Teaching content generation method, system and device based on artificial intelligence and medium

Abstract

The invention discloses a teaching content generation method, system and device based on artificial intelligence and a medium, and relates to the technical field of teaching content generation. A plurality of content tags in a teaching outline sent by a client are received and preliminarily screened in a database to obtain a first target set; obtaining the complexity level of each module according to the character vocabulary number, the picture number and the video number of each module in the first target set; screening out a second target set according to the complexity level of each module in the first target set; for each module in the second target set, calculating a priority selection index, and combining a plurality of modules with the maximum priority selection index in the second target set to generate final teaching content, so that a teacher can conveniently select, and the finally combined teaching content meets the requirements of teaching courseware; and meanwhile, the time and energy of the teacher during selection and arrangement are reduced, the burden of teaching courseware preparation is relieved, and the influence on lesson preparation efficiency is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of teaching content generation, and in particular to a method, system, device and medium for generating teaching content based on artificial intelligence. Background Art

[0002] Teaching content generation based on artificial intelligence refers to the use of AI technology to automatically analyze the teaching materials required by users, screen and generate personalized corresponding teaching materials from massive teaching resources, so as to customize teaching content suitable for each user, thereby improving teaching efficiency and teaching effect; especially when teachers need to generate teaching courseware content, teachers can input the teaching material labels they need on the client (device storing teaching content), screen and generate corresponding teaching content, which greatly saves teachers' time in making courseware and enables teachers to devote more energy to teaching methods and classroom interaction;

[0003] When teachers are screening teaching courseware, in addition to entering the teaching material labels they need on the client, they also screen separately according to the approximate time of the entire teaching courseware, each teaching content type module (such as introduction module, explanation module, interactive module, etc.) and the corresