Teaching management method and equipment based on artificial intelligence, and medium
Through artificial intelligence management methods, new teachers can design curriculum teaching more effectively, solve the problem of low teaching efficiency, and achieve personalized teaching and improve teaching effects.
Patent Information
- Application Number
- CN202411334264.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-24
- Publication Date
- 2025-01-28
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
When designing course teaching, new teachers lack teaching experience, it is difficult for them to design the most suitable teaching method, resulting in low teaching efficiency.
Using artificial intelligence -based teaching management methods, through obtaining static teaching resources, assessing its reliability, combining static and dynamic teaching resources to generate teaching resources groups, and to screen the most adaptive teaching resource group according to students as the curriculum examples Essence
It has improved the understanding of students 'learning needs and preferences, realized personalized teaching, stimulated students' interest in learning, improved teaching effects, and provided new teachers with a reference for teaching to learn to improve teaching efficiency.
Smart Images
Figure CN119379501A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of data processing for management purposes, and specifically to a teaching management method, device and medium based on artificial intelligence. Background Art
[0002] The design of courses and the arrangement of teaching content usually rely on teachers' personal teaching experience and professional judgment. However, with the continuous expansion and diversification of the education system, as well as the continuous emergence of new technologies and new knowledge fields, this approach of relying on human experience has gradually revealed its limitations, especially when facing new teachers. For new teachers with relatively little teaching experience, when facing course teaching, due to the lack of certain teaching references, it is often difficult to design the most appropriate teaching method for different teaching contents, resulting in relatively low teaching efficiency. Summary of the invention
[0003] In order to solve the above problems, this application proposes a teaching management method based on artificial intelligence, including: For a target course, in response to a user's operation instruction, a static teaching resource corresponding to the target course is obtained from a preset data channel; wherein the static teaching resource refers to teaching content in written or digital form; Determine the dependency type between the static teaching resource and the target course, and evaluate the reliability between the static teaching resource and the target course according to the dependency type to obtain the resource reliability corresponding to the static teaching resource; When the resource reliability is greater than a preset value, at least one teaching resource group corresponding to the target course is generated according to the combination of the static teaching resources and the preset dynamic teaching resources; wherein the dynamic teaching resources are used to represent different forms of teaching content, and the forms of expression include lecturing, interaction, demonstration and inspection; After completing the study of the teaching resource group, collecting feedback information of the students for each teaching resource group, and determining the adaptability of the students to the teaching resource group according to the feedback information; A target teaching resource group with the largest corresponding adaptability is selected from the teaching resource group as a course example of the target course, and the course example is uploaded to a course example library so that teachers can implement the teaching of the target course according to the course example library.
[0004] In one example, for a target course, a static teaching resource corresponding to the target course is obtained from a preset data channel; wherein the static teaching resource refers to teaching content in written or digital form; Determine the dependency type between the static teaching resource and the target course, and evaluate the reliability between the static teaching resource and the target course according to the dependency type to obtain the resource reliability corresponding to the static teaching resource; When the resource reliability is greater than a preset value, at least one teaching resource group corresponding to the target course is generated according to the combination of the static teaching resources and the preset dynamic teaching resources; wherein the dynamic teaching resources are used to represent different forms of teaching content, and the forms of expression include lecturing, interaction, demonstration and inspection; After completing the study of the teaching resource group, collecting feedback information of the students for each teaching resource group, and determining the adaptability of the students to the teaching resource group according to the feedback information; A target teaching resource group with the largest corresponding adaptability is selected from the teaching resource group as a course example of the target course, and the course example is uploaded to a course example library so that teachers can implement the teaching of the target course according to the course example library.
[0005] In one example, after obtaining the resource reliability corresponding to the static teaching resource, the method further includes: According to the course progress of the target course, determine the standard preset value corresponding to the static teaching resource, and try to obtain the course example corresponding to the target course from the course example library; In the case that the course examples are not obtained or the number of the course examples is less than the preset number, the standard preset value is corrected according to the number of the course examples and the preset proportional factor, so as to obtain the corresponding preset value according to the corrected standard preset value; In a case where the number of the course examples is not less than the preset number, the standard preset value is used as the preset value of the resource reliability.
[0006] In one example, students' feedback information for each teaching resource group is collected, including: For each teaching resource group, video data generated during the teaching process is collected, and the video data is recognized by a pre-trained support vector machine to recognize the teaching actions generated by the teacher during the teaching process, and to recognize the dynamic teaching resources corresponding to the teaching actions; For different dynamic teaching resources, respectively collect video data within the corresponding time range as classroom performance data corresponding to the dynamic teaching resources; Identify the classroom performance data to identify feedback actions generated by the target student within a corresponding time range, and identify multiple action categories included in the feedback actions, where the action categories include: standing, sitting, leaning forward, raising the head, lowering the head, reading, writing, and communicating; Each of the dynamic teaching resources is combined with an action category within a corresponding time range to obtain feedback information corresponding to the teaching resource group.
