English teaching method and system based on artificial intelligence

Abstract

The invention provides an English teaching method and system based on artificial intelligence, and relates to the technical field of artificial intelligence, and the method comprises the steps: obtaining the English learning condition of a user; based on artificial intelligence, according to an English learning condition, matching an optimal learning material for the user; and based on the virtual teacher, performing English teaching on the user according to the optimal learning material. According to the invention, the optimal learning material is matched for the user according to the English learning condition of the user based on artificial intelligence, and English teaching is carried out on the user according to the optimal learning material based on the virtual teacher. Personalized instant adjustment of teaching content based on the English language basis, the learning progress, the learning ability, the learning preference and the like of the learner is realized, the learning experience of the learner is prompted, and the learning effect is improved.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to an English teaching method and system based on artificial intelligence. Background Art

[0002] At present, the traditional English teaching model mostly adopts unified teaching for multiple learners in the classroom, usually relying on teachers to teach at a fixed time and place, and all learners learn according to the same progress and content.

[0003] However, there are significant differences among learners in terms of English language foundation, learning progress, learning ability and learning preferences. The traditional English teaching model cannot adjust the teaching content for learners based on this, resulting in poor learning experience for learners and may also affect their learning results. For example, some learners feel unable to keep up with the progress in class, while others may feel bored because the course content is too simple. For another example, for learners with poor language foundation, the rapid advancement and unified content in the traditional teaching model may make them feel frustrated and even averse to learning. For learners with strong language skills, the slow teaching progress and unified teaching content may not satisfy their desire for knowledge and affect their learning motivation.

[0004] Therefore, a solution is urgently needed. Summary of the invention

[0005] One of the purposes of the present invention is to provide an English teaching method based on artificial intelligence, which matches the best learning materials for users according to the users' English learning status based on artificial intelligence, and teaches English to users based on the best learning materials based on a virtual teacher. When teaching English to learners, personalized and instant adjustment of teaching content is achieved based on the learners' English language foundation, learning progress, learning ability and learning preferences, which prompts the learners' learning experience and improves the learning effect.

[0006] An embodiment of the present invention provides an English teaching method based on artificial intelligence, comprising:

[0007] Get the user's English learning status;

[0008] Based on artificial intelligence, the best learning materials are matched to users according to their English learning status;

[0009] Based on virtual teachers, users are taught English according to the best learning materials.

[0010] Optionally, the artificial intelligence-based method matches the best learning materials for the user according to the English learning situation, including:

[0011] Retrieve the pre-trained artificial intelligence model;

[0012] Based on the artificial intelligence model, match the best learning materials for the user according to the English learning situation;

[0013] Among them, the pre-training steps of the artificial intelligence model are as follows:

[0014] Collect training samples; among them, the training samples include: a large number of pre-labeled preparatory learning materials adapted to the English learning situation;

[0015] Based on the machine learning algorithm, train the artificial intelligence model according to the training samples.

[0016] Optionally, the English teaching of the user based on the virtual teacher according to the best learning materials includes:

[0017] Guide the user into the virtual classroom;

[0018] Based on the best learning materials, control the virtual teacher to start English teaching for the user in the virtual classroom; among them, the virtual teacher self-decides the teaching link sequence and conducts corresponding teaching according to the teaching link sequence;

[0019] When the virtual teacher conducts English teaching to the i-th teaching link in the teaching link sequence, identify whether the user is distracted during the class; where i≥D, and the importance of the teaching of the D-th teaching link in the teaching link sequence exceeds the importance threshold for the first time;

[0020] When the recognition is yes, identify whether the user generates effective distraction content;

[0021] When the recognition is yes, based on the effective distraction content, optimize and replace the i-th to the i+j-th teaching links in the teaching link sequence; among them, the virtual teacher continues to conduct teaching based on the optimized and replaced teaching link sequence of the i-th to the i+j-th teaching links; j is the proportional quantization value corresponding to the number of times the user has generated different effective distraction contents in the quantization value library;

[0022] Otherwise, control the virtual teacher to give distraction reminders and distraction remedy assistance to the user in sequence.

