 9 As shown, Figure 9 A flowchart illustrating the method for calculating student participation provided in this application.

[0124] Step S901: Determine the first sampling time according to the preset first sampling strategy. For each first sampling time, analyze the teaching process data, identify the preset student behaviors, and statistically obtain the distribution of student behaviors. It should be noted that the sampling interval of the first sampling strategy can be of equal or unequal length; no limitation is made here. Regarding the sampling interval in the first sampling strategy, in practical use, a shorter sampling interval leads to a reduction in analysis granularity, which usually improves the quality of the analysis results. However, a shorter sampling interval also results in a greater computational load, requiring higher computing power and potentially increasing the cost of computing equipment. In a preferred embodiment, the sampling interval is set within 1 to 2 seconds. This means that the first sampling strategy samples at a interval of 1 to 2 seconds, which ensures good computational results while controlling the computational load and equipment cost to a certain extent.

[0125] For each initial sampling time, the teaching process data of the classroom is analyzed. This data includes images of the student area, identifying pre-defined student behaviors in the classroom, and counting the number of students exhibiting each pre-defined behavior and their percentage to obtain the student behavior distribution. This application does not limit the pre-defined student behaviors; depending on specific needs, they may include several of the following: raising hands, standing, listening, reading, writing, turning one's back, and leaning on the desk, or other types of student behaviors.

[0126] It should be noted that, in addition to images of the student area, the teaching process data can also include other types of data to enhance the accuracy of student behavior recognition; this is not a limitation.

[0127] The technology for identifying student behavior has reached a fairly mature level, and will not be elaborated on here.

[0128] It should be noted that the teaching process data analyzed at each first sampling moment is not limited to the data at the current moment. In one specific embodiment, student behavior is identified solely based on the student area images at the first sampling moment. This method is relatively simple and has lower performance requirements for computing devices. In another specific embodiment, student behavior is identified based on several frames of student area images before and after the first sampling moment to enhance recognition accuracy. For example, continuously analyzing 15 consecutive frames of student area images before and after the first sampling moment to identify student behavior. This technical solution has higher performance requirements for computing devices but achieves better results. It is worth noting that, typically, the interval between the acquisition of teaching process data is shorter than the interval of the first sampling moment; for example, the acquisition of student area images is 30 frames per second.

[0129] In practical use, student behavior recognition may be affected by occlusion or other factors, potentially generating some recognition results with low reliability. Therefore, in a preferred embodiment, a first sampling time is determined according to a preset first sampling strategy. For each first sampling time, the student area image in the teaching process data is analyzed to identify preset student behaviors in the student area image. The reliability of each preset student behavior is calculated according to a preset reliability algorithm. Preset student behaviors with a reliability greater than a preset reliability threshold are extracted. The distribution of student behaviors at each first sampling time is statistically analyzed. This application does not limit the specific reliability algorithm. As an example, parameters for calculating the reliability of a student behavior may include: the reliability value output by the neural network, the behavioral changes of the student before and after the behavior, the degree of occlusion of the student, the size of the student's head, etc. Combining various reliability parameters, a usable reliability algorithm can be easily implemented, which will not be elaborated here.

[0130] As can be seen, by employing a preset first sampling strategy to determine each first sampling moment and analyzing the data of the teaching process at each first sampling moment, preset student behaviors in the student area screen during the teaching process are identified. A preset confidence algorithm is used to calculate the confidence level of each preset student behavior, and only preset student behaviors with a confidence level greater than a preset confidence threshold are extracted. This not only improves the accuracy of student behavior identification and reduces the impact of misidentification on student participation calculation, but also, in some other embodiments, all student behaviors identified at the first sampling moment can be directly counted; this is not limited here.

[0131] The above embodiments of the data collection system for the teaching process have been described in detail and will not be repeated here.

[0132] Step S902: Determine the second sampling time according to the preset second sampling strategy. For each second sampling time, analyze the teaching process data and determine the classroom scenario category based on the distribution of various classroom activity parameters within the preset time period before and after the second sampling time. The sampling interval of the second sampling strategy can be of equal or unequal length; this is not limited here. In some embodiments, the second sampling strategy differs from the first sampling strategy, and different sampling intervals are determined according to actual conditions; this is not limited here. In other embodiments, the second sampling strategy is the same as the first sampling strategy; this is not limited here. Using the same first and second sampling strategies simplifies the system flow.