[0007] In one example, students' feedback information for each teaching resource group is collected, including: For each teaching resource group, video data generated during the teaching process is collected, and the video data is recognized by a pre-trained support vector machine to recognize the teaching actions generated by the teacher during the teaching process, and to recognize the dynamic teaching resources corresponding to the teaching actions; For different dynamic teaching resources, respectively collect video data within the corresponding time range as classroom performance data corresponding to the dynamic teaching resources; Identify the classroom performance data to identify feedback actions generated by the target student within a corresponding time range, and identify multiple action categories included in the feedback actions, where the action categories include: standing, sitting, leaning forward, raising the head, lowering the head, reading, writing, and communicating; Each of the dynamic teaching resources is combined with an action category within a corresponding time range to obtain feedback information corresponding to the teaching resource group.
[0008] In one example, after determining whether the feedback information is a positive performance behavior, the method further includes: If it is determined that the active performance behavior corresponds to the dynamic teaching resource as interaction, then determining the expression type of the student's corresponding facial expression, the expression type including positive expression and negative expression, the positive expression at least including: excitement, calmness, the negative expression at least including: tension, fear; If the expression type of the student belongs to the positive expression, confirming that the positive expression behavior corresponding to the dynamic teaching resource is interactive is valid; Otherwise, the positive performance behavior is confirmed to be invalid.
[0009] In one example, a target teaching resource group with the largest degree of adaptation is selected from the teaching resource group as a course example of the target course, specifically including: For the target teaching resource group, determine the teaching form corresponding to the target teaching resource group, wherein the teaching form includes: teacher lecture form, interactive question form, experiment form, and self-study form, which respectively correspond to the lecture, interaction, demonstration, and inspection in the dynamic teaching resources; According to the performance score corresponding to the action category, the appearance time of the corresponding teaching form in the course example is determined.
[0010] In one example, a target teaching resource group with the largest degree of adaptation is selected from the teaching resource group as a course example of the target course, specifically including: For the target teaching resource group, determine the teaching form corresponding to the target teaching resource group, wherein the teaching form includes: teacher lecture form, interactive question form, experiment form, and self-study form, which respectively correspond to the lecture, interaction, demonstration, and inspection in the dynamic teaching resources; According to the performance score corresponding to the action category, the appearance time of the corresponding teaching form in the course example is determined.
[0011] On the other hand, the present application also proposes a teaching management device based on artificial intelligence, comprising: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the artificial intelligence-based teaching management method as described in the above example.
[0012] On the other hand, the present application also proposes a non-volatile computer storage medium storing computer executable instructions, wherein the computer executable instructions are configured as: the artificial intelligence-based teaching management method described in the above example.
[0013] The teaching management method based on artificial intelligence proposed in this application can bring the following beneficial effects: By collecting students' feedback on each teaching resource group and determining the students' adaptability to the teaching resource group, teachers can more accurately understand students' learning needs and preferences, and then select the target teaching resource group with the greatest adaptability as a course example. This personalized teaching method helps to stimulate students' interest and enthusiasm in learning and improve teaching effectiveness.
[0014] Uploading the selected course examples to the course example library can provide teachers with teaching references, help new teachers quickly adapt to teaching work, reduce the time of exploration and trial and error, and improve teaching efficiency. BRIEF DESCRIPTION OF THE DRAWINGS
[0015] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings: Figure 1 A flowchart of an artificial intelligence-based teaching management method provided in an embodiment of the present application; Figure 2 A structural diagram of an artificial intelligence-based teaching management device provided in an embodiment of the present application. DETAILED DESCRIPTION
[0016] In order to make the purpose, technical solution and advantages of the present application clearer, the technical solution of the present application will be clearly and completely described below in combination with the specific embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present application.
[0017] The technical solutions provided by various embodiments of the present application are described in detail below in conjunction with the accompanying drawings.
[0018] like Figure 1 As shown, the embodiment of the present application provides a teaching management method based on artificial intelligence, including: S101: For a target course, in response to a user's operation instruction, obtain static teaching resources corresponding to the target course from a preset data channel; wherein the static teaching resources refer to teaching content in digital form.
[0019] The embodiments of the present application are mainly aimed at the teaching process in primary and secondary schools and high schools. At this stage, students are relatively concentrated and the course learning progress is relatively uniform. By comprehensively evaluating course feedback, teaching examples can be formed for teachers to refer to.
[0020] A target course refers to a learning task of a subject at a certain stage. For example, the Chinese subject needs to complete the study of Text 1 and Text 2 within a week, so completing the study of Text 1 and Text 2 can be called a target course. The duration of the learning stage and the amount of learning tasks of the target course are not limited in this application and can be set according to actual learning needs. For the target course, teachers not only need to screen specific teaching content during the teaching process, but also need to adaptively adjust the teaching content or teaching form according to different teaching content and students' teaching feedback information. Usually, in the process of screening teaching content, teachers will obtain static teaching resources of the target course from preset data channels, and write lesson plans or courseware according to different organizational forms of static teaching resources, and finally form a complete teaching plan. Therefore, when a user (teacher) queries static teaching resources through an online system, the server can respond to the user's operation instructions and obtain the corresponding resources. The static teaching resources mentioned in the above content refer to teaching content that exists on a static carrier (a carrier in digital form). Electronic textbooks and teaching materials, videos or written materials provided by educational resource websites, etc. are all called static teaching resources.
