Content interaction method for AI online education

Abstract

The invention discloses a content interaction method for AI online education, which is applied to a content interaction system for online AI education, and the content interaction system for online AI education comprises a teacher end, a server end and a student end. According to the AI online education content interaction method, through interaction among the teacher terminal, the server terminal and the student terminal, students can play back at any time point through the student terminal, timely and repeatedly listen to important knowledge points, and can also join in a live course again without missing follow-up real-time explanation content; meanwhile, the playback efficiency is improved through the judgment of the wrong playback operation at the student end. According to the playback dynamic state of the student end, the teacher end generates a message prompt of a playback hotspot interval to assist a teacher to adjust an explanation strategy in a targeted manner, so that the content is fully understood and mastered, and the teaching effect is effectively improved; and meanwhile, the repeated content explanation is identified and the classroom rest time period is reminded, so that the time is effectively saved when the student side plays back.

Description

Technical Field

[0001] The present invention belongs to the technical field of online education, and particularly relates to a content interaction method for AI online education. Background Art

[0002] Online learning refers to learning carried out in an electronic environment composed of communication technology, microcomputer technology, computer technology, artificial intelligence, network technology, multimedia technology, etc., which is technology-based learning. It is a way of network teaching and learning through the computer Internet or through the mobile phone wireless network in a network virtual classroom. Interactive learning refers to a new learning method that uses multimedia computer technology and network technology to realize the communication and interaction between teachers and learners and among learner groups. This learning method is more likely to achieve one-on-one communication between learning and teaching, fully respects the individuality of learning, stimulates the motivation of learning, and is not restricted by time, place, and space, and can achieve the same interaction as in reality.

[0003] In the existing online learning, the way that requires students to give active feedback may have the situation that students do not give active feedback due to shyness or other reasons, so that teachers cannot obtain accurate information and adjust the teaching speed or teaching content; at the same time, the relatively independent methods of live broadcast and playback in the existing technology have two problems for students: one is that during the live broadcast, if some knowledge points are not understood, it will affect the subsequent listening of this class; the other is that after the live broadcast, when playing back again, it takes more time for students to learn. Summary of the Invention

[0004] The purpose of the present invention is to propose a content interaction method for AI online education to solve the problems raised in the background art.

[0005] To achieve the above purpose, the technical solution adopted by the present invention is as follows: A content interaction method for AI online education proposed by the present invention is applied to a content interaction system for online AI education. The content interaction system for online AI education includes a teacher terminal, a server terminal, and a student terminal, wherein: The teacher terminal sends the live teaching video stream to the server terminal in a multicast manner, and each student terminal joins the multicast group to play the live teaching video stream. For each student terminal: When the student terminal triggers the playback of the live teaching video stream, it sends the first behavior information to the server terminal, and according to whether the corresponding playback video stream is cached locally at the student terminal, directly plays back the corresponding playback video stream or the server terminal sends the corresponding playback video stream to the student terminal for playback; When the student terminal finishes the playback, it rejoins the multicast group to play the live teaching video stream and sends the second behavior information to the server terminal; The server generates the playback information corresponding to each student terminal in real time according to the first-line information or the second-line information received from each student terminal, and sends each playback information to the teacher terminal in real time; The teacher terminal analyzes the latest playback information received from each student terminal, dynamically counts the current latest number of students and the proportion within each preset playback time period, and takes the playback time period with a proportion exceeding the first threshold as the playback hot interval, and generates a message reminder for the playback hot interval; The server also automatically identifies whether the current content segment is a repeated content explanation. If it is a repeated content explanation, a reminder of a repeated content segment will be given when each student terminal replays to the current content, and for the classroom break time period set by the teacher terminal, a reminder of the classroom break time period will be given when each student terminal replays to the corresponding time period.

