Teaching behavior analysis system and method based on artificial intelligence

Abstract

The invention provides a teaching behavior analysis system and analysis method based on artificial intelligence, and the method comprises the steps: recognizing each knowledge point explained in a teaching process based on the teaching voice of a teacher, and determining the explanation time of each knowledge point; based on the image data of each student in the teaching process, analyzing the understanding state of each student, and determining the understanding duration of each student corresponding to each knowledge point; teaching feedback information is generated based on the explanation duration of each knowledge point and the understanding duration of each student corresponding to each knowledge point; wherein the teaching feedback information comprises knowledge point explanation duration rationality evaluation and knowledge point understanding efficiency analysis of students. By automatically identifying knowledge points and associating the understanding duration of students, the teacher can conveniently and scientifically optimize classroom time distribution in subsequent teaching. And by analyzing the understanding duration of each student on different knowledge points, a class with synchronous cognition is formed, so that a teacher can grasp a uniform teaching rhythm conveniently.

Description

Technical Field

[0001] This invention relates to the field of education system technology, and in particular to an artificial intelligence-based teaching behavior analysis system and analysis method. Background Technology

[0002] In traditional teaching, teachers often rely on teaching experience or subjective judgment to allocate time for explaining classroom knowledge points, lacking objective data to support their assessment of students' actual understanding. On the one hand, for knowledge points that are easy for students to understand, teachers may extend the explanation time because they cannot promptly perceive students' grasp of the material, resulting in wasted class time. On the other hand, for knowledge points that are difficult for students to understand, teachers may shorten the explanation time because they cannot accurately know the degree of students' confusion, leading to some students not fully mastering the knowledge points and affecting the teaching effect.

[0003] Meanwhile, current student grouping or class assignment methods primarily rely on exam scores, neglecting the non-necessary correlation between scores and knowledge comprehension efficiency. Different students grasp different types of knowledge at varying paces, and classes divided solely by scores cannot guarantee a convergence of comprehension efficiency among students. This makes it difficult for teachers to cater to different students' learning speeds during instruction, failing to meet the needs of faster learners and adapting to the absorption pace of slower learners, further reducing the relevance and effectiveness of classroom teaching. Summary of the Invention

[0004] Based on this, the present invention proposes an artificial intelligence-based teaching behavior analysis method, which aims to provide reasonable feedback on the required explanation time for each knowledge point in order to improve teaching effectiveness; at the same time, it provides suggestions for classifying students with large differences in comprehension efficiency so as to provide unified instruction to students with similar comprehension efficiency, thereby improving the pertinence and effectiveness of classroom teaching.

[0005] Firstly, the technical solution provided in this application includes:

[0006] An artificial intelligence-based method for analyzing teaching behavior, the method comprising:

[0007] Based on the teacher's voice, the system identifies each knowledge point explained during the lecture and determines the duration of each knowledge point's explanation.

[0008] Based on the image data of each student during the teaching process, analyze the understanding status of each student and determine the understanding time of each student for each knowledge point.

[0009] Based on the explanation time for each knowledge point and the understanding time for each student for each knowledge point, teaching feedback information is generated; wherein, the teaching feedback information includes an assessment of the reasonableness of the explanation time for each knowledge point and an analysis of the efficiency of students' understanding of the knowledge points.

[0010] Furthermore, the method of identifying various knowledge points explained during the lecture based on the teacher's voice includes:

[0011] The teacher's lecture audio is processed to convert it into text, resulting in text information.

[0012] Natural language analysis is performed on the text information to extract keywords, semantic topics, or teaching terms, and these are matched with a pre-set knowledge point database to identify each knowledge point.

[0013] Furthermore, determining the explanation time for each knowledge point includes:

[0014] Record the start timestamp when each knowledge point is first mentioned and the end timestamp when it is last mentioned; calculate the difference between the end timestamp and the start timestamp to obtain the explanation duration of the knowledge point.

[0015] Furthermore, the analysis of each student's comprehension state includes:

[0016] Based on the student's image data, by analyzing their head posture and gaze direction, it can be determined whether they are attentive or not.

[0017] For students in the focused listening state, their comprehension state is analyzed using a micro-expression recognition model based on their facial images; the comprehension state includes understanding and not understanding.

[0018] Furthermore, determining the comprehension time for each student for each knowledge point includes:

[0019] Within the duration of an explanation of a knowledge point, analyze the student's comprehension only during the time period when the student is attentively listening.