[0021] S102: Determine the dependency type between the static teaching resources and the target course, and evaluate the reliability between the static teaching resources and the target course according to the dependency type to obtain the resource reliability corresponding to the static teaching resources.
[0022] As the first step in teaching, static teaching resources are of great importance to teaching effectiveness. If the static teaching resources are set unreasonably or inaccurately, it will seriously affect students' learning effect on the target course. Therefore, after collecting static teaching resources, it is necessary to evaluate their reliability. By analyzing the dependency type between static teaching resources and target courses, the resource reliability corresponding to the static teaching resources can be determined. The dependency type is used to characterize the degree of support of users for the acquired static teaching resources. According to the order of decreasing support, it is divided into three types: strong active dependency, weak active dependency and passive dependency. Among them, active dependency refers to static teaching resources actively selected by users, while passive dependency refers to users who are not involved in the screening but are directly selected into the teaching content due to the binding relationship between different static teaching resources. For example, when a user downloads the course content of the target course on an educational resource website, the course content is directly bound to supporting exercises, so the course content and supporting exercises will be downloaded together. This process does not require users to perform additional content screening.
[0023] In one embodiment, for each static teaching resource collected, it is determined whether it is bound to other acquired static teaching resources. This process can be achieved by checking whether the static teaching resource contains an associated resource identifier pointing to other resources. The associated resource identifier can be a link, a reference, a directory item, etc., which is used to indicate the binding relationship between different static teaching resources. If there is an associated resource identifier in a static teaching resource, it indicates that there are other static teaching resources bound to it. At this time, the associated resource identifier needs to be identified. The associated resource identifier can point to the passively bound static teaching resource. Then, the static teaching resource located according to the associated resource identifier is the teaching content that is passively screened out. In other words, the static teaching resource has not actually been actively screened by the user, and the type of dependency between it and the target course is passive dependency.
[0024] If static teaching resources are not bound to other static teaching resources, it means that all static educational resources are obtained through active screening by users. However, there are certain differences in the operation of users when screening static educational resources. Some teaching content may be carefully screened by users, while some teaching content is relatively simple or widely used in the subject area, or is of low importance, so users only screen it out after a simple browsing. Therefore, based on the user's screening operation for static educational resources, the dependency types of static educational resources can be divided more finely.
[0025] Specifically, the operation event information for static teaching resources is obtained, and the operation event information contains sufficient information to identify the user's operation behavior, operation time, and the specific resources involved. According to the operation event information, the user's operation process and operation attributes for static teaching resources can be determined. The operation process refers to a number of continuous operation behaviors of the user on the resource within a period of time, such as inputting keywords, browsing static teaching resource A, browsing static teaching resource B, downloading resources, modifying resources, etc. The operation attributes include the operation time of each operation behavior. The operation behavior of "viewing resource details" usually indicates that the user has an in-depth understanding of the resource, and the operation time of this operation behavior can be used to evaluate the user's attention and investment in the resource. Therefore, after obtaining the operation process and operation attributes, the specified static teaching resources with the operation behavior of viewing resource details and the corresponding operation time greater than the preset operation time can be screened out from all static teaching resources. For the specified static teaching resources screened out, since the user not only views the resource details but also invests a long time, it can be considered that there is a strong active dependency relationship between these resources and the target course, that is, these resources are more important to the target course, and because they have been carefully screened, they are relatively more reliable. As for other static teaching resources whose resource details have not been viewed or whose viewing time has not reached the preset operation time, although they may have been visited by users, it is not enough to prove that users have a deep understanding of these resources. Even if they are filtered into the teaching content, their reliability is slightly lower because they have not been carefully screened.
[0026] It should be noted that for other static teaching resources whose resource details have not been viewed or whose corresponding operation time is less than the preset operation time, although they have not been carefully screened, there may be the following reasons that lead to this situation, for example, the resources are widely used in the teaching of the target course, the teaching content is relatively simple and does not require in-depth browsing, and the resources are only used for teaching links with low importance such as course introduction. Therefore, for such static teaching resources, even if the dependency type between them and the target course is determined to be a weak active dependency based on the operation event information, this determination result needs to be corrected. Therefore, after screening out static teaching resources with weak active dependencies, it is necessary to determine whether the static teaching resources have special attributes, which are the reasons listed above. If there are special attributes, the user already has a high degree of support for the static teaching resource, and it can be directly identified as having a strong active dependency relationship with the target course. At this time, the dependency type needs to be corrected.
[0027] Classifying dependency types actually provides a quantifiable basis for evaluating the reliability of teaching resources. By distinguishing different dependency types, the reliability of static teaching resources can be more accurately evaluated. Reliability can effectively give play to the leading role of teachers in screening teaching resources, so as to select teaching resources that better meet actual teaching needs. In this way, it can provide more accurate underlying data support for subsequent teaching and improve teaching quality.