[0023] Optionally, the identification of whether the user is distracted during the class includes:

[0024] Continuously obtain multiple listening behaviors of the user;

[0025] Based on the concentration analysis library, determine the concentration of each listening behavior;

[0026] Based on the concentration of each listening behavior and the generation time of each listening behavior, draw a concentration curve; where the horizontal axis of the concentration curve is time and the vertical axis is the concentration value;

[0027] Identify whether a target valley value lower than the valley value threshold first appears in the concentration curve;

[0028] When the identification result is yes, determine that the user is distracted during the class;

[0029] Otherwise, determine that the user is not distracted during the class.

[0030] Optionally, the identifying whether the user generates effective distraction content includes:

[0031] Take the occurrence time of the target valley value as the first boundary time;

[0032] Identify whether more than the number threshold of target peak values higher than the peak value threshold continuously appear immediately after the first boundary time in the concentration curve, and the last target peak value that appears is higher than all other target peak values;

[0033] When the identification result is yes, take the occurrence time of the last target peak value that appears as the second boundary time; otherwise, determine that the user does not generate effective distraction content;

[0034] Identify whether the user accesses an autonomous operation scenario that has a standard association relationship with the i-th teaching link between the first boundary time and the second boundary time;

[0035] When the identification result is yes, determine that the user generates effective distraction content, obtain the scenario operation dynamics of the autonomous operation scenario at the second boundary time, and use it as the effective distraction content; otherwise, determine that the user does not generate effective distraction content.

[0036] Optionally, the optimizing and replacing the i-th to the (i + j)-th teaching links in the teaching link sequence based on the effective distraction content includes:

[0037] Extract features from the effective distraction content, the i-th to the (i + j)-th teaching links, and the link association relationships between the i-th to the (i + j)-th teaching links pairwise to obtain a multi-dimensional feature set;

[0038] Construct an optimized replacement situation vector based on the multi-dimensional feature set;

[0039] Determine the optimized replacement strategy corresponding to the optimized replacement situation vector from the optimized replacement strategy library;

[0040] Optimize and replace the i-th to the (i + j)-th teaching links based on the optimized replacement strategy.

[0041] Optionally, the controlling the virtual teacher to give distraction reminders and distraction remedy assistance to the user successively includes:

[0042] Control the virtual teacher to pause the teaching of the i-th teaching link;

[0043] Control the virtual teacher to interact with the user for distraction reminder;

[0044] Generate learning review materials based on the taught content of the virtual teacher in the i-th teaching session before and after the first boundary moment within a preset time;

[0045] Based on the learning review materials, control the virtual teacher to conduct review learning for the user.

[0046] An English teaching system based on artificial intelligence provided by an embodiment of the present invention includes:

[0047] An English learning status acquisition module for acquiring the English learning status of the user;

[0048] An optimal learning material matching module for matching the optimal learning materials for the user based on artificial intelligence according to the English learning status;

[0049] An English teaching module for teaching the user English based on the virtual teacher according to the optimal learning materials.

[0050] Optionally, the optimal learning material matching module matches the optimal learning materials for the user based on artificial intelligence according to the English learning status, including:

[0051] Retrieve a pre-trained artificial intelligence model;

[0052] Based on the artificial intelligence model, match the optimal learning materials for the user according to the English learning status;

[0053] Among them, the pre-training steps of the artificial intelligence model are as follows:

[0054] Collect training samples; among them, the training samples include: a large number of pre-labeled preparatory learning materials adapted to the English learning status;

[0055] Based on the machine learning algorithm, train the artificial intelligence model according to the training samples.

[0056] Optionally, the English teaching module teaches the user English based on the virtual teacher according to the optimal learning materials, including:

[0057] Guide the user into the virtual classroom;

[0058] Based on the optimal learning materials, control the virtual teacher to start teaching the user English in the virtual classroom; among them, the virtual teacher self-decides the teaching session sequence and conducts corresponding teaching according to the teaching session sequence;

[0059] When the virtual teacher conducts English teaching up to the i-th teaching link in the teaching link sequence, it is identified whether the user is distracted during the class; where i ≥ D, and the importance of the teaching of the D-th teaching link in the teaching link sequence first exceeds the importance threshold;

[0060] When it is identified as yes, it is identified whether the user generates effective distraction content;

[0061] When it is identified as yes, based on the effective distraction content, the i-th to the i + j-th teaching links in the teaching link sequence are optimized and replaced; where the virtual teacher continues to conduct teaching in relay based on the optimized and replaced teaching link sequence of the i-th to the i + j-th teaching links; j is the proportional quantization value corresponding to the number of times the user has generated different effective distraction contents in the quantization value library;

[0062] Otherwise, control the virtual teacher to give distraction reminders and distraction remedy assistance to the user successively.