[0020] If, during the duration of the explanation of a knowledge point, a student transitions from a state of non-listening to a state of focused listening and is identified as having understood the knowledge point, then the difference between the moment the student is identified as having understood the knowledge point and the moment the student enters a state of focused listening is the duration of the student's understanding of that knowledge point.

[0021] If, during the explanation time for a particular knowledge point, a student is found to be not paying attention and their comprehension status remains unchanged, then the student's comprehension time for that knowledge point is marked as invalid data.

[0022] If a student remains attentive throughout the explanation of a knowledge point but does not understand it, the student's understanding time for that knowledge point will be recorded as the total explanation time, and this time will be marked.

[0023] Furthermore, the assessment of the reasonableness of the explanation time for the aforementioned knowledge points includes:

[0024] For a given knowledge point, based on the time students spend understanding it and the time spent explaining it, we can calculate the percentage of students who can understand the knowledge point within the allotted time.

[0025] If the percentage of people who understand a knowledge point within the allotted time is lower than the first threshold, it is assessed that the allotted time for that knowledge point needs to be increased.

[0026] If the percentage of people who understand a knowledge point within the allotted time is higher than the second threshold, then the allotted time for that knowledge point can be shortened.

[0027] If the percentage of people who understand a knowledge point within the allotted time is between the first and second thresholds, then the allotted time for that knowledge point can be maintained.

[0028] Furthermore, the student's efficiency in understanding knowledge points is analyzed, including:

[0029] An individual comprehension efficiency profile is generated for each student. The individual comprehension efficiency profile is composed of the student's standardized comprehension efficiency score across all knowledge points. For a knowledge point, the efficiency score is the ratio of the explanation time for that knowledge point to the student's comprehension time for that knowledge point.

[0030] Based on the individual comprehension efficiency profiles of all students, a clustering algorithm is used to divide all students into high, medium, and low comprehension efficiency groups, and teaching class assignment suggestions are output.

[0031] Secondly, the technical solution provided in this application includes:

[0032] An artificial intelligence-based teaching behavior analysis device, the device comprising:

[0033] The knowledge point recognition module is used to identify the various knowledge points explained during the lecture based on the teacher's voice and determine the explanation time of each knowledge point.

[0034] The student comprehension analysis module is used to analyze the comprehension status of each student based on the image data of each student during the teaching process, and to determine the comprehension time of each student for each knowledge point.

[0035] The feedback information generation module is used to generate teaching feedback information based on the explanation time of each knowledge point and the understanding time of each student for each knowledge point; wherein, the teaching feedback information includes an assessment of the rationality of the explanation time of the knowledge points and an analysis of the efficiency of students' understanding of the knowledge points.

[0036] Thirdly, the technical solution provided in this application includes:

[0037] A computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the steps of the above-described method.

[0038] Fourthly, the technical solution provided in this application includes:

[0039] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described method.

[0040] The technical solution provided in this application has at least the following advantages over the prior art:

[0041] This application provides teachers with objective and quantitative data support regarding the reasonableness of their explanation time allocation by automatically identifying knowledge points and linking them to students' comprehension time. This enables teachers to accurately identify which knowledge points are over-explained and which are under-explained, allowing them to scientifically optimize classroom time allocation in subsequent teaching and effectively improve classroom teaching efficiency.

[0042] This invention analyzes the time each student takes to understand different knowledge points and can group students based on their comprehension efficiency. This allows for the aggregation of students with similar learning paces to form a class with synchronized cognition, greatly facilitating teachers in maintaining a unified teaching rhythm and truly achieving "teaching according to aptitude" and "precision teaching," thereby improving the relevance and effectiveness of teaching from the source. Attached Figure Description

[0043] Figure 1 This is an exemplary flowchart of an AI-based teaching behavior analysis method according to some embodiments of this application;

[0044] Figure 2 This is a structural diagram of an artificial intelligence-based teaching behavior analysis device according to some embodiments of this application; Detailed Implementation

[0045] This specific embodiment is merely an explanation of this application and is not intended to limit it. Those skilled in the art, after reading this specification, can make modifications to this embodiment without contributing any inventive step, but such modifications are protected by patent law as long as they fall within the scope of the claims of this application. To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without inventive effort are within the scope of protection of this application.