[0133] The teaching process data in step S902 includes student area screens, teacher's area screens, classroom audio, and courseware screens. It may also include other teaching process data, such as operation records from handheld smart teaching devices; this is not limited. It should be noted that the teaching process data used in step S902 actually includes the teaching process data used in step S901. The teaching process data in step S901 mainly reflects student behavior, while the teaching process data in step S902 is more comprehensive, including not only student behavior but also teacher behavior and the behavior of the entire classroom. Although the content and focus of the data differ, both in step S901 and step S902, this data is uniformly referred to as teaching process data.

[0134] It should be noted that the teaching process data analyzed at each second sampling moment is not limited to the data at the current moment. For example, when analyzing classroom sound, it is necessary to consider continuous sound segments, not just the sound at the current second sampling moment. Similarly, the analysis of teacher and student behavior may require considering the images of several frames before and after the current second sampling moment.

[0135] Specifically, classroom activity parameters include teacher behavior, student behavior, courseware behavior, and classroom sound. Classroom sound includes one or more of the following: teacher voice, individual student voice, student discussion, choral student voice, and silence. In essence, the aforementioned teacher behavior, student behavior, courseware behavior, and classroom sound are all generalizations. For example, teacher behavior can be further subdivided into: teacher's blackboard writing, teacher's facial orientation, and teacher's position in the classroom; student behavior can be subdivided into: standing, raising hands, listening, writing, reading, and turning away; courseware behavior refers to actions involving the courseware, such as changes in the courseware screen or clicking on courseware content; classroom sound can also be further subdivided, commonly including teacher voice (teacher speaking), individual student voice (single student speaking), student discussion (multiple students discussing a question), choral student voice (multiple students reading or answering together), and silence (no sound or volume below a preset threshold). Specific behavior categories are not limited here. A more comprehensive and detailed definition of classroom activity parameters can capture classroom activity characteristics more deeply and comprehensively, thereby improving the accuracy of classroom scene recognition.

[0136] Specifically, classroom scene categories include two or more of the following: teacher lecturing, presentation slides, teacher-student Q&A, student practice, student-student interaction, student presentations, and classroom organization. In practice, more classroom scenes can be included; this is not limited here. It should be noted that too few classroom scenes may fail to accurately depict the teaching process, while too many scenarios will break down the teaching process into excessively fine segments, resulting in fragmented classroom scene diagrams that are not conducive to reviewing the teaching process. The specific granularity of segmentation needs to be determined based on actual user needs. One alternative approach is to identify a wider range of classroom scene types, which can improve the accuracy of student participation calculations. When displaying these scenarios to users through the UI, several classroom scenes with similar teaching meanings can be merged into one classroom scene. For example, a teacher-led verbal lecture scenario (the teacher lectures verbally), a teacher-led blackboard lecture scenario (the teacher lectures while writing on the blackboard), and a teacher-led demonstration lecture scenario (the teacher lectures while demonstrating) can be identified and merged into a single teacher-led lecture scenario for display.

[0137] In one embodiment of step S902, it includes: Step S9021: Determine the second sampling time according to the preset second sampling strategy; Step S9022: For each second sampling time, analyze the teaching process data, identify classroom activity parameters, and generate a classroom activity parameter dataset; For each second sampling time, the teaching process data is analyzed to identify various classroom activity parameters, obtaining classroom activity parameter data, such as teacher behavior categories and durations, classroom sound categories and durations, and the number of students exhibiting various behaviors. Specific classroom activity parameters and their data are not limited here. Each classroom activity parameter data point at each second sampling time corresponds to an element in the classroom activity parameter dataset. All classroom activity parameter data points from all second sampling times are used to generate the classroom activity parameter dataset.

[0138] The technology for identifying classroom activity parameters (i.e., teacher behavior, student behavior, courseware behavior, and classroom sound, etc.) is a mature technology and will not be elaborated here. Algorithms for obtaining classroom activity parameter data (including teacher behavior categories and durations, classroom sound categories and durations, and the number of students exhibiting various behaviors, etc.) based on the identified parameters are readily available and can be easily determined according to specific needs; therefore, they will not be elaborated here.