[0063] Other features and advantages of the present invention will be described in the subsequent specification, and, in part, will be obvious from the specification, or will be understood by implementing the present invention. The objectives and other advantages of the present invention can be realized and obtained by the structures specifically pointed out in the written specification and the drawings.

[0064] The technical solution of the present invention will be further described in detail below through the drawings and embodiments. Description of the Drawings

[0065] The drawings are used to provide a further understanding of the present invention, and constitute a part of the specification. They are used together with the embodiments of the present invention to explain the present invention, and do not constitute a limitation to the present invention. In the drawings:

[0066] Figure 1 It is a schematic diagram of an English teaching method based on artificial intelligence in an embodiment of the present invention;

[0067] Figure 2 It is a schematic diagram of an English teaching system based on artificial intelligence in an embodiment of the present invention. Detailed Embodiments

[0068] The following describes the preferred embodiments of the present invention with reference to the drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0069] The embodiment of the present invention provides an English teaching method based on artificial intelligence, as Figure 1 shown, including:

[0070] S1. Obtain the English learning status of the user;

[0071] S2. Based on artificial intelligence, match the best learning materials for the user according to the English learning situation;

[0072] S3. Based on a virtual teacher, conduct English teaching for the user according to the best learning materials.

[0073] The user is a learner of English; correspondingly, the English learning situation at least includes: English language foundation, learning progress, learning ability, and learning preferences. The acquisition of the English learning situation can be obtained by conducting a questionnaire survey on the user, or can also be analyzed and extracted from the user's historical English learning records. Use the English learning situation as the basis for matching the best learning materials, and use artificial intelligence technology to match the best learning materials. The best learning materials are the materials most suitable for teaching English to the user, and they are adapted to the user's English language foundation, learning progress, learning ability, and learning preferences. The virtual teacher is a virtual teacher model created based on 3D technology, and it can conduct English teaching for the user according to the best learning materials.

[0074] The present invention matches the best learning materials for the user based on artificial intelligence according to the user's English learning situation, and conducts English teaching for the user based on the virtual teacher according to the best learning materials, so that when teaching English to learners, it realizes personalized and immediate adjustment of teaching content based on the learner's English language foundation, learning progress, learning ability, and learning preferences, improves the learner's learning experience, and further improves the learning effect.

[0075] In one embodiment, the step of matching the best learning materials for the user based on artificial intelligence according to the English learning situation includes:

[0076] Retrieve a pre-trained artificial intelligence model;

[0077] Based on the artificial intelligence model, match the best learning materials for the user according to the English learning situation;

[0078] Among them, the pre-training steps of the artificial intelligence model are as follows:

[0079] Collect training samples; among them, the training samples include: a large number of pre-labeled preparatory learning materials adapted to the English learning situation;

[0080] Based on a machine learning algorithm, train the artificial intelligence model according to the training samples.

[0081] When collecting training samples, a large number of preparatory learning materials are collected. The preparatory learning materials are a large number of materials for English teaching. Then, these preparatory learning materials are handed over to English teaching experts for annotation of adaptation to the English learning situation. The adaptation to the English learning situation is the English learning situation of the learners for whom the preparatory learning materials are suitable for English teaching. After annotation, these preparatory learning materials with the annotated adaptation to the English learning situation are used as training samples, and machine learning algorithms are used for model training to obtain an artificial intelligence model. The artificial intelligence model can then match the best learning materials for the user according to the user's English learning situation.