[0028] In one embodiment, after determining the dependency type corresponding to each static teaching resource, the resource reliability of the obtained static teaching resource needs to be determined according to the dependency type. First, the corresponding dependency type weight is assigned to different dependency types, and the dependency type weight is reduced in the order of strong active dependency, weak active dependency and passive dependency according to the different support levels of the user. In other words, for the static teaching resources obtained by the user through active screening behavior, it can be shown that the teacher has a high degree of recognition and has been carefully screened, and the teaching effect it can play in the teaching process should be relatively strong. For the static teaching resources obtained through passive binding, the teacher did not actively participate in the screening process of such resources. Since the teacher did not properly screen the content adaptability and rationality, in the actual teaching process, the corresponding reliability of such static teaching resources is slightly lower. It should be noted that the specific value of the dependency type weight can be set according to the actual situation and teaching needs. For example, it can be set by using a proportional setting method such as 1.0 (strong active dependency), 0.5 (weak active dependency), 0.2 (passive dependency), or determined according to expert evaluation and actual teaching feedback.
[0029] Each static teaching resource has a different resource base reliability according to the different data channels it is obtained from. Data channels include at least one or more of the following: textbooks, teaching materials, educational resource websites, official government websites, and internal school resource libraries. Resource base reliability is used to characterize the authority and credibility of data channels. This has nothing to do with the user's acquisition method and only depends on which data channel the static teaching resource comes from. For example, textbooks and official government websites usually have high authority and credibility, so their resource base reliability is higher; while educational resource websites and internal school resource libraries may have relatively low resource base reliability due to diverse sources and uneven quality. The specific value of resource base reliability also needs to be set based on expert evaluation or actual teaching feedback.
[0030] After clarifying the resource base reliability corresponding to each static teaching resource, the resource base reliability is weighted according to the dependency type weight corresponding to the dependency type to which it belongs, so as to obtain the resource reliability corresponding to the static teaching resource. This method can comprehensively consider the reliability of resources in teaching (dependency type weight) and the authority of the source (resource base reliability), so as to more accurately evaluate the reliability of teaching resources.
[0031] The above process describes the specific process of calculating the resource reliability corresponding to a static teaching resource. In the actual teaching process, teachers usually need to obtain multiple static teaching resources. Therefore, when there are multiple static teaching resources, it is necessary to conduct an overall evaluation of the resource reliability of multiple static teaching resources. At this time, it is necessary to calculate the mean of the resource reliabilities corresponding to all static teaching resources. This mean is the resource reliability corresponding to the static teaching resources finally obtained.
[0032] S103: When the resource reliability is greater than a preset value, at least one teaching resource group corresponding to the target course is generated according to the combination of static teaching resources and preset dynamic teaching resources; wherein the dynamic teaching resources are used to represent different forms of teaching content, including lectures, interactions, demonstrations and inspections.
[0033] After obtaining the resource reliability, a preset value for limiting the lower limit of resource reliability needs to be set to ensure that only static teaching resources with a certain reliability can be applied to teaching, which helps reduce the teaching risks brought by the use of unreliable resources and ensure the teaching quality and learning effect. According to the preset value, resources with resource reliability greater than the preset value can be screened out from all available static teaching resources, and these resources will serve as the basis for building teaching resource groups. Dynamic teaching resources are used to represent different forms of teaching content, such as lectures, interactions, demonstrations, and inspections. Each dynamic teaching resource has its unique teaching value and applicable scenarios. For example, lectures are suitable for conveying new knowledge, interactions can improve students' participation, demonstrations help to intuitively demonstrate skills, and inspections make it easier for teachers to understand students' learning situations. Therefore, according to the applicable teaching methods of static teaching resources, static teaching resources and dynamic teaching resources can be dynamically combined, so as to generate at least one teaching resource group corresponding to the target course according to the obtained combination method. Each group can contain a combination of multiple static teaching resources and dynamic teaching resources to meet the needs of different teaching links. For example, the static teaching resources required for a certain target course include static teaching resource 1 and static teaching resource 2. The dynamic teaching resources that can be used by static teaching resource 1 are lectures and demonstrations, and the dynamic teaching resources that can be used by static teaching resource 2 are interactions. By freely combining them, the final teaching resource groups obtained are of two types. The first type is static teaching resource 1: lectures, static teaching resource 2: interactions; the second type is static teaching resource 1: demonstrations, static teaching resource 2: interactions.
[0034] In one embodiment, the preset value mentioned in the above content for determining whether the static teaching resources can be put into the teaching design can be adjusted in real time according to whether there are existing course examples. Generally speaking, if there is no course example for a certain target course or the number of course examples is small, then the teacher will lack examples for reference when designing the teaching resource group. At this time, in order to expand the course examples as soon as possible and improve the richness of the course example library, it is necessary to appropriately lower the standard of the static teaching resources. In this way, even if the static teaching resources selected by the teacher are not very safe, in order to add new course examples, the static teaching resources should be put into the teaching design and their teaching effect should be tested. Therefore, the preset value needs to be corrected according to the existing course examples in the course example library.