[0006] Preferably, the content interaction method of the AI online education further includes: when the student terminal is in the process of video stream playback, the student terminal adjusts the progress bar and / or the playback speed of the playback video stream. If the corresponding playback video stream is cached locally by the student terminal, the student terminal directly uses the locally cached playback video stream to perform the playback of the corresponding playback video stream and / or the corresponding playback speed. If the corresponding playback video stream is not cached locally by the student terminal, the third-line information is sent to the server. After receiving the third-line information, the server sends the corresponding playback video stream and / or the playback video stream with the corresponding playback speed to the student terminal according to the progress bar adjustment and / or the playback speed adjustment.

[0007] Preferably, the first-line information includes the teaching course ID, the trigger playback behavior, the student ID, and the playback start time point, and the trigger playback behavior includes the trigger behavior when the corresponding playback video stream is not cached locally by the student terminal and the trigger behavior when the corresponding playback video stream is cached locally by the student terminal; The second-line information includes the teaching course ID, the end of playback, and the student ID; The third-line information includes the teaching course ID, the adjusted playback, the student ID, the adjusted playback start time point, and the adjusted playback speed.

[0008] Preferably, when the student terminal triggers the playback of the live teaching video stream, the first-line information is sent to the server, and according to whether the corresponding playback video stream is cached locally by the student terminal, the direct playback of the corresponding playback video stream or the server sends the corresponding playback video stream to the student terminal for playback, including: The student terminal drags the progress bar of the live teaching video stream to trigger playback. If the corresponding playback video stream is cached locally by the student terminal, first judge whether it is a misplayback operation. If it is a misplayback operation, the first-line information is not sent to the server. Otherwise, the first-line information is sent to the server, and the student terminal directly uses the locally cached playback video stream to replay the corresponding content; If there is no cached corresponding playback video stream on the student side, a pop-up window will appear on the student side to prompt whether to confirm the video stream playback. After confirmation, the student side will send the first-line information to the server. After the server recognizes that the triggering playback in the first-line information is due to no cached video stream on the student side, it will send the video stream of the corresponding playback content to the student side.

[0009] Preferably, the student side determines whether it is a misplayback operation, including: When the camera of the student side is turned on, through image recognition, check whether the student is watching. If the student is not watching, it is determined as a misplayback operation; otherwise, it is determined as not a misplayback operation. When the camera of the student side is not turned on, determine whether the time maintained after the student drags the progress bar of the live teaching video stream is greater than the second threshold. If it is greater than the second threshold, it is determined as not a misplayback operation; otherwise, it is determined as a misplayback operation.

[0010] Preferably, for the first-line information, the corresponding playback information includes: teaching course ID, triggering playback, student ID, and playback start time point. For the second-line information, the corresponding playback information includes: teaching course ID, ending playback, and student ID.

[0011] Preferably, the teacher side analyzes the latest playback information received from each student side, dynamically counts the current latest student quantity and proportion within each preset playback time period, and takes the playback time period with a proportion exceeding the first threshold as the playback hot spot interval, and generates a message reminder for the playback hot spot interval, including: The teacher side pre-divides the live teaching video stream into multiple time periods, and each time period is used as a preset playback time period. The teacher side, according to the playback information of each student side, counts the quantity and proportion of the playback start time points of the currently playing student sides within each preset playback time period, and this proportion is the proportion of the quantity of playback start time points in each playback time period to the total number of students. Among them, the quantity of playback start time points is used as the student quantity within the playback time period. The teacher side takes the playback time period with a proportion exceeding the first threshold as the playback hot spot interval and sends a message reminder of the playback hot spot interval to the teacher. The teacher re-explains the corresponding content according to the message reminder of the playback hot spot interval.

[0012] Preferably, the server also automatically identifies whether the current content segment is a repeated content explanation. If it is a repeated content explanation, a reminder of the repeated content segment will be given when each student side plays back to the current content, including: The server divides the live teaching video stream into video segments of a preset length, converts the explanatory speech of each video segment into text, then extracts the keywords and themes of each text, calculates the similarity between different texts through the keywords and themes of each text, and performs clustering analysis according to the similarity. The video segments corresponding to each category divided by the clustering analysis are used as duplicate content segments; The server sets a reminder for the duplicate content segments in the playback video stream, so that each student terminal will be reminded that it is a duplicate content segment when playing back to the duplicate content; At the same time, when the student terminal plays back to the duplicate content, the student terminal pops up a reminder asking whether to play at a preset multiple speed.