[0082] In one embodiment, the virtual teacher conducts English teaching for the user based on the best learning materials, including:

[0083] Guide the user into the virtual classroom;

[0084] Based on the best learning materials, control the virtual teacher to start English teaching for the user in the virtual classroom; wherein, the virtual teacher self-determines the teaching session sequence and conducts corresponding teaching according to the teaching session sequence;

[0085] When the virtual teacher conducts English teaching to the i-th teaching session in the teaching session sequence, identify whether the user is distracted during the class; wherein, i≥D, and the importance of the teaching in the D-th teaching session in the teaching session sequence first exceeds the importance threshold;

[0086] When it is identified as yes, identify whether the user generates effective distraction content;

[0087] When it is identified as yes, based on the effective distraction content, optimize and replace the i-th to the i + j-th teaching sessions in the teaching session sequence; wherein, the virtual teacher continues teaching based on the optimized and replaced teaching session sequence of the i-th to the i + j-th teaching sessions; j is the proportional quantization value corresponding to the number of times the user has generated different effective distraction contents in the quantization value library;

[0088] Otherwise, control the virtual teacher to give the user distraction reminders and distraction remedy assistance in sequence.

[0089] The virtual classroom is a virtual space for users to have teaching interactions with virtual teachers. When guiding a user into the virtual classroom, an entry link to the virtual space can be pushed to intelligent terminals such as the mobile phone used by the user. After the user clicks on this link, they enter. The virtual teacher will independently decide on multiple teaching links that need to be taught successively based on the best learning materials. A teaching link is a link for teaching using a certain sub-material in the best learning materials. For example, if a certain sub-material is an English word speaking exercise, the corresponding teaching link is the link where the virtual teacher leads the user to conduct the English word speaking exercise. Sort the teaching links in the order of teaching sequence to obtain a teaching link sequence. The teaching importance of a teaching link represents the degree of teaching importance of the teaching link, and can be counted as the difficulty of the teaching objectives of different teaching links. The importance threshold is a threshold representing a relatively large importance, for example: 8. The first time exceeding the importance threshold refers to the teaching importance of the first teaching link from the front to the back in the teaching link sequence exceeding the importance threshold, which makes the value of D unique. The teaching importance of the D-th teaching link is relatively high, and the subsequent teaching links depend on its teaching effect. Therefore, set i≥D, and starting from the D-th teaching link, identify whether the user is distracted during the lecture to improve the identification efficiency.

[0090] When it is recognized that the user is distracted during the lecture, identify whether they generate effective distraction content. Effective distraction content refers to content that is beneficial to further optimizing subsequent teaching links after the user is distracted during the lecture. When it is recognized that the user generates effective distraction content, based on the effective distraction content, optimize and replace the teaching links from the i-th to the i + j-th in the teaching link sequence. In the quantization value library, quantization values corresponding to the number of times of generating different effective distraction contents in different histories are set. The larger the number of times of generating different effective distraction contents in history, the more times it represents that teaching link optimization is needed in history, and the higher the depth of teaching link optimization required this time, and the corresponding quantization value is larger. Therefore, there is a direct proportional relationship between the quantization value and the number of times of generating different effective distraction contents in history. Determine the number of subsequent teaching links for optimized replacement based on the proportional quantization value, which improves the suitability of the optimized replacement. When it is recognized that the user does not generate effective distraction content, it means that the user needs to be reminded in time after being distracted, and control the virtual teacher to successively give distraction reminders and distraction remedy assistance to the user.

[0091] In the prior art, when it is recognized that the learner in class is distracted, the learner is directly prompted not to be distracted and to focus on the class. However, the degree of freedom of operation of the intelligent terminal used by the learner in class is very large. When the learner may be operating a note-taking software to take class notes or searching for other supporting exercises to complete while listening to the class, etc., it may be recognized by the system that the learner is distracted and reminded. This relatively rigid distraction recognition and reminder mechanism seriously reduces the learning experience of the learner. The embodiment of the present invention can solve this problem: recognize whether the user is distracted during class. When it is recognized as yes, recognize whether the user generates effective distracting content. When it is recognized as yes, based on the effective distracting content, optimize and replace the i-th to (i + j)-th teaching links in the teaching link sequence. Otherwise, control the virtual teacher to successively give the user distraction reminders and distraction remedy assistance, introduce the effective distracting content, and further judge whether to give the user distraction prompts based on whether it generates, which greatly improves the learning experience of the learner and improves the applicability of the system.