[0046] The term "comprising" and any variations thereof in the specification and claims of this application are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or units that are not explicitly listed or that are inherent to such process, method, product or device.

[0047] In the embodiments of this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design that is described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design. Specifically, the use of the terms "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0048] The embodiments of this application will now be described in further detail with reference to the accompanying drawings.

[0049] Figure 1 This application illustrates an exemplary embodiment of an artificial intelligence-based teaching behavior analysis method, the method comprising:

[0050] Step S100: Based on the teacher's lecture voice, identify each knowledge point explained during the lecture and determine the explanation duration of each knowledge point;

[0051] For example, the system uses a directional microphone array in the classroom to capture the teacher's audio. By analyzing this continuous audio signal, it identifies the individual knowledge points contained within. For instance, in a math lesson, it identifies three core knowledge points: "the definition of a quadratic equation," "the derivation of the quadratic formula," and "the application of the discriminant." The system automatically records the explanation time for each knowledge point; for example, the explanation of "the definition of a quadratic equation" is from the 2nd minute to the 8th minute after the start of the lesson, totaling 6 minutes.

[0052] Step S200: Based on the image data of each student during the teaching process, analyze the understanding status of each student and determine the understanding time of each student for each knowledge point.

[0053] For example, multiple cameras can be placed in the classroom to collect image data of students. By analyzing this image data, the students' understanding of the knowledge points can be determined, thereby determining how long it will take for the students to understand the knowledge points.

[0054] Step S300: Based on the explanation time of each knowledge point and the understanding time of each student for each knowledge point, generate teaching feedback information; wherein, the teaching feedback information includes an assessment of the rationality of the explanation time of the knowledge points and an analysis of the efficiency of students' understanding of the knowledge points.

[0055] The above data is then analyzed to generate teaching feedback information. For example, it may be found that a significant number of students are unable to understand a particular knowledge point within the allotted explanation time. In this case, the explanation time for that knowledge point may be deemed too short, and it may be recommended to increase it. Furthermore, the comprehension efficiency of students can be judged based on the time it takes for them to understand a knowledge point, thus providing a more reasonable basis for grouping students with similar learning and comprehension abilities.

[0056] Thus, in this embodiment, by automatically identifying knowledge points and linking them to students' comprehension time, teachers are provided with objective and quantitative data support regarding the reasonableness of the allocation of explanation time. This allows teachers to accurately identify which knowledge points are over-explained and which are under-explained, thereby scientifically optimizing classroom time allocation in subsequent teaching and effectively improving classroom teaching efficiency. By analyzing each student's comprehension time for different knowledge points, classes can be grouped based on students' comprehension efficiency. Students with similar learning paces can be grouped to form a class with synchronized cognition, greatly facilitating teachers in maintaining a unified teaching pace and truly achieving "individualized instruction" and "precision teaching," thereby improving the relevance and effectiveness of teaching from the source.

[0057] In some embodiments, the step of identifying various knowledge points explained during the lecture based on the teacher's voice specifically includes:

[0058] The teacher's lecture audio is processed to convert it into text, resulting in text information.

[0059] Natural language analysis is performed on the text information to extract keywords, semantic topics, or teaching terms, and these are matched with a pre-set knowledge point database to identify each knowledge point.

[0060] Specifically, for example, in a biology lesson on photosynthesis, the teacher's continuous audio explanation is captured in real time. The system uses a speech recognition engine deployed in the cloud to convert the audio stream into a continuous text sequence. The converted text might contain something like: "Today we are going to learn about photosynthesis... Its definition is... Green plants use light energy... to convert carbon dioxide and water into organic matter... and release oxygen..." This step transforms unstructured audio data into text information that can be further processed by the computer, laying the foundation for subsequent analysis.

[0061] Then, key teaching terms (such as "photosynthesis", "light energy", "carbon dioxide", "organic matter", "oxygen") or semantic themes (such as "process definition", "energy conversion", "matter transformation") can be extracted from the text.

[0062] The system has a pre-built structured knowledge point database. This database contains the core knowledge points in the course syllabus and their related terminology. For example, the database may contain a knowledge point titled "Definition of Photosynthesis," with associated keywords including "photosynthesis," "light energy," "carbon dioxide," "water," "organic matter," "oxygen," and "process."