[0013] Preferably, for the classroom break time period set by the teacher terminal, when each student terminal plays back to the corresponding time period, it is reminded as the classroom break time period, including: The teacher terminal is provided with a classroom break button and a break end button. When each button operation is performed, a classroom break notification message is sent to the server. The classroom break notification message includes the teaching course ID, the button operation type, and the time point of the button operation. The button operation type includes classroom break and break end. When the button operation type is classroom break, the time point of the button operation is the start time point of the classroom break. When the button operation type is break end, the time point of the button operation is the end time point of the classroom break; The server identifies the start time point and the end time point of the classroom break according to each classroom break notification message, and sends a break reminder message to the student terminal. The break reminder message includes the teaching course ID, the classroom break start point, and the classroom break end time point; The student terminal automatically skips the classroom break time period according to the classroom break start point and the classroom break end time point in the break reminder message.

[0014] Preferably, when the teacher terminal ends the live teaching, the server stops sending the live teaching video stream, and the student terminal synchronously stops playing the live teaching video stream. And when the student terminal plays back to the time point when the live teaching ends, it is reminded that the playback is completed.

[0015] Compared with the prior art, the beneficial effects of the present invention are: Through the interaction between the teacher terminal, the server and the student terminal, the content interaction method of this AI online education enables students to play back at any time point through the student terminal, repeat and listen to important knowledge points in time, and can also rejoin the live course without missing subsequent real-time explanation content; at the same time, through the judgment of mis-playback operations on the student terminal, the playback efficiency is improved; According to the playback dynamics on the student side, the content interaction method of this AI online education generates message reminders for the playback hot spot intervals on the teacher side, assisting the teacher to adjust the teaching strategy targeted, ensuring that the content is fully understood and mastered, and effectively improving the teaching effect; at the same time, the recognition of repeated content explanations and the reminder of the classroom break time period enable the student side to effectively save time during playback. Brief Description of the Drawings

[0016] Figure 1 It is a flowchart of the content interaction method of the AI online education of the present invention; Figure 2 It is a flowchart of triggering playback of the present invention. Detailed Embodiments

[0017] In order to make the purpose, technical solutions and advantages of this application clearer, the following further details this application in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not used to limit this application.

[0018] In one embodiment, as Figure 1 - Figure 2 shown, a content interaction method for AI online education is provided, which is applied to the content interaction system of online AI education. The content interaction system of online AI education includes a teacher side, a server side and a student side, where: Step 1: The teacher side sends the live teaching video stream to the server side in a multicast manner, and each student side joins the multicast group to play the live teaching video stream; It should be noted that the teacher starts live teaching on a smart terminal (such as a computer, PAD, mobile phone, etc.) through education software (such as an APP, a small program, etc.). The education software is called the teacher side. The teacher side sends the live teaching video stream to the central server (i.e., the server side) in a multicast manner. Multiple students join the multicast group through the education software client of the smart terminal (called the student side) to watch the live teaching video.