[0139] Step S9023: For the current second sampling time, determine several time windows that include the current second sampling time according to the preset time window selection strategy; the current second sampling time is any second sampling time. This application does not limit the selection strategy for time windows. In an exemplary example, according to the preset time window selection strategy, all time periods containing the current second sampling moment with a duration of 16 seconds and 20 seconds are selected as time windows. For example, if the current second sampling moment is the 100th second of the teaching process, then the time periods of 85 seconds to 100 seconds, 86 seconds to 101 seconds, ..., 100 seconds to 115 seconds are selected as time windows, and the time periods of 81 seconds to 100 seconds, 82 seconds to 101 seconds, ..., 100 seconds to 119 seconds are also selected as time windows.

[0140] Step S9024: For the current time window, statistically analyze the distribution of classroom activity parameters based on the classroom activity parameter dataset, and calculate the characteristic values ​​of classroom activities. The current time window can be any time window. For the current time window, the distribution of various classroom activity parameters is statistically analyzed based on the classroom activity parameter dataset. Clearly, the classroom activity parameter data for all second sampling moments within the current time window is also a set, a subset of the classroom activity parameter dataset. The distribution of various classroom activity parameters is statistically analyzed on this subset. This application does not limit the specific statistical method; a technical solution can be easily determined based on specific needs. For ease of understanding, the following is an exemplary description of the distribution of classroom activity parameters: for student behavior, the distribution of various student behaviors at each moment within the time window is statistically analyzed; for teacher voice, the duration and percentage of teacher voice throughout the entire time window are statistically analyzed.

[0141] Then, based on the distribution of the various classroom activity parameters mentioned above, the characteristic values ​​of classroom activities in the current time window are calculated. The distribution data of the classroom activity parameters reflects the characteristics of various classroom activities in this time window. By combining the distribution data of various classroom activity parameters in the current time window, the characteristic values ​​of classroom activities in this time window are calculated to identify the classroom scene category. The characteristic value of classroom activities can be a scalar value or a vector.

[0142] Step S9025: Determine the classroom scene category at the current second sampling moment based on the classroom activity feature values.

[0143] For all time windows of the current second sampling moment, the corresponding classroom activity feature values ​​were calculated. The classroom activity feature values ​​of all time windows were comprehensively analyzed, that is, the distribution of various classroom activity parameters within the preset time period before and after the current second sampling moment was comprehensively analyzed. The classroom activity features within a period of time before and after the current second sampling moment were combined to determine the classroom scene category of the current second sampling moment. This application does not impose limitations on the calculation method for classroom activity feature values ​​or the algorithm for class scene category determination. As a suggestive point, the design principle of the classroom activity feature value calculation method is to highlight the most important classroom activity features within the current time window as much as possible, thereby improving the accuracy of classroom scene identification.

[0144] In one specific embodiment, for each time window, a score is calculated for each classroom scenario. The scores of all classroom scenarios in a time window are the classroom activity feature values ​​for that time window. According to a preset strategy, the classroom scenario category is determined based on the most prominent classroom activity feature value among all time windows. For example, if the teacher-led teaching scenario has the highest score, then the classroom scenario category at the current second sampling moment is determined to be "teacher-led teaching".

[0145] In another embodiment, a pre-set classroom scenario table is used. This table contains multiple columns, each corresponding to a specific classroom scenario. Within this table, each classroom scenario corresponds to a certain range of classroom activity characteristic values. Thus, the corresponding classroom scenario can be determined by comparing the calculated classroom activity characteristic values ​​with the range in the table.

[0146] As can be seen, by analyzing rich teaching process data, a detailed dataset of classroom activity parameters reflecting the characteristics of classroom activities can be obtained. By using a preset time window selection strategy, not only is the distribution of various classroom activity parameters at the current second sampling moment analyzed, but also the distribution of various classroom activity parameters within the preset time periods before and after it. This avoids the error problems caused by a single moment and enhances the accuracy of classroom scene category identification.

[0147] In another embodiment of step S902, it includes: The second sampling time is determined according to the preset second sampling strategy; For each second sampling moment, the teaching process data within a preset time period before and after the second sampling moment is input into a preset classroom scene classification model to obtain the classroom scene category. The training samples of the classroom scene classification model are historical teaching process data. The classroom scene category of each classroom moment in the training samples is determined according to the distribution of classroom activity parameters within a certain period before and after that classroom moment.

[0148] The training samples for the classroom scene classification model are historical teaching process data, which includes images of historical courseware, the historical podium area, the historical student area, and historical classroom audio. For each training sample, corresponding to a classroom teaching process, the classroom scene category for each classroom moment needs to be labeled. For each classroom moment, the labelers repeatedly observe the distribution of various classroom activity parameters over a period of time before and after that moment to determine the specific classroom scene category.