[0035] Specifically, according to the course progress and general teaching requirements of the target course, determine a standard preset value for the reliability of static teaching resources. This value can be used as a benchmark for subsequent comparisons and adjustments. Generally speaking, as the course progresses, the teaching difficulty gradually increases, the reliability of the required static teaching resources becomes higher, and the standard preset value should also increase accordingly. Next, try to obtain course examples corresponding to the target course from the course example library. If the number of course examples obtained is not less than the preset number, it means that there is sufficient reference information. At this time, the standard preset value can be directly used as the preset value for resource reliability. If no course examples are obtained, or the number of course examples obtained is less than the preset number (indicating that there is not enough reference information), the standard preset value needs to be corrected based on the number of course examples and a preset proportional factor. This can be achieved specifically through the following formula:
[0036] in, Indicates the corrected preset value. Indicates the standard preset value, Indicates the preset quantity, represents the number of course examples, and P represents the proportional factor, which is a positive number.
[0037] After correcting the marked preset value in the above way, if the original number of course examples is very small, the corrected preset value may be reduced accordingly, so that relatively more static teaching resources will be involved in the subsequent teaching process, thereby forming course examples added to the course example library, providing sufficient reference information for subsequent teaching of other teachers.
[0038] S104: After completing the study of the teaching resource groups, collect students' feedback information on each teaching resource group, and determine the students' adaptability to the teaching resource groups based on the feedback information.
[0039] After teaching students according to different teaching resource groups, we collect students' feedback information for each teaching resource group and determine the students' adaptability to the teaching resource group based on the feedback information. The adaptability is used to evaluate the teaching effect of the teaching resource group. The higher the adaptability, the more students adapt to the organization of the teaching resource group and the more positive feedback they give overall.
[0040] Feedback information needs to be obtained with the help of video data. Video data can be collected by a camera installed in the classroom. Support vector machine (SVM) is a binary classification model that can achieve similar results in a smart campus compared to a deep learning model, and requires lower computing power. The number of students in a school is huge. If a deep learning model is used, the computing power consumed, the time spent in the early training process, and the amount of training data are relatively large. Therefore, in the embodiment of the present application, a support vector machine is used to process the video data accordingly, so as to obtain feedback information corresponding to the target students.
[0041] Feedback information is mainly expressed by the student's actions. The main function of the support vector machine is to identify the student's actions through video data and use the student's actions as feedback information. In the actual processing process, the corresponding actions can also be performed by neural networks, deep learning models or other methods instead of support vector machines, which are not limited here.
[0042] Specifically, the video data generated during the teaching process is collected and recognized by the support vector machine to identify the teaching actions generated by the teacher during the teaching process and the dynamic teaching resources corresponding to the teaching actions. The teaching actions are pre-classified into multiple categories (referred to as dynamic teaching resources herein), which may include: lecturing, interaction, demonstration, and inspection.
[0043] For lectures, teachers need to open their mouths to lecture, and there will be corresponding body movements, such as pointing to the screen, picking up textbooks, etc., and the position is relatively fixed, the eyes will patrol within a certain range, and the duration of this action is often long. For interaction, it can be accompanied by asking students questions and listening to their answers. Compared with lectures, the teacher's eyes and the frequency of opening their mouths will change accordingly. The eyes will be fixed on a single student, and the speaking time will also be shortened, leaving time for students to answer. For demonstration, it is mainly accompanied by the teacher's body movements, and the amplitude of the body movements is often large, and it is necessary to hold the corresponding models, teaching aids, experimental instruments, etc. For patrols, it is easy to appear in self-study classes or when teaching tasks are assigned (for example, exercises are issued), and then wait for students to complete the tasks. At this time, the teacher usually does not open his mouth, and the movement amplitude is very small. Sitting on the podium, or performing a certain action in a loop (for example, stepping on the spot, walking around the classroom for inspection).
[0044] For different dynamic teaching resources, the video data within the corresponding time range is collected as the classroom performance data corresponding to the dynamic teaching resources. According to different dynamic teaching resources, the time in the teaching process is divided into multiple time ranges. In each time range, the dynamic teaching resources are unified, and the video data within the time range is used as the classroom performance data of the dynamic teaching resource.
[0045] Classroom performance data is essentially still video data, which includes target students in addition to teachers. Therefore, classroom performance data is identified to identify the feedback actions generated by target students within the corresponding time range (referring to students' feedback on teachers' teaching actions). Similarly, multiple action categories contained in feedback actions are identified, including: standing, sitting, leaning forward, raising head, lowering head, reading, writing, and communicating.
[0046] Since the embodiment of the present application is mainly aimed at the generation of teaching examples, unlike dynamic teaching resources, the division of action categories is more detailed. It simply recognizes the students' body movements and does not conduct a deeper recognition of the meaning behind the movements. Therefore, the recognition accuracy of action categories is higher.