[0019] For each student side: Step 2: When the student side triggers the playback of the live teaching video stream, send the first behavior information to the server side, and according to whether the corresponding playback video stream is cached locally on the student side, perform direct playback of the corresponding playback video stream or the server side sends the corresponding playback video stream to the student side for playback, including: Step 2.1: The student side drags the progress bar of the live teaching video stream to trigger playback (the student drags the progress bar on the student side to trigger playback); Step 2.2: If the corresponding playback video stream is cached locally on the student side, first determine whether it is an incorrect playback operation (the student side may readjust the playback point multiple times, so in order to avoid misjudgment of the playback behavior, two determination methods are adopted); Step 2.2.1. The student side determines whether it is a misplayed operation, including: When the camera of the student side is turned on, through image recognition (where image recognition technology is a mature technology and will not be elaborated in this solution), check whether the student is watching (if a face is recognized and the eyes are not closed, it is considered watching). If the student is not watching (if no face is recognized or the eyes are closed, it is considered not watching), then it is determined as a misplayed operation; otherwise, it is determined as not a misplayed operation. When the camera of the student side is not turned on, determine whether the time maintained after the student drags the progress bar of the live teaching video stream is greater than the second threshold. If it is greater than the second threshold, it is determined as not a misplayed operation; otherwise, it is determined as a misplayed operation. Step 2.2.2. If it is a misplayed operation, the student side does not send the first behavior information to the server; otherwise, it sends the first behavior information to the server, and the student side directly plays back the corresponding content using the locally cached playback video stream. The first behavior information includes: teaching course ID, playback trigger behavior, student ID, and playback start time point. In this step, the playback trigger behavior is the trigger behavior when the student side locally caches the corresponding playback video stream. When the server recognizes that the playback trigger behavior is the trigger behavior when the student side locally caches the corresponding playback video stream, the server does not need to perform any operations.

[0020] Step 2.3. If the student side does not locally cache the corresponding playback video stream, the student side pops up a prompt box (i.e., a prompt box pops up after triggering the playback) to remind whether to confirm the video stream playback. When confirmed, the student side sends the first behavior information to the server (where the playback trigger behavior in the first behavior information in this step is the trigger behavior when the student side does not locally cache the video stream; when not performing the video stream playback, the first behavior information is not sent to the server). When the server recognizes that the playback trigger behavior in the first behavior information is the trigger behavior after the student side does not locally cache the video stream, it sends the video stream of the corresponding playback content to the student side with the corresponding student ID.

[0021] Step 3. When the student side is performing video stream playback, the student side adjusts the progress bar and / or the playback speed of the playback video stream. If the student side locally caches the corresponding playback video stream, the student side directly plays back the corresponding playback video stream and / or the corresponding playback speed using the locally cached playback video stream. If the student side does not locally cache the corresponding playback video stream, it sends the third behavior information to the server. After receiving the third behavior information, the server sends the corresponding playback video stream and / or the playback video stream with the corresponding playback speed to the student side according to the progress bar adjustment and / or the playback speed adjustment. Among them, the information in the third line includes the teaching course ID, adjusted playback, student ID, adjusted playback start time point (i.e., adjusting the progress bar to a certain time point), and adjusted playback speed (when the student side does not adjust the progress bar, the adjusted playback start time point is not included in the information in the third line; when the student side does not adjust the playback speed, the adjusted playback speed is not included in the information in the third line).

[0022] Step 4: After the student side finishes playing back (students watching the playback explanation content, whether it is local cached playback or server video stream playback, can click the exit playback button on the student side and then re-join the multicast group of the video explanation live broadcast to watch the live broadcast), re-join the multicast group to play the live teaching video stream and send the information in the second line to the server; after finishing the playback, the student side can still restart the playback; Among them, the information in the second line includes the teaching course ID, end of playback, and student ID.

[0023] Step 5: The server generates the corresponding playback information for each student side in real time according to the information in the first line or the second line received from each student side, and sends each playback information to the teacher side in real time (each time the server generates a playback information, it is sent to the teacher side in real time); Among them, for the information in the first line, the corresponding playback information includes: teaching course ID, triggered playback, student ID, and playback start time point; For the information in the second line, the corresponding playback information includes: teaching course ID, end of playback, and student ID.