[0149] This application does not impose any limitations on the classroom scene classification model. It can be an end-to-end model, that is, receiving teaching process data and directly outputting classroom scene categories, or it can be a two-stage or more-stage model, where the first stage of the model receives teaching process data, outputs intermediate results to the next stage of the model, and so on, with the final stage of the model outputting the classroom scene category.

[0150] In one embodiment, the classroom scene classification model is trained as follows: Relevant data, including historical courseware images, historical lectern area images, historical student area images, and historical classroom audio, are collected in advance. This data can come from a historical database or be manually input. Based on the collected data, the model is trained using supervised learning. When labeling each training sample (each training sample corresponds to a classroom teaching process), for each classroom moment, the labelers repeatedly observe the distribution of various classroom activity parameters before and after that moment to determine the specific classroom scene category. That is, a correlation is established between the classroom scene category of each classroom moment and the distribution of various classroom activity parameters before and after that moment.

[0151] Of course, in other embodiments, the classroom scene classification model can also be trained using other methods, which are not limited here.

[0152] Step S903: Generate a sequence of classroom scenes for the entire teaching process based on the classroom scene category; A sequence of classroom scenes representing the entire teaching process can be generated using the classroom scene category at each second sampling time. This application does not limit the specific method; a simple approach is to divide consecutive time periods with the same classroom scene category during the teaching process into independent classroom scenes, thus generating a sequence of classroom scenes representing the entire teaching process.

[0153] However, the above method may result in an excessive number of teaching scenarios with overly fine granularity. To address this issue, a filtering algorithm can be introduced: if the duration of a particular classroom scenario is less than a set threshold, it is merged into its adjacent teaching scenarios. This approach not only reduces the number of classroom scenarios but also ensures that each scenario has sufficient length, which is more conducive to subsequent analysis and understanding.

[0154] Step S904: Determine the participation weight table based on the conformity between the preset student behavior and the classroom scenario; Referring to Table 1, which is an example of an engagement weighting table, the rows and columns correspond to preset student behavior categories and classroom scenario categories, respectively. The value of each item is determined by the degree to which the corresponding student behavior matches the classroom scenario. The larger the value, the higher the match between the student behavior and the classroom scenario; in other words, the higher the student's engagement with that behavior in that classroom scenario. This approach helps to more clearly understand and quantify student engagement in different classroom scenarios.

[0155] Table 1 categorizes student behaviors into three types: listening, writing, and slumping over the desk; and class scenarios into five types: teacher lecturing, presentation slides, student presentations, student practice, and teacher-student Q&A. In this table, the weight value for "listening" corresponding to "teacher lecturing" is 10, indicating a high match between listening behavior and the teacher lecturing scenario, and high student engagement. Conversely, the weight value for "slumping over the desk" corresponding to "teacher lecturing" is -10, indicating a mismatch between student slumping over the desk and the teacher lecturing scenario, and low student engagement. The weight value for "writing" corresponding to "teacher lecturing" is 3, indicating a partial match between student writing and the teacher lecturing scenario, and moderate student engagement.

[0156] Table 1: Participation Weighting Table As can be seen from the above, the consistency between student behavior and classroom scenario describes the degree to which students participate in the current classroom teaching. If the behavior of each student in the classroom is accurately extracted and the current classroom scenario is accurately identified, the overall degree of student participation in classroom teaching activities can be calculated.

[0157] It should be noted that the participation weight table can be static, meaning the value (weight) of each item remains constant throughout the lesson; or it can be dynamic, meaning the value of each item differs for different classroom moments and is dynamically adjusted based on the characteristics of teaching activities within a time period before and after the current classroom moment. The specific format of the participation weight table, the weight values ​​of the items and their determination methods, and the number or types of student behavior categories and classroom scenario categories are not limited here.

[0158] Step S905: Calculate the student participation value for each classroom moment based on the participation weight table, student behavior distribution, and classroom scene sequence.

[0159] The classroom time in step S905 is not limited here; an interval of 1 second is sufficient.

[0160] Specifically, for any first sampling moment, the classroom scenario at that moment can be determined based on the classroom scenario sequence. Then, the weight values ​​of various student behaviors in that classroom scenario can be found from the participation weight table, thereby calculating the student participation value for each first sampling moment.