[0063] The system matches the extracted keywords and semantic topics with the knowledge point database. When it detects a dense cluster of terms highly related to the "definition of photosynthesis" in the text, and semantic analysis indicates that the paragraph is describing the definition of a process, the system can determine that the current explanation is indeed of the knowledge point "definition of photosynthesis".

[0064] Thus, this embodiment can accurately identify the various knowledge points contained in the teacher's lecture audio.

[0065] In some embodiments, determining the explanation duration for each knowledge point specifically includes:

[0066] Record the start timestamp when each knowledge point is first mentioned and the end timestamp when it is last mentioned; calculate the difference between the end timestamp and the start timestamp to obtain the explanation duration of the knowledge point.

[0067] Specifically, when the system determines through semantic analysis that a teacher has begun teaching a specific knowledge point and utters the first core term strongly related to that knowledge point or begins a defining statement, it immediately records that moment. For example, in a math class, when the teacher says, "Now let's look at the Pythagorean theorem...", the system recognizes that the explanation of the "Pythagorean theorem" has officially begun and records the current audio stream time as the start timestamp;

[0068] When the system detects that the teacher has reached the end of their core explanation of a knowledge point and determines that the lecture is complete through context analysis, it records this moment. This is usually determined by the following signals: the teacher begins explaining the next knowledge point (e.g., "Okay, now that we understand the Pythagorean theorem, let's look at one of its applications..."), or there is a long pause in teaching, or the topic is clearly changed. The time recorded at this moment is the end timestamp;

[0069] Assuming the start timestamp for the Pythagorean theorem is 10 minutes after the start of the course and the end timestamp is 25 minutes, then the explanation time for this knowledge point is 15 minutes.

[0070] In this embodiment, timestamps enable the system to perfectly synchronize the teacher's "teaching" and the student's "learning" on a unified timeline. This facilitates subsequent analysis to determine at which point in the explanation of a knowledge point the student demonstrated understanding.

[0071] In some embodiments, the analysis of each student's comprehension state includes:

[0072] Based on the student's image data, by analyzing their head posture and gaze direction, it can be determined whether they are attentive or not.

[0073] For students in the focused listening state, their comprehension state is analyzed using a micro-expression recognition model based on their facial images; the comprehension state includes understanding and not understanding.

[0074] Specifically, facial landmark detection technology can be used to establish a three-dimensional spatial orientation model of the head. For example, when a student's head is consistently facing the blackboard or the teacher's direction, and the tilt angle is within the normal range, it is judged as being attentive; when a student is consistently looking down (e.g., playing on a mobile phone) or frequently turning their head (talking to their neighbor), it is judged as not paying attention. Alternatively, eye positioning and gaze estimation algorithms can be used to determine the student's visual focus. For example, when a student's gaze is consistently focused on the blackboard, projection screen, or teacher, it is judged as being attentive; when their gaze frequently deviates from the teaching area (e.g., looking out the window or at the ground), it is judged as not paying attention.

[0075] For students identified as "focused on listening," their facial images (including eyebrows, eyes, and mouth areas) are captured and input into a pre-trained micro-expression recognition model.

[0076] This micro-expression recognition model is trained based on a convolutional neural network (CNN). The input is a facial feature image, and the output is a classification result of "understood" or "not understood". The training data contains a large number of student classroom micro-expression samples labeled as "understood" (such as nodding, raising the corners of the mouth, and bright eyes) and "not understood" (such as frowning, unfocused eyes, and pursing lips).

[0077] In this embodiment, by distinguishing between students who "did not listen and therefore did not understand" and those who "were focused but did not understand," the subsequent analysis of comprehension status is ensured to focus only on effective listening behavior, thus improving data reliability. Furthermore, this facilitates targeted interventions by teachers, such as reminding students who were not listening or providing supplementary explanations to students who were focused but did not understand.

[0078] In some embodiments, determining the comprehension time for each student for each knowledge point specifically includes:

[0079] Within the duration of an explanation of a knowledge point, analyze the student's comprehension only during the time period when the student is attentively listening.

[0080] If, during the duration of the explanation of a knowledge point, a student transitions from a state of non-listening to a state of focused listening and is identified as having understood the knowledge point, then the difference between the moment the student is identified as having understood the knowledge point and the moment the student enters a state of focused listening is the duration of the student's understanding of that knowledge point.