[0092] In addition, based on the effective distracting content, the i-th to (i + j)-th teaching links in the teaching link sequence are optimized and replaced, and the virtual teacher continues to teach based on the optimized and replaced teaching link sequence of the i-th to (i + j)-th teaching links, which improves the effect of the virtual teacher's English teaching for the user.

[0093] In one embodiment, the recognition of whether the user is distracted during class includes:

[0094] Continuously obtain multiple listening behaviors of the user;

[0095] Based on the concentration analysis library, determine the concentration of each listening behavior;

[0096] Based on the concentration of each listening behavior and the time when each listening behavior occurs, draw a concentration curve; wherein, the horizontal axis of the concentration curve is time, and the vertical axis is the magnitude of the concentration;

[0097] Identify whether there is a target trough value lower than the trough value threshold for the first time in the concentration curve;

[0098] When it is recognized as yes, determine that the user is distracted during class;

[0099] Otherwise, determine that the user is not distracted during class.

[0100] The listening behavior at least includes: the action behavior, language behavior, etc. of the user during listening. When obtaining the listening behavior, the action images, speech, etc. can be obtained based on the camera device, sound pickup device, etc. of the intelligent terminal used by the user, and corresponding analysis can be carried out to obtain it. In the focus analysis library, there is a focus corresponding to different listening behaviors, and the focus is the degree of the user's listening focus reflected by the listening behavior. In the focus curve, first, based on the focus of each listening behavior and the moment when the behavior occurs, coordinates are determined in the curve, and then the coordinates are connected in sequence according to the time sequence of the coordinates to obtain the focus curve. The trough value threshold is the threshold representing a relatively low focus, for example: 4. The target trough value that first appears below the trough value threshold refers to the newly appeared trough value and it is lower than the trough value threshold. If the target trough value that first appears below the trough value threshold appears in the focus curve, it means that the user newly generates a situation with relatively low listening focus after this recognition, that is, distraction occurs. By introducing the focus curve and determining whether the user has listening distraction based on whether the target trough value that first appears below the trough value threshold appears in the recognized focus curve, the recognition accuracy, comprehensiveness, and efficiency of whether the user has listening distraction are improved.

[0101] In one embodiment, the recognition of whether the user generates effective distraction content includes:

[0102] Taking the occurrence moment of the target trough value as the first boundary moment;

[0103] Recognizing whether there are more than the number threshold of target peak values higher than the peak value threshold continuously appearing immediately after the first boundary moment in the focus curve, and the last target peak value appearing is higher than all other target peak values;

[0104] When the recognition result is yes, taking the occurrence moment of the last target peak value as the second boundary moment; otherwise, determining that the user does not generate effective distraction content;

[0105] Recognizing whether the user accesses an autonomous operation scenario that has a standard association relationship with the i-th teaching link between the first boundary moment and the second boundary moment;

[0106] When the recognition result is yes, determining that the user generates effective distraction content, obtaining the scenario operation dynamics of the autonomous operation scenario at the second boundary moment, and taking it as the effective distraction content; otherwise, determining that the user does not generate effective distraction content.

[0107] Take the occurrence time of the target trough value as the first boundary time, which is also the time when the user is distracted during the lecture. When the learner operates the note-taking software to record lecture notes or searches for other supporting exercises to complete while listening to the lecture, etc., there will be continuous and multiple situations of being more focused during the lecture (for example: while looking at the note interface, while facing the virtual teacher and continuing to listen to the lecture). In addition, if the learner finishes operating the note-taking software to record lecture notes or searches for other supporting exercises to complete while listening to the lecture, etc., the last more focused situation generated will be higher than the previous more focused situations (stop operating and fully engage in continuing to listen to the virtual teacher's lecture), that is, in the focus curve, immediately after the first boundary time, there are more than the number threshold of target peak values higher than the peak value threshold continuously appearing, and the last target peak value appearing is higher than all other target peak values. Take the occurrence time of the last target peak value as the second boundary time, which is also the time when the user may end generating effective distracting content. The autonomous operation scenario is the scenario where the user operates autonomously on the intelligent terminal, such as: the note scenario, the exercise scenario, etc. The standard association relationship represents the relationship between the autonomous operation scenario and the teaching link, such as: the notes taken in the note scenario are related to the teaching link, etc. When it is recognized as yes, it is determined that the user generates effective distracting content, and the scenario operation dynamics of the autonomous operation scenario at the second boundary time are obtained and used as the effective distracting content. The scenario operation dynamics are the information finally completed by the user in the autonomous operation scenario, such as: the final draft notes, etc.