[0161] The specific calculation method is not limited in this application. In one embodiment, a simple method is used, which involves weighting and summing the student behavior distribution at each first sampling time and dividing by the total number of students to obtain the student participation value. For ease of understanding, using the weight table in Table 1 as an example, suppose the classroom scenario at the current first sampling time is "teacher lecturing", with a total of 50 students and a student behavior distribution of (listening = 30, writing = 15, and slumped over the desk = 5). Then, the student participation value at the current first sampling time is (30*10 + 3*15 - 5*10) / 50 = 5.9. In another embodiment, for each first sampling time, the student participation value is calculated by comprehensively considering the student behavior distribution within a time interval including the current first sampling time and the classroom scenario.

[0162] In this way, after calculating the student engagement values ​​for all the first sampling moments, interpolation techniques are used to fill in the student engagement values ​​between these sampling moments, thus obtaining the student engagement value for each moment in the class. Based on the student engagement values ​​for all class moments, a student engagement curve can be plotted graphically.

[0163] In summary, it is evident that by analyzing teaching process data and classroom activity parameters, the correlation between classroom scenario categories and student participation is clearly considered. By generating a sequence of classroom scenarios throughout the entire teaching process and determining a participation weight table based on the pre-set conformity of student behavior with the classroom scenarios, student participation at each stage of the teaching process can be quantified more accurately, reflecting actual student participation in different classroom scenarios. Furthermore, since this invention can calculate student participation values ​​at each classroom moment, it reduces misjudgments of student participation and improves the accuracy of student participation results.

[0164] The following are device embodiments of this application, which can be used to execute the method embodiments of this application. For details not disclosed in the device embodiments of this application, please refer to the method embodiments of this application.

[0165] refer to Figure 10 This application provides a student participation calculation system, including a computer, which includes: The student behavior distribution module 1001 is used to determine the first sampling time according to the preset first sampling strategy. For each first sampling time, it analyzes the teaching process data, identifies the preset student behaviors, and statistically obtains the student behavior distribution. The classroom scenario category module 1002 is used to determine the second sampling time according to a preset second sampling strategy. For each second sampling time, the teaching process data is analyzed, and the classroom scenario category is determined based on the distribution of various classroom activity parameters within a preset time period before and after the second sampling time. Classroom Scene Sequence Module 1003 is used to generate a sequence of classroom scenes for the entire teaching process based on the classroom scene category; The participation weight table module 1004 is used to determine the participation weight table based on the conformity between preset student behavior and classroom scenario; the student participation value module 1005 is used to calculate the student participation value for each classroom moment based on the participation weight table, student behavior distribution and classroom scenario sequence.

[0166] In some embodiments, the first sampling strategy and the second sampling strategy are the same.

[0167] In some embodiments, the student behavior distribution module specifically includes: The first sampling strategy submodule is used to determine the first sampling time according to the preset first sampling strategy; The preset student behavior submodule is used to analyze the student area screen in the teaching process data for each first sampling time and identify the preset student behavior in the student area screen; The credibility submodule is used to calculate the credibility of each preset student behavior according to a preset credibility algorithm; The credibility threshold submodule is used to extract preset student behaviors with credibility values ​​greater than a preset credibility threshold. The student behavior distribution submodule is used to statistically analyze the student behavior distribution at each first sampling time.

[0168] In some embodiments, classroom activity parameters include teacher behavior, student behavior, courseware behavior, and classroom sound; classroom sound includes one or more of the following: teacher sound, individual student sound, student discussion sound, student chorus sound, and silence.

[0169] In some embodiments, the classroom scenario categories include two or more of the following: teacher lecturing, courseware presentation, teacher-student Q&A, student practice, student-student interaction, student presentation, and classroom organization.

[0170] In some embodiments, the classroom scenario category module specifically includes: The second sampling time submodule is used to determine the second sampling time according to the preset second sampling strategy; The teaching process data submodule is used to analyze teaching process data, identify classroom activity parameters, and generate a classroom activity parameter dataset for each second sampling time. The time window submodule is used to determine, based on a preset time window selection strategy, several time windows that include the current second sampling moment; the current second sampling moment can be any second sampling moment. The Classroom Activity Feature Value Submodule is used to calculate the feature value of classroom activities for the current time window by statistically analyzing the distribution of classroom activity parameters based on the classroom activity parameter dataset. The current time window can be any time window. The classroom scene category submodule is used to determine the classroom scene category at the current second sampling moment based on the feature values ​​of classroom activities.