[0081] For example, a student is not paying attention at the beginning of a lesson on a certain knowledge point (minutes 10-13) (playing with a pen with their head down), but starts paying attention at the 13th minute, and is identified by the system as having "understood" the concept through micro-expressions at the 18th minute. In this case, the effective learning time that the student actually spent understanding the knowledge point was 5 minutes.

[0082] If, during the explanation time for a particular knowledge point, a student is found to be not paying attention and their comprehension status remains unchanged, then the student's comprehension time for that knowledge point is marked as invalid data.

[0083] For example, a student is not paying attention for 10-20 minutes, but becomes attentive at 20 minutes. However, by the end of the explanation (25 minutes), the student still does not show any micro-expressions of understanding. This indicates that the student's effective learning time was only 5 minutes, and they failed to reach a state of "understanding." Therefore, it is unclear whether the student actually grasped the concept within the allotted explanation time, and this data cannot be used to evaluate teaching effectiveness; thus, it is excluded.

[0084] If a student remains attentive throughout the explanation of a knowledge point but does not understand it, the student's understanding time for that knowledge point will be recorded as the total explanation time, and this time will be marked.

[0085] For example, if a student remains attentive throughout the entire explanation of a knowledge point (let's say 15 minutes) but does not demonstrate any understanding, then their longest period of understanding (15 minutes) is marked (e.g., "not understood") to indicate that the student did not comprehend the knowledge point.

[0086] In this embodiment, by distinguishing between "focused / not listening" states, the comprehension time is calculated only based on the effective listening time, avoiding misjudging "lack of understanding due to not listening" as "the knowledge point is difficult", ensuring that the data truly reflects the students' ability to accept the knowledge points, so as to provide teachers with more objective teaching feedback information in the future.

[0087] In some embodiments, the assessment of the reasonableness of the explanation time for the knowledge points specifically includes:

[0088] For a given knowledge point, based on the time students spend understanding it and the time spent explaining it, we can calculate the percentage of students who can understand the knowledge point within the allotted time.

[0089] If the percentage of people who understand a knowledge point within the allotted time is lower than the first threshold, it is assessed that the allotted time for that knowledge point needs to be increased.

[0090] If the percentage of people who understand a knowledge point within the allotted time is higher than the second threshold, then the allotted time for that knowledge point can be shortened.

[0091] If the percentage of people who understand a knowledge point within the allotted time is between the first and second thresholds, then the allotted time for that knowledge point can be maintained.

[0092] Specifically, for example, suppose the explanation of the "quadratic formula" takes 15 minutes, and there are 40 students in the class. The understanding time and state of each student regarding this knowledge point have been determined. After excluding invalid data (i.e., data from students who did not listen), the valid statistical sample is 38 students. The first threshold can be set to 80%, and the second threshold can be set to 95%.

[0093] "Understanding within the explanation time" means that the student's understanding time is less than 15 minutes (the explanation time); if the student's understanding time is equal to 15 minutes (the same as the explanation time), it can be considered that the student has not understood.

[0094] Of the 38 valid samples: 18 students took less than 15 minutes to understand the material (e.g., 5 minutes, 8 minutes, 12 minutes), and 20 students took 15 minutes or more to understand it. This indicates that 18 students could understand the material within the allotted time, while the remaining 20 students could not. In this case, the percentage of students who understood the material within the allotted time was approximately 47.4%, which is less than the first threshold (80%). Therefore, the assessment is that "the allotted time for this knowledge point needs to be increased."

[0095] If the percentage of people who understand the knowledge within the allotted time (e.g., 97%) is greater than the second threshold (95%), it is assessed as "the allotted time for this knowledge point can be shortened".

[0096] If the percentage of people who understand the knowledge within the allotted time (e.g., 86%) is between the first threshold (80%) and the second threshold (95%), it is assessed as "the allotted time for this knowledge point can remain unchanged".

[0097] The first threshold and the second threshold can be determined according to actual needs.

[0098] This embodiment provides teachers with a clear direction for adjusting class time through specific proportional thresholds, replacing the vague feedback such as "speaking too fast / slowly" in the traditional sense.

[0099] In some embodiments, the student's knowledge point comprehension efficiency analysis includes:

[0100] An individual comprehension efficiency profile is generated for each student. The individual comprehension efficiency profile is composed of the student's standardized comprehension efficiency score across all knowledge points. For a knowledge point, the efficiency score is the ratio of the explanation time for that knowledge point to the student's comprehension time for that knowledge point.