[0024] Step 6: The teacher side analyzes the latest playback information received from each student side (that is, if the teacher side receives both the playback information corresponding to the information in the first line of a certain student side and the playback information corresponding to the information in the second line, only the latest received behavior information of this student side needs to be analyzed), dynamically counts the current latest number of students and the proportion within each preset playback time period, and uses the playback time period with a proportion exceeding the first threshold as the playback hot spot interval, and generates a message reminder for the playback hot spot interval, including: The teacher side pre-divides the live teaching video stream into multiple time periods (such as 1 - 10 minutes as a time period, 11 - 20 minutes as a time period, and so on), and each time period is used as a preset playback time period; Based on the latest playback information of each student terminal, the teacher terminal counts the number and proportion of the playback start time points of the currently playing student terminals within each preset playback time period (when the playback behavior type in the latest playback information received by the teacher terminal is end of playback, it is considered that the corresponding student terminal is not currently playing and does not need to be counted), and this proportion is the proportion of the number of playback start time points within each playback time period to the total number of students. The number of playback start time points is used as the number of students within the playback time period; The teacher terminal takes the playback time period with a proportion exceeding the first threshold as the playback hot interval and sends a message reminder of the playback hot interval to the teacher; Based on the message reminder of the playback hot interval, the teacher re-explains the corresponding content (for example, repeats the review of the already explained content, slows down the explanation speed, etc.).

[0025] Step 7: The server also automatically identifies whether the current content segment is a repeated content explanation. If it is a repeated content explanation, a reminder of the repeated content segment is given when each student terminal plays back to this content, including: The server divides the live teaching video stream into video segments of a preset length (such as every ten minutes as a video segment), and converts the explanatory speech of each video segment into text (using audio signal processing technology to extract the audio from each video segment through decoding, and then using a deep learning model (such as a deep learning-based voice separation model (Conv-TasNet model)) to separate the speech component in the audio signal from other sounds (such as ambient noise, etc.), and then through a deep learning model (such as a convolutional neural network CNN, long short-term memory network LSTM) for speech recognition and conversion into text, and this process belongs to the prior art and will not be described in detail). Then, keywords and themes of each text are extracted (natural language processing technology can be used), the similarity between different texts is calculated through the keywords and themes of each text (such as cosine similarity), and clustering analysis is performed according to the similarity (this process belongs to the well-known prior art and will not be described in detail). The corresponding video segments in each category divided by the clustering analysis are used as repeated content segments (for example, there are a total of ten video segments, and after clustering analysis, the first, sixth, and ninth video segments are in the same category, then the first, sixth, and ninth video segments are used as repeated content segments); The server sets a reminder for the repeated content segment in the playback video stream, so that a reminder of the repeated content segment is given when each student terminal plays back to the repeated content; At the same time, when the student terminal plays back to the repeated content, a pop-up window on the student terminal reminds whether to play back at a preset multiple speed (at this time, the student can play back at a preset multiple speed, such as 1.2 times, or adjust the playback speed by himself).

[0026] Step 8: The server reminds each student terminal of the classroom break time period set by the teacher as the classroom break time period when the student terminal plays back the corresponding time period, including: The teacher's end is equipped with a button for class break and a button for ending the break. When each button operation is performed, the class break notification information is sent to the server, where the class break notification information includes the teaching course ID, the button operation type and the time point of the button operation. The button operation type includes class break and end of break. When the button operation type is class break, the time point of the button operation is the start time point of the class break. When the button operation type is end of break, the time point of the button operation is the end time point of the class break. The server identifies the start time and end time of the class break according to the class break notification information, and sends a break reminder message to the student end, where the break reminder message includes the teaching course ID, the class break start time and the class break end time; The student terminal automatically skips the class break period according to the class break start and end time points in the break reminder information.

[0027] Step 9: When the teacher ends the live teaching, the server stops sending the live teaching video stream, and the student ends simultaneously stop playing the live teaching video stream. When the student ends playback to the time point when the live teaching ends, the student ends are reminded that the playback is complete.

[0028] In another embodiment, the present application also includes a content interaction device for online AI education, including a processor and a memory storing a plurality of computer instructions, and when the computer instructions are executed by the processor, the steps of step 1 to step 9 of the method are implemented.

[0029] This AI online education content interaction method enables students to replay at any time through the student end through the interaction between the teacher end, the server end and the student end, and to repeat the important knowledge points in time. Students can also rejoin the live course and will not miss the subsequent real-time explanation content. At the same time, the efficiency of playback is improved by judging the wrong playback operation on the student end. This AI online education content interaction method generates message reminders for playback hot spots based on the playback dynamics of the student side, assisting teachers to adjust the explanation strategy in a targeted manner to ensure that the content is fully understood and mastered, effectively improving the teaching effect; at the same time, the identification of repeated content explanations and reminders of class break time periods enable students to effectively save time when replaying.