[0171] In some embodiments, the classroom scenario category module specifically includes: The second sampling time submodule is used to determine the second sampling time according to the preset second sampling strategy; The classroom scene classification model submodule is used to input the teaching process data within a preset time period before and after the second sampling time into the preset classroom scene classification model for each second sampling time to obtain the classroom scene category. The training samples of the classroom scene classification model are historical teaching process data, and the classroom scene category of each classroom moment in the training samples is determined according to the distribution of classroom activity parameters within a certain period before and after that classroom moment.

[0172] This application also discloses a system for calculating student engagement. (See reference...) Figure 11 This is a schematic diagram of the physical device of the student engagement calculation system provided in this application. The computer 1100 may include: at least one processor 1101, at least one network interface 1104, a user interface 1103, a memory 1105, and at least one communication bus 1102.

[0173] The communication bus 1102 is used to realize the connection and communication between these components.

[0174] The user interface 1103 may include a display screen and a camera. Optionally, the user interface 1103 may also include a standard wired interface and a wireless interface.

[0175] The network interface 1104 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface).

[0176] The processor 1101 may include one or more processing cores. The processor 1101 connects to various parts of the server using various interfaces and lines, and performs various server functions and processes data by running or executing instructions, programs, code sets, or instruction sets stored in memory 1105, and by calling data stored in memory 1105. Optionally, the processor 1101 may be implemented using at least one hardware form of Digital Signal Processing (DSP), Field-Programmable Gate Array (FPGA), or Programmable Logic Array (PLA). The processor 1101 may integrate one or a combination of several of the following: Central Processing Unit (CPU), Graphics Processing Unit (GPU), and modem. The CPU primarily handles the operating system, user interface, and applications; the GPU is responsible for rendering and drawing the content required for display; and the modem handles wireless communication. It is understood that the modem may also not be integrated into the processor 1101 and may be implemented as a separate chip.

[0177] The memory 1105 may include random access memory (RAM) or read-only memory. Optionally, the memory 1105 may include non-transitory computer-readable storage medium. The memory 1105 can be used to store instructions, programs, code, code sets, or instruction sets. The memory 1105 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as touch function, sound playback function, image playback function, etc.), instructions for implementing the above-described method embodiments, etc.; the data storage area may store data involved in the above-described method embodiments, etc. Optionally, the memory 1105 may also be at least one storage device located remotely from the aforementioned processor 1101. (Refer to...) Figure 5 The memory 1105, which serves as a computer storage medium, may include an operating system, a network communication module, a user interface module, and an application program for calculating student participation.

[0178] exist Figure 5In the computer 1100 shown, the user interface 1103 is mainly used to provide an input interface for the user and obtain user input data; while the processor 1101 can be used to call the application program for calculating student participation stored in the memory 1105. When executed by one or more processors 1101, the computer 1100 performs one or more of the methods described in the above embodiments. It should be noted that, for the foregoing method embodiments, for the sake of simplicity, they are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, because according to this application, some steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also understand that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0179] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0180] In the various embodiments provided in this application, it should be understood that the disclosed apparatus can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some service interface; the indirect coupling or communication connection between apparatuses or units may be electrical or other forms.

[0181] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0182] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0183] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage device (CMD). Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned memory includes various media capable of storing program code, such as USB flash drives, portable hard drives, magnetic disks, or optical disks.

[0184] The above description is merely an exemplary embodiment of this disclosure and should not be construed as limiting the scope of this disclosure. Any equivalent changes and modifications made in accordance with the teachings of this disclosure shall still fall within the scope of this disclosure. Other embodiments of this disclosure will be readily apparent to those skilled in the art upon consideration of the specification and the disclosure of practical truths.

[0185] This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not described in this disclosure. The specification and embodiments are to be considered exemplary only, and the scope and spirit of this disclosure are defined by the claims.

Claims

1. A method for calculating student participation, characterized in that, include: The first sampling time is determined according to the preset first sampling strategy. For each first sampling time, the teaching process data is analyzed, the preset student behaviors are identified, and the distribution of student behaviors is statistically obtained. The second sampling time is determined according to the preset second sampling strategy. For each second sampling time, the teaching process data is analyzed, and the classroom scene category is determined according to the distribution of various classroom activity parameters within the preset time period before and after the second sampling time. The classroom activity parameters include teacher behavior, student behavior, courseware behavior, and classroom sound. The classroom sound includes one or more of the following: teacher voice, student voice, student discussion, student chorus, and silence. Generate a sequence of classroom scenarios for the entire teaching process based on the aforementioned classroom scenario categories; The participation weighting table is determined based on the conformity between the pre-set student behavior and the classroom scenario; The student participation value for each classroom moment is calculated based on the participation weight table, the student behavior distribution, and the classroom scene sequence.