[0101] Based on the individual comprehension efficiency profiles of all students, a clustering algorithm is used to divide all students into high, medium, and low comprehension efficiency groups, and teaching class assignment suggestions are output.

[0102] For example, a physics course contains three core knowledge points: "Newton's First Law" (15 minutes of explanation), "Work and Energy" (20 minutes of explanation), and "Conservation of Momentum" (18 minutes of explanation).

[0103] First, calculate the standardized comprehension efficiency score for each student on each knowledge point:

[0104] Student A's comprehension efficiency score is calculated as follows: Newton's First Law: 15 minutes / 5 minutes = 3.0; Work and Energy: 20 minutes / 8 minutes = 2.5; Conservation of Momentum: 18 minutes / 6 minutes = 3.0; Therefore, Student A's comprehension efficiency profile is: [3.0, 2.5, 3.0].

[0105] Student B's comprehension efficiency score calculation: Newton's First Law: 15 minutes / 12 minutes = 1.25; Work and Energy: 20 minutes / 18 minutes = 1.11; Conservation of Momentum: 18 minutes / 16 minutes = 1.125; Therefore, Student B's comprehension efficiency profile is: [1.25, 1.11, 1.125].

[0106] Then, the K-means clustering algorithm can be used to analyze the students' comprehension efficiency profile:

[0107] For example, clustering results:

[0108] High-efficiency group: Student A (Characteristics: all knowledge points comprehension efficiency scores are greater than or equal to 1.5)

[0109] Medium-efficiency group: Student B (Characteristics: all knowledge point comprehension efficiency scores are greater than 1 and less than 1.5)

[0110] Thus, it can be analyzed that student A's comprehension speed is significantly faster than the teaching pace, resulting in a large amount of "waiting time"; student B's comprehension speed is basically in sync with the teaching pace, but there is no extra time for digestion.

[0111] In this embodiment, class grouping suggestions are made based on comprehension efficiency, avoiding the limitations of traditional class grouping based on grades (for example, high-achieving students may understand slowly but have reviewed thoroughly), ensuring that students in the same class have similar cognitive pace, making it easier for teachers to unify the teaching progress, and providing accurate data support for differentiated teaching.

[0112] like Figure 2 As shown, one embodiment of this application also provides a teaching behavior analysis device based on artificial intelligence, the device comprising:

[0113] The knowledge point recognition module is used to identify the various knowledge points explained during the lecture based on the teacher's voice and determine the explanation time of each knowledge point.

[0114] The student comprehension analysis module is used to analyze the comprehension status of each student based on the image data of each student during the teaching process, and to determine the comprehension time of each student for each knowledge point.

[0115] The feedback information generation module is used to generate teaching feedback information based on the explanation time of each knowledge point and the understanding time of each student for each knowledge point; wherein, the teaching feedback information includes an assessment of the rationality of the explanation time of the knowledge points and an analysis of the efficiency of students' understanding of the knowledge points.

[0116] Each module in the above-mentioned device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0117] An embodiment of this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-described artificial intelligence-based teaching behavior analysis method.

[0118] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, NFC (Near Field Communication), or other technologies. The computer program, when executed by the processor, implements the aforementioned methods. The display screen can be an LCD screen or an e-ink display. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad mounted on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0119] An embodiment of this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described artificial intelligence-based teaching behavior analysis method.

[0120] An embodiment of this application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the above-described artificial intelligence-based teaching behavior analysis method.

[0121] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0122] In summary, this application provides an AI-based teaching behavior analysis system and method. By automatically identifying knowledge points and linking them to students' comprehension time, it provides teachers with objective and quantitative data support regarding the rationality of their explanation time allocation. This enables teachers to accurately identify which knowledge points are over-explained and which are under-explained, allowing them to scientifically optimize classroom time allocation in subsequent teaching and effectively improve classroom teaching efficiency. By analyzing each student's comprehension time for different knowledge points, classes can be grouped based on students' comprehension efficiency. Students with similar learning paces can be grouped to form a cohesive learning class, greatly facilitating teachers in maintaining a unified teaching pace and truly achieving "individualized instruction" and "precision teaching," thereby improving the relevance and effectiveness of teaching from the source.

[0123] 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 of other embodiments.