[0108] In an embodiment of the present invention, when identifying whether the user generates effective distracting content, it is identified whether there are more than the number threshold of target peak values higher than the peak value threshold continuously appearing immediately after the first boundary time in the focus curve, and the last target peak value appearing is higher than all other target peak values to determine whether the user generates effective distracting content, which improves the accuracy and efficiency of the first determination; secondly, take the occurrence time of the last target peak value as the second boundary time, and combine the first boundary time to determine the recognition time period for identifying whether the user accesses the autonomous operation scenario having a standard association relationship with the i-th teaching link, further improving the recognition efficiency, obtaining the scenario operation dynamics of the autonomous operation scenario at the second boundary time and using it as the effective distracting content, which improves the accuracy, comprehensiveness and efficiency of the selection of the effective distracting content.

[0109] In one embodiment, the optimizing and replacing the i-th to the (i + j)-th teaching links in the teaching link sequence based on the effective distracting content includes:

[0110] Extract features from the effective distracting content, the i-th to the (i + j)-th teaching links, and the link association relationships between every two of the i-th to the (i + j)-th teaching links to obtain a multi-dimensional feature set;

[0111] Construct an optimized replacement scenario vector based on the multi-dimensional feature set;

[0112] Determine the optimized replacement strategy corresponding to the optimized replacement scenario vector from the optimized replacement strategy library;

[0113] Based on the optimized replacement strategy, perform optimized replacement on the \(i\)th to the \((i + j)\)th teaching links.

[0114] The features in the multi-dimensional feature set at least include: the content type of the effective distraction content, the teaching link type of the \(i\)th to the \((i + j)\)th teaching links, the link association relationship type, etc.; the multi-dimensional feature set is constructed into an optimized replacement scenario vector in vector form, which represents the scenario that requires optimized replacement of teaching links. Different optimized replacement strategies corresponding to the optimized replacement scenario vectors are preset in the optimized replacement strategy library, retrieved from the library, and based on it, the \(i\)th to the \((i + j)\)th teaching links are optimized and replaced. For example: if the effective distraction content is the listening notes recorded by the user, the types of the \(i\)th to the \((i + j)\)th teaching links are: interactive Q&A, practice questions, group discussion, and the link association relationship types include link associations such as concept application and cooperative learning, then the optimized replacement strategy corresponding to the constructed optimized replacement scenario vector is that if the learner has difficulty understanding after the interactive Q&A, it can be transitioned to the practice questions by adding examples or simpler questions, or some of the practice questions can be replaced with more interactive content (such as animations, interactive videos).

[0115] In one embodiment, controlling the virtual teacher to give distraction reminders and distraction remediation assistance to the user successively includes:

[0116] Control the virtual teacher to pause the teaching of the \(i\)th teaching link;

[0117] Control the virtual teacher to interact with the user for distraction reminder;

[0118] Generate learning review materials based on the taught content of the \(i\)th teaching link by the virtual teacher within the preset time before and after the first boundary moment;

[0119] Based on the learning review materials, control the virtual teacher to conduct review learning for the user.

[0120] When conducting distraction reminder interaction, the virtual teacher can be controlled to emit a voice prompting the user to focus on learning and identify whether the user is focused. The preset time can be 20 seconds. Generate learning review materials based on the taught content of the \(i\)th teaching link by the virtual teacher within the preset time before and after the first boundary moment. The learning review materials are materials to help the user review the taught content. Based on the learning review materials, control the virtual teacher to conduct review learning for the user. After the review learning, then continue with the normal learning teaching.

[0121] An embodiment of the present invention provides an English teaching system based on artificial intelligence, as Figure 2 shown, including:

[0122] An English learning status acquisition module 1 for acquiring the English learning status of a user;

[0123] A best learning material matching module 2 for matching the best learning materials for the user based on artificial intelligence according to the English learning status;

[0124] An English teaching module 3 for teaching the user English based on a virtual teacher according to the best learning materials.