[0030] It should be understood that although Figure 1 and Figure 2The steps in the flowchart are shown in sequence according to the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, Figure 1 and Figure 2 At least some of the steps in may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed alternately or in turn with at least some of the sub-steps or stages of other steps or other steps.

[0031] The above-described embodiments merely represent several implementation manners of the present application. The description is relatively specific and detailed, but it should not be construed as a limitation on the scope of the invention patent. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the appended claims.

Claims

1. A content interaction method for AI online education, characterized by: The content interaction method of AI online education is applied to the content interaction system of online AI education, and the content interaction system of online AI education includes a teacher end, a server end and a student end, wherein: The teacher sends the live teaching video stream through the server in multicast mode. Each student joins the multicast group to play the live teaching video stream. For each student: When the student end triggers the playback of the live teaching video stream, the first behavior information is sent to the server end, and according to whether the student end has the corresponding playback video stream cached locally, the corresponding playback video stream is directly played back or the server end sends the corresponding playback video stream to the student end for playback; When the student end finishes playback, it rejoins the multicast group to play the live teaching video stream and sends the second behavior information to the server end; The server generates the playback information of the corresponding student terminal in real time according to the first behavior information or the second behavior information received from each student terminal, and sends the playback information to the teacher terminal in real time; The teacher terminal analyzes the latest playback information received from each student terminal, dynamically counts the current latest number and proportion of students in each pre-set playback time period, and takes the playback time period whose proportion exceeds the first threshold as the playback hotspot interval, and generates a message reminder for the playback hotspot interval; The server also automatically identifies the current content segment to see if it is a repeated content explanation. If it is a repeated content explanation, the server will remind each student that it is a repeated content segment when they play back the current content. For the class break time period set by the teacher, the server will remind each student that it is a class break time period when they play back the corresponding time period.

2. The content interaction method for AI online education according to claim 1, characterized in that: The content interaction method of AI online education also includes: when the student end is playing back the video stream, the student end adjusts the progress bar and / or the playback speed of the playback video stream. If the student end has a corresponding playback video stream cached locally, the student end uses the locally cached playback video stream to directly play back the corresponding playback video stream and / or the corresponding playback speed. If the student end has no corresponding playback video stream cached locally, a third behavior information is sent to the server end. After receiving the third behavior information, the server end sends the corresponding playback video stream and / or the playback video stream at the corresponding playback speed to the student end according to the progress bar adjustment and / or the playback speed adjustment.

3. The content interaction method of AI online education according to claim 2, characterized in that: The first behavior information includes a teaching course ID, a triggering playback behavior, a student ID, and a playback start time point, and the triggering playback behavior includes a triggering behavior of a playback video stream that is not locally cached on the student end and a triggering behavior of a playback video stream that is locally cached on the student end; The second behavior information includes the teaching course ID, end playback and student ID; The third behavior information includes the teaching course ID, adjusted playback, student ID, adjusted playback start time point and adjusted play speed.

4. The content interaction method for AI online education according to claim 3, characterized in that: When the student terminal triggers the playback of the live teaching video stream, the first behavior information is sent to the server terminal, and according to whether the student terminal has the corresponding playback video stream cached locally, the corresponding playback video stream is directly played back or the server terminal sends the corresponding playback video stream to the student terminal for playback, including: The student end drags the progress bar of the live teaching video stream to trigger playback. If the student end has the corresponding playback video stream cached locally, it first determines whether it is an erroneous playback operation. If it is an erroneous playback operation, the first behavior information is not sent to the server end. Otherwise, the first behavior information is sent to the server end, and the student end directly plays back the corresponding content using the locally cached playback video stream; If the student end does not have the corresponding playback video stream cached locally, a pop-up box will pop up on the student end to remind whether to confirm the video stream playback. If confirmed, the student end sends the first behavior information to the server end. After the server end recognizes that the situation that triggers playback in the first behavior information is that the student end does not have the corresponding cached video stream, it will send the video stream of the corresponding playback content to the student end.