[0047] At this time, each dynamic teaching resource is combined with the action category within the corresponding time range to obtain the feedback information corresponding to the dynamic teaching resource. Generally speaking, there are multiple action categories. For example, when a teacher is giving a lecture, the dynamic teaching resource identified is a lecture. At this time, the student is sitting on the table and looking up to listen to the class. At this time, the action categories include: sitting, standing, and looking up. At this time, the generated feedback information is: (lecture; sitting, standing, and looking up).
[0048] Among them, video data or classroom performance data are essentially video data, so their recognition process can be recognized by a support vector machine. In actual work, since the input of the recognition process of video data or classroom performance data is video data and the output is action categories, in order to reduce costs, only one support vector machine can be trained for recognition. Of course, since the recognition targets and recognition accuracy of the two are different, in order to improve the recognition accuracy, multiple support vector machines can be trained, which are used for the recognition of dynamic teaching resources and action categories respectively.
[0049] Specifically, when identifying dynamic teaching resources or action categories, the video data or classroom performance data are split and extracted to obtain multiple frame images, and the frame images are preprocessed (for example, image denoising, image enhancement, size adjustment, etc.) to obtain a standard image.
[0050] Based on the pre-trained support vector machine, the standard image is recognized to extract the image features (such as color, texture, shape, etc.) of the target person's area based on the scale-invariant feature transform SIFT. The target persons include teachers and target students, which correspond to the dynamic teaching resources and action categories corresponding to this recognition respectively.
[0051] Perform feature preprocessing (e.g., feature dimension reduction) on the image features to obtain standard features, and determine the key points of the target person in the frame image based on the standard features. The key points include: head, face, torso, shoulders, and hands. For the key points of the face, you can also continue to identify the key points contained therein, such as eyes, mouth, etc., based on demand.
[0052] Based on the changes between the corresponding key points in multiple frame images (such as position changes, distance changes, angle changes, speed changes, shape changes, etc.), the action features of the target person can be captured. For example, the shape changes of the mouth key points can be identified to determine whether the teacher is speaking, and the position changes of the student's chest key points can be identified to determine whether the student is standing or sitting. At this time, the corresponding action category is determined by the action features.
[0053] Furthermore, after the feedback information is determined, it can be determined whether the feedback information is a positive performance behavior based on a preset mapping relationship table and the dynamic teaching resources and action categories appearing in the feedback information.
[0054] When the teacher performs different teaching actions, the students give different feedback actions, which represent the students' current corresponding status. These statuses can include positive and negative behaviors, indicating whether the students are interested in the teacher's current teaching action. Of course, if the student's action category is identified as leaning forward, that is, lying on the desk, it can be directly considered as a negative behavior.
[0055] The mapping relationship table is set based on expert experience, and the positive performance behaviors include: When the dynamic teaching resource is a lecture, the action categories include: sitting, raising the head, indicating that the students are listening carefully, or sitting, lowering the head, reading, or sitting, lowering the head, writing, indicating that the students are reading books or taking notes according to the course they are listening to. Of course, if students sit, lower their heads, and communicate, it can be considered that the students are secretly chatting and are not listening carefully, which is considered a negative behavior.
[0056] When the dynamic teaching resource is interactive, the action categories include: standing and communicating, which means that the student is actively answering questions, or sitting and looking up, which means that the student is sitting and listening carefully. Of course, if the student sits and lowers his head, it can be considered that the student dares not look directly at the interactive behavior, and that the student is afraid of or not interested in this teaching action, which is considered a negative behavior.
[0057] When the dynamic teaching resource is a demonstration, the action categories include: sitting and standing, raising the head, indicating that the student is sitting and standing and observing the demonstration carefully. Similarly, if the student sits and stands and lowers the head, it can be considered that the student is not interested in the demonstration and is considered to be a negative behavior.
[0058] When the dynamic teaching resource is patrolling, the action categories include: sitting, bowing, reading, or sitting, bowing, writing, indicating that the students are completing the task or reading seriously. Of course, if the students stand, sit, bow, and communicate at this time, it is considered a negative behavior.
[0059] Based on this, the target students' active performance behaviors under each dynamic teaching resource can be determined based on the feedback information, thereby determining the target students' interest level in each dynamic teaching resource, that is, in the teacher's various teaching actions.
[0060] In addition, if it is determined that the active performance behavior corresponds to the dynamic teaching resources as interaction, at this time, it is also possible that the students do not like this teaching action, but are forced to be called upon by the teacher to answer questions to achieve interaction.
[0061] At this time, the key points corresponding to the face of the target student extracted based on the support vector machine are determined, and the face area corresponding to the target student is obtained according to the key points. Of course, the corresponding key points of the mouth, eyes, etc. can also be obtained.
[0062] Based on Local Binary Patterns (LBP), facial features corresponding to the facial area are extracted, and based on a support vector machine (similarly, the support vector machine can be additionally trained, or the same support vector machine as used in the video recognition above), the expression type of the facial expression corresponding to the target student is determined through changes between the corresponding facial features in multiple frame images (for example, angle changes, speed changes, shape changes), and the expression types include positive expressions and negative expressions. Positive expressions include at least: excitement and calmness, and negative expressions include at least: tension and fear.