[0125] The best learning material matching module matches the best learning materials for the user based on artificial intelligence according to the English learning status, including:

[0126] Retrieving a pre-trained artificial intelligence model;

[0127] Based on the artificial intelligence model, matching the best learning materials for the user according to the English learning status;

[0128] Among them, the pre-training steps of the artificial intelligence model are as follows:

[0129] Collecting training samples; among them, the training samples include: a large number of pre-labeled preparatory learning materials adapted to the English learning status;

[0130] Training the artificial intelligence model based on a machine learning algorithm according to the training samples.

[0131] The English teaching module teaches the user English based on a virtual teacher according to the best learning materials, including:

[0132] Guiding the user into a virtual classroom;

[0133] Based on the best learning materials, controlling the virtual teacher to start teaching the user English in the virtual classroom; among them, the virtual teacher self-decides the teaching link sequence and conducts corresponding teaching according to the teaching link sequence;

[0134] When the virtual teacher conducts English teaching to the i-th teaching link in the teaching link sequence, identifying whether the user is distracted during the class; where i≥D, and the importance of the teaching of the D-th teaching link in the teaching link sequence first exceeds the importance threshold;

[0135] When it is identified as yes, identifying whether the user generates effective distracting content;

[0136] When it is recognized as yes, based on the effective distraction content, the i-th to the (i + j)-th teaching links in the teaching link sequence are optimized and replaced; wherein, the virtual teacher continues teaching relay based on the optimized and replaced teaching link sequence of the i-th to the (i + j)-th teaching links; j is the proportional quantization value corresponding to the number of times the user has generated different effective distraction contents in the quantization value library.

[0137] Otherwise, control the virtual teacher to give distraction reminders and distraction remedy assistance to the user successively.

[0138] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these changes and modifications.

Claims

1. An English teaching method based on artificial intelligence, characterized in that: include: Get the user's English learning status; Based on artificial intelligence, the best learning materials are matched to users according to their English learning status; Based on virtual teachers, users are taught English according to the best learning materials.

2. The English teaching method based on artificial intelligence as claimed in claim 1, characterized in that: Based on artificial intelligence, the best learning materials are matched for users according to their English learning status, including: Retrieve pre-trained AI models; Based on the artificial intelligence model, the best learning materials are matched for users according to their English learning status; Among them, the pre-training steps of the artificial intelligence model are as follows: Collect training samples; wherein the training samples include: a large amount of preparatory learning materials that have been labeled and adapted to English learning conditions; Based on machine learning algorithms, artificial intelligence models are trained according to training samples.

3. The English teaching method based on artificial intelligence as claimed in claim 1, characterized in that: The virtual teacher provides English teaching to users based on the best learning materials, including: Guide users into the virtual classroom; Based on the best learning materials, the virtual teacher is controlled to start teaching English to the user in the virtual classroom; wherein the virtual teacher decides the teaching link sequence by itself and conducts corresponding teaching according to the teaching link sequence; When the virtual teacher teaches English to the i-th teaching link in the teaching link sequence, it is identified whether the user is distracted in the class; wherein, i≥D, the importance of the teaching of the D-th teaching link in the teaching link sequence exceeds the importance threshold for the first time; When the identification is yes, identifying whether the user generates effective distracting content; When the identification is yes, based on the effective distracting content, the teaching links from the ith to the i+jth in the teaching link sequence are optimized and replaced; wherein the virtual teacher continues to teach based on the teaching link sequence after the optimization and replacement of the ith to the i+jth teaching links; j is the proportional quantitative value corresponding to the number of times the user has generated different effective distracting contents in the quantitative value library in the history; Otherwise, the virtual teacher is controlled to provide distraction reminders and distraction remediation assistance to the user in turn.

4. The English teaching method based on artificial intelligence as claimed in claim 3, characterized in that: The identifying whether the user is distracted during class includes: Continuously obtain multiple lecture behaviors of users; Based on the concentration analysis library, determine the concentration level of each listening behavior; Based on the concentration of each listening behavior and the time when each listening behavior occurs, a concentration curve is drawn; wherein the horizontal axis of the concentration curve is time, and the vertical axis is the concentration level; Identify whether the target trough value below the trough value threshold appears for the first time in the concentration curve; When the identification is yes, it is determined that the user is distracted in the class; Otherwise, it is determined that the user is not distracted while listening to the class.