[0124] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and substitutions can be made without departing from the technical principles of the present invention, and these improvements and substitutions should also be considered within the scope of protection of the present invention.

Claims

1. A teaching behavior analysis method based on artificial intelligence, characterized in that, The method includes: Based on the teacher's voice, the system identifies each knowledge point explained during the lecture and determines the duration of each knowledge point's explanation. Based on the image data of each student during the teaching process, analyze the understanding status of each student and determine the understanding time of each student for each knowledge point. Based on the explanation time for each knowledge point and the understanding time for each student for each knowledge point, teaching feedback information is generated; wherein, the teaching feedback information includes an assessment of the reasonableness of the explanation time for each knowledge point and an analysis of the efficiency of students' understanding of the knowledge points.

2. The method according to claim 1, characterized in that, The system identifies various knowledge points explained during the lecture based on the teacher's voice, including: The teacher's lecture audio is processed to convert it into text, resulting in text information. Natural language analysis is performed on the text information to extract keywords, semantic topics, or teaching terms, and these are matched with a pre-set knowledge point database to identify each knowledge point.

3. The method according to claim 2, characterized in that, The determination of the explanation time for each knowledge point includes: Record the start timestamp when each knowledge point is first mentioned and the end timestamp when it is last mentioned; calculate the difference between the end timestamp and the start timestamp to obtain the explanation duration of the knowledge point.

4. The method according to claim 1, characterized in that, The analysis of each student's comprehension status includes: Based on the student's image data, by analyzing their head posture and gaze direction, it can be determined whether they are attentive or not. For students in the focused listening state, their comprehension state is analyzed using a micro-expression recognition model based on their facial images; the comprehension state includes understanding and not understanding.

5. The method according to claim 4, characterized in that, The determination of the comprehension time for each student for each knowledge point includes: Within the duration of an explanation of a knowledge point, analyze the student's comprehension only during the time period when the student is attentively listening. If, during the duration of the explanation of a knowledge point, a student transitions from a state of non-listening to a state of focused listening and is identified as having understood the knowledge point, then the difference between the moment the student is identified as having understood the knowledge point and the moment the student enters a state of focused listening is the duration of the student's understanding of that knowledge point. If, during the explanation time for a particular knowledge point, a student is found to be not paying attention and their comprehension status remains unchanged, then the student's comprehension time for that knowledge point is marked as invalid data. If a student remains attentive throughout the explanation of a knowledge point but does not understand it, the student's understanding time for that knowledge point will be recorded as the total explanation time, and this time will be marked.

6. The method according to claim 1, characterized in that, The assessment of the reasonableness of the explanation time for the knowledge points includes: For a given knowledge point, based on the time students spend understanding it and the time spent explaining it, we can calculate the percentage of students who can understand the knowledge point within the allotted time. If the percentage of people who understand a knowledge point within the allotted time is lower than the first threshold, it is assessed that the allotted time for that knowledge point needs to be increased. If the percentage of people who understand a knowledge point within the allotted time is higher than the second threshold, then the allotted time for that knowledge point can be shortened. If the percentage of people who understand a knowledge point within the allotted time is between the first and second thresholds, then the allotted time for that knowledge point can be maintained.

7. The method according to claim 1, characterized in that, The analysis of students' efficiency in understanding knowledge points includes: An individual comprehension efficiency profile is generated for each student. The individual comprehension efficiency profile is composed of the student's standardized comprehension efficiency score across all knowledge points. For a knowledge point, the efficiency score is the ratio of the explanation time for that knowledge point to the student's comprehension time for that knowledge point. Based on the individual comprehension efficiency profiles of all students, a clustering algorithm is used to divide all students into high, medium, and low comprehension efficiency groups, and teaching class assignment suggestions are output.

8. A teaching behavior analysis device based on artificial intelligence, characterized in that, The device includes: The knowledge point recognition module is used to identify the various knowledge points explained during the lecture based on the teacher's voice and determine the explanation time of each knowledge point. The student comprehension analysis module is used to analyze the comprehension status of each student based on the image data of each student during the teaching process, and to determine the comprehension time of each student for each knowledge point. The feedback information generation module is used to generate teaching feedback information based on the explanation time of each knowledge point and the understanding time of each student for each knowledge point; wherein, the teaching feedback information includes an assessment of the rationality of the explanation time of the knowledge points and an analysis of the efficiency of students' understanding of the knowledge points.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images