[0063] If the target student's expression type is a positive expression, it can be considered that the student is actively answering, or is interested in answering the question this time, and the positive expression behavior corresponding to the dynamic teaching resource for interaction is confirmed to be valid. Otherwise, it is considered that the student is forced to answer, and the positive expression behavior is confirmed to be invalid.
[0064] For teaching methods, determine the number of occurrences and duration of the positive performance behaviors corresponding to each dynamic teaching resource. At this time, a fixed duration can be used as a collection time period. For example, one day can be used as a fixed duration to determine the number of occurrences and duration of the target students' positive performance behaviors within the duration.
[0065] According to the number of occurrences and / or duration, the performance score of the target student in each dynamic teaching resource is determined, and the performance score is normalized to obtain the student's adaptability to the current teaching resource group. Among them, the performance score corresponding to lectures, demonstrations, and inspections is positively correlated with the duration. The longer the duration, the higher the performance score. The performance score corresponding to interaction has nothing to do with the duration, but is mainly related to the number of occurrences. Therefore, it is positively correlated with the number of occurrences. The more times it appears, the higher the performance score.
[0066] In the dynamic teaching resources, the specified action categories whose performance scores exceed the preset scores are determined. The higher the performance scores are, the better the target students perform under the teaching action and the more interested they are in the teaching action.
[0067] At this time, the teaching form corresponding to the specified action category is determined, wherein the teaching forms include: teacher lecture form, interactive question form, experimental form, and self-study form, which respectively correspond to lectures, interactions, demonstrations, and inspections in dynamic teaching resources.
[0068] In addition, the duration of the corresponding teaching form in the personalized teaching environment can be determined according to the performance score corresponding to the specified action category. The higher the performance score, the longer the duration of the corresponding dynamic teaching resource will appear when teaching the target student. This teaching form can serve as a guide. When teaching the target student, the instructor can refer to this teaching form to prepare lessons and perform teaching tasks.
[0069] S105: Filter out a target teaching resource group with the largest degree of adaptation from the teaching resource group as a course example of the target course, and upload the course example to the course example library so that the teacher can teach the target course according to the course example library.
[0070] After obtaining the fitness corresponding to different teaching resource groups, the target resource group with the largest fitness is selected, and the teaching plan is designed according to the combination of the target resource group, thereby forming a course example for the target course. The course example can be uploaded to the course example library as an excellent teaching model, so that other teachers can understand which static teaching resources should be obtained and how to combine these static teaching resources with dynamic teaching resources based on the course examples in the course example library when facing new courses, so as to improve teaching effects.
[0071] It should be noted that in order to improve the referenceability of the course example library, the number of course examples that can be stored for each target course is limited. When the maximum storage number is exceeded, the course example library needs to be optimized according to the adaptability of the course examples. By deleting the course examples with the lowest adaptability, the accumulation of inefficient content can be reduced, making the course example library more streamlined and efficient.
[0072] like Figure 2 As shown, the embodiment of the present application also proposes a teaching management device based on artificial intelligence, including: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the artificial intelligence-based teaching management method as described in any of the above embodiments.
[0073] The embodiment of the present application also proposes a non-volatile computer storage medium storing computer executable instructions, wherein the computer executable instructions are configured as: the teaching management method based on artificial intelligence described in any of the above embodiments.
[0074] The above is only the embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the present application.
Claims
1. A teaching management method based on artificial intelligence, characterized in that: include: For a target course, in response to a user's operation instruction, a static teaching resource corresponding to the target course is obtained from a preset data channel; wherein the static teaching resource refers to teaching content in digital form; Determine the dependency type between the static teaching resource and the target course, and evaluate the reliability between the static teaching resource and the target course according to the dependency type to obtain the resource reliability corresponding to the static teaching resource; When the resource reliability is greater than a preset value, at least one teaching resource group corresponding to the target course is generated according to the combination of the static teaching resources and the preset dynamic teaching resources; wherein the dynamic teaching resources are used to represent different forms of teaching content, and the forms of expression include lecturing, interaction, demonstration and inspection; After completing the study of the teaching resource group, collecting feedback information of the students for each teaching resource group, and determining the adaptability of the students to the teaching resource group according to the feedback information; A target teaching resource group with the largest corresponding adaptability is selected from the teaching resource group as a course example of the target course, and the course example is uploaded to a course example library so that teachers can implement the teaching of the target course according to the course example library.
2. According to claim 1, a teaching management method based on artificial intelligence is characterized in that: Determining the dependency type between the static teaching resource and the target course specifically includes: For each static teaching resource, determining whether the static teaching resource is bound to other acquired static teaching resources; If yes, identifying the associated resource identifier in the static teaching resource, and setting the dependency type between the static teaching resource located by the associated resource identifier and the target course as passive dependency; If not, obtain the operation event information for the static teaching resource, and determine the corresponding operation process and operation attributes according to the operation event information; wherein the operation process is composed of a number of continuous operation behaviors, and the operation attributes include the operation time of each operation behavior; According to the operation process and the operation attributes, the designated static teaching resources having the operation behavior of viewing resource details and the corresponding operation time greater than the preset operation time are screened out from the static teaching resources, and the dependency type between the designated static teaching resources and the target course is determined to be a strong active dependency; For other static teaching resources for which there is no operation behavior of viewing resource details or the corresponding operation time is less than the preset operation time, it is determined that the dependency type between the other static teaching resources and the target course is a weak active dependency.