2. The method for calculating student participation according to claim 1, characterized in that: The first sampling strategy and the second sampling strategy are the same.

3. The method for calculating student participation according to claim 1, characterized in that, The steps of determining the first sampling time according to the preset first sampling strategy, analyzing teaching process data, identifying preset student behaviors, and statistically obtaining the distribution of student behaviors for each first sampling time specifically include: determining the first sampling time according to the preset first sampling strategy; For each first sampling moment, analyze the student area screen in the teaching process data and identify the preset student behaviors in the student area screen; The credibility of each student behavior is calculated based on a preset credibility algorithm. Extract preset student behaviors with a credibility level greater than a preset credibility threshold; The distribution of student behavior at each first sampling time is statistically analyzed.

4. The method for calculating student participation according to claim 1, characterized in that: The classroom scenario categories include two or more of the following: teacher lecturing, courseware presentation, teacher-student Q&A, student practice, student-student interaction, student presentations, and classroom organization.

5. The method for calculating student participation according to claim 1, characterized in that, The step of determining the second sampling time according to the preset second sampling strategy, and for each second sampling time, analyzing the teaching process data and determining the classroom scene category based on the distribution of various classroom activity parameters within the preset time period before and after the second sampling time, specifically includes: determining the second sampling time according to the preset second sampling strategy; For each second sampling time, analyze the teaching process data, identify classroom activity parameters, and generate a classroom activity parameter dataset; For the current second sampling moment, a number of time windows containing the current second sampling moment are determined according to a preset time window selection strategy; The current second sampling time can be any of the second sampling times; For the current time window, the distribution of classroom activity parameters is statistically analyzed based on the classroom activity parameter dataset, and the characteristic value of classroom activity is calculated. The current time window can be any of the aforementioned time windows. The classroom scene category at the current second sampling moment is determined based on the classroom activity feature values.

6. The method for calculating student participation according to claim 1, characterized in that, The step of determining the second sampling time according to the preset second sampling strategy, and for each second sampling time, analyzing the teaching process data and determining the classroom scene category based on the distribution of various classroom activity parameters within the preset time period before and after the second sampling time, specifically includes: determining the second sampling time according to the preset second sampling strategy; For each second sampling moment, the teaching process data within a preset time period before and after the second sampling moment is input into a preset classroom scene classification model to obtain the classroom scene category. The training samples of the classroom scene classification model are historical teaching process data. The classroom scene category of each classroom moment in the training samples is determined according to the distribution of classroom activity parameters within a certain period before and after that classroom moment.

7. A system for calculating student participation, characterized in that, The computer includes a student behavior distribution module, which is used to determine a first sampling time according to a preset first sampling strategy, and for each first sampling time, analyze teaching process data, identify preset student behaviors, and statistically obtain the student behavior distribution. The classroom scenario category module is used to determine the second sampling time according to a preset second sampling strategy. For each second sampling time, the teaching process data is analyzed, and the classroom scenario category is determined based on the distribution of various classroom activity parameters within a preset time period before and after the second sampling time. The classroom activity parameters include teacher behavior, student behavior, courseware behavior, and classroom sound. The classroom sound includes one or more of the following: teacher voice, student voice, student discussion, student chorus, and silence. The classroom scene sequence module is used to generate a sequence of classroom scenes for the entire teaching process based on the classroom scene category. The participation weighting table module is used to determine the participation weighting table based on the conformity between preset student behavior and classroom scenario; The student participation value module is used to calculate the student participation value for each classroom moment based on the participation weight table, the student behavior distribution, and the classroom scene sequence.

8. A system for calculating student participation, characterized in that, include: One or more processors and memory; The memory is coupled to the one or more processors, the memory being used to store computer program code, the computer program code including computer instructions, the one or more processors invoking the computer instructions to cause the student engagement calculation system to perform the method as described in any one of claims 1-6.

9. A computer-readable storage medium comprising instructions, characterized in that, When the instruction is executed on the student engagement calculation system, the student engagement calculation system performs the method as described in any one of claims 1-6.

Images