5. The English teaching method based on artificial intelligence as claimed in claim 4, characterized in that: The identifying whether the user generates effective distracting content includes: The moment when the target trough value appears is taken as the first boundary moment; Identify whether, in the concentration curve, more than a threshold number of target wave peaks higher than the wave peak threshold appear successively after the first limit moment and whether the last target wave peak that appears is higher than all other target wave peaks; When the identification is yes, the appearance time of the last target peak value is used as the second limit time; otherwise, it is determined that the user has not generated effective distracting content; Identify whether the user accesses an autonomous operation scene having a standard association relationship with the i-th teaching link between the first boundary time and the second boundary time; When the identification is yes, it is determined that the user has generated effective distracting content, and the scene operation dynamics of the autonomous operation scene at the second boundary moment are obtained and used as the effective distracting content; otherwise, it is determined that the user has not generated effective distracting content.

6. The English teaching method based on artificial intelligence as claimed in claim 3, characterized in that: The optimizing and replacing the teaching links from the i-th to the i+j-th in the teaching link sequence based on the effective distraction content includes: Extract features of effective distraction content, the i-th to i+j-th teaching links, and the link correlations between the i-th to i+j-th teaching links to obtain a multi-dimensional feature set; Based on the multi-dimensional feature set, an optimized replacement scenario vector is constructed; Determine the optimization replacement strategy corresponding to the optimization replacement situation vector from the optimization replacement strategy library; Based on the optimization replacement strategy, the i-th to i+j-th teaching links are optimized and replaced.

7. The English teaching method based on artificial intelligence as claimed in claim 3, characterized in that: The controlling virtual teacher successively reminds the user of distraction and assists in distraction remediation, including: Control the virtual teacher to suspend the teaching of the i-th teaching session; Control the virtual teacher to interact with the user to remind them of distraction; Generate learning review materials based on the teaching content of the virtual teacher in the i-th teaching session within a preset time before and after the first boundary moment; Based on the learning review materials, the virtual teacher is controlled to review the learning of the user.

8. An English teaching system based on artificial intelligence, characterized in that: include: English learning status acquisition module, used to obtain the user's English learning status; The best learning material matching module is used to match the best learning material for users based on artificial intelligence and their English learning status; The English teaching module is used to teach English to users based on a virtual teacher and the best learning materials.

9. The artificial intelligence-based English teaching system according to claim 8, characterized in that: The optimal learning material matching module is based on artificial intelligence and matches the best learning material for the user according to the English learning situation, including: Retrieve pre-trained AI models; Based on the artificial intelligence model, the best learning materials are matched for users according to their English learning status; Among them, the pre-training steps of the artificial intelligence model are as follows: Collect training samples; wherein the training samples include: a large amount of preparatory learning materials that have been labeled and adapted to English learning conditions; Based on machine learning algorithms, artificial intelligence models are trained according to training samples.

10. The artificial intelligence-based English teaching system according to claim 8, characterized in that: The English teaching module is based on a virtual teacher and provides English teaching to users according to the best learning materials, including: Guide users into the virtual classroom; Based on the best learning materials, the virtual teacher is controlled to start teaching English to the user in the virtual classroom; wherein the virtual teacher decides the teaching link sequence by itself and conducts corresponding teaching according to the teaching link sequence; When the virtual teacher teaches English to the i-th teaching link in the teaching link sequence, it is identified whether the user is distracted in the class; wherein, i≥D, the importance of the teaching of the D-th teaching link in the teaching link sequence exceeds the importance threshold for the first time; When the identification is yes, identifying whether the user generates effective distracting content; When the identification is yes, based on the effective distracting content, the teaching links from the ith to the i+jth in the teaching link sequence are optimized and replaced; wherein the virtual teacher continues to teach based on the teaching link sequence after the optimization and replacement of the ith to the i+jth teaching links; j is the proportional quantitative value corresponding to the number of times the user has generated different effective distracting contents in the quantitative value library in the history; Otherwise, the virtual teacher is controlled to provide distraction reminders and distraction remediation assistance to the user in turn.

Images