3. The teaching management method based on artificial intelligence according to claim 1, characterized in that: After obtaining the resource reliability corresponding to the static teaching resource, the method further includes: According to the course progress of the target course, determine the standard preset value corresponding to the static teaching resource, and try to obtain the course example corresponding to the target course from the course example library; In the case that the course examples are not obtained or the number of the course examples is less than the preset number, the standard preset value is corrected according to the number of the course examples and the preset proportional factor, so as to obtain the corresponding preset value according to the corrected standard preset value; In a case where the number of the course examples is not less than the preset number, the standard preset value is used as the preset value of the resource reliability.
4. The teaching management method based on artificial intelligence according to claim 1, characterized in that: Collect student feedback for each teaching resource group, including: For each teaching resource group, video data generated during the teaching process is collected, and the video data is recognized by a pre-trained support vector machine to recognize the teaching actions generated by the teacher during the teaching process, and to recognize the dynamic teaching resources corresponding to the teaching actions; For different dynamic teaching resources, respectively collect video data within the corresponding time range as classroom performance data corresponding to the dynamic teaching resources; Identify the classroom performance data to identify feedback actions generated by the target student within a corresponding time range, and identify multiple action categories included in the feedback actions, where the action categories include: standing, sitting, leaning forward, raising the head, lowering the head, reading, writing, and communicating; Each of the dynamic teaching resources is combined with an action category within a corresponding time range to obtain feedback information corresponding to the teaching resource group.
5. The teaching management method based on artificial intelligence according to claim 1, characterized in that: Determining the student's suitability for the teaching resource group according to the feedback information specifically includes: Based on a preset mapping relationship table, as well as the dynamic teaching resources and action categories appearing in the feedback information, determining whether the feedback information is a positive performance behavior; If yes, determine the number of occurrences and duration of the positive performance behavior corresponding to each action category, and determine the performance score of the student in each action category according to the number of occurrences and / or the duration, and normalize the performance score to obtain the student's fitness for the teaching resource group; Among them, the performance scores corresponding to teaching, demonstration, and inspection are positively correlated with the duration, and the performance score corresponding to interaction is positively correlated with the number of occurrences.
6. The teaching management method based on artificial intelligence according to claim 5 is characterized in that: After determining whether the feedback information is a positive performance behavior, the method further includes: If it is determined that the active performance behavior corresponds to the dynamic teaching resource as interaction, then determining the expression type of the student's corresponding facial expression, the expression type including positive expression and negative expression, the positive expression at least including: excitement, calmness, the negative expression at least including: tension, fear; If the expression type of the student belongs to the positive expression, confirming that the positive expression behavior corresponding to the dynamic teaching resource is interactive is valid; Otherwise, the positive performance behavior is confirmed to be invalid.
7. The teaching management method based on artificial intelligence according to claim 1, characterized in that: A target teaching resource group with the largest degree of adaptation is selected from the teaching resource group as a course example of the target course, specifically including: For the target teaching resource group, determine the teaching form corresponding to the target teaching resource group, wherein the teaching form includes: teacher lecture form, interactive question form, experiment form, and self-study form, which respectively correspond to the lecture, interaction, demonstration, and inspection in the dynamic teaching resources; According to the performance score corresponding to the action category, the appearance time of the corresponding teaching form in the course example is determined.
8. The teaching management method based on artificial intelligence according to claim 2, characterized in that: According to the dependency type, the reliability between the static teaching resource and the target course is evaluated to obtain the resource reliability corresponding to the static teaching resource, specifically including: For different dependency types, corresponding dependency type weights are assigned to them; wherein the dependency type weights corresponding to the strong active dependency, the weak active dependency and the passive dependency decrease in sequence; According to the data channel corresponding to the static teaching resource, the corresponding resource base reliability is determined, and according to the dependency type weight and the resource base reliability, the resource reliability corresponding to the static teaching resource is obtained; wherein the data channel includes at least one or more of the following: teaching materials, supplementary teaching materials, educational resource websites, official government websites, and internal resource libraries of schools; When there are multiple static teaching resources, the average of the resource reliabilities corresponding to all static teaching resources is calculated, and the average is used as the resource reliability finally calculated.
9. A teaching management device based on artificial intelligence, characterized in that: include: at least one processor; as well as, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the artificial intelligence-based teaching management method as described in any one of claims 1-8.
10. A non-volatile computer storage medium storing computer executable instructions, characterized in that: The computer executable instructions are configured as: the teaching management method based on artificial intelligence as described in any one of claims 1 to 8.