5. The content interaction method for AI online education according to claim 4, characterized in that: The student terminal determines whether the playback operation is an erroneous operation, including: When the student turns on the camera, the system uses image recognition to check whether the student is watching. If the student is not watching, it is judged as an erroneous playback operation. Otherwise, it is judged as not an erroneous playback operation. When the camera is not turned on on the student side, it is determined whether the time the student keeps dragging the progress bar of the live teaching video stream is greater than the second threshold. If it is greater than the second threshold, it is determined that it is not an erroneous playback operation, otherwise it is determined that it is an erroneous playback operation.

6. The content interaction method for AI online education according to claim 1, characterized in that: For the first line of information, the corresponding playback information includes: teaching course ID, trigger playback, student ID and playback start time point; For the second behavior information, the corresponding playback information includes: teaching course ID, end playback and student ID.

7. The content interaction method of AI online education according to claim 6, characterized in that: The teacher terminal analyzes the latest playback information received from each student terminal, dynamically counts the current latest number and proportion of students in each preset playback time period, and uses the playback time period whose proportion exceeds the first threshold as the playback hotspot interval, and generates a message reminder for the playback hotspot interval, including: The teacher divides the live teaching video stream into multiple time periods in advance, and each time period is used as a pre-set playback time period; The teacher end counts the number and proportion of playback start time points of the student end that are currently playing back in each pre-set playback time period based on the playback information of each student end, and the proportion is the ratio of the number of playback start time points in each playback time period to the total number of students, where the number of playback start time points is taken as the number of students in the playback time period; The teacher terminal uses the playback time period that exceeds the first threshold as the playback hotspot interval, and sends a message reminder of the playback hotspot interval to the teacher; The teacher will re-explain the corresponding content based on the message reminders in the playback hot spots.

8. The content interaction method for AI online education according to claim 1, characterized in that: The server also automatically identifies the current content segment to determine whether it is a repeated content explanation. If it is a repeated content explanation, each student terminal will be reminded that it is a repeated content segment when playing back the current content, including: The server divides the live teaching video stream into video segments of preset length, converts the explanation voice of each video segment into text, and then extracts the keywords and themes of each text, calculates the similarity between different texts through the keywords and themes of each text, and performs cluster analysis according to the similarity, and uses the video segments corresponding to each category divided by the cluster analysis as repeated content segments; The server sets a reminder for repeated content segments in the playback video stream, so that each student terminal is reminded of the repeated content segments when playing back the repeated content; At the same time, when the student end is playing back repeated content, a pop-up window will appear on the student end to remind whether to play at the preset speed.

9. The content interaction method for AI online education according to claim 1, characterized in that: The server end sets the classroom break time period set by the teacher end, and reminds each student end as the classroom break time period when playing back the corresponding time period, including: The teacher end is provided with a button for class break and a button for ending the break. When each button operation is performed, a class break notification message is sent to the server end, wherein the class break notification message includes the teaching course ID, the button operation type and the time point of the button operation, wherein the button operation type includes class break and ending the break. When the button operation type is class break, the time point of the button operation is the start time point of the class break, and when the button operation type is ending the break, the time point of the button operation is the end time point of the class break; The server identifies the start time and end time of the class break according to the class break notification information, and sends a break reminder message to the student end, where the break reminder message includes the teaching course ID, the class break start time and the class break end time; The student terminal automatically skips the class break period according to the class break start and end time points in the break reminder information.

10. The content interaction method for AI online education according to claim 1, characterized in that: When the teacher ends the live teaching, the server stops sending the live teaching video stream, and the student ends synchronously stop playing the live teaching video stream. When the student ends playback to the time point when the live teaching ends, the student ends are reminded that the playback is complete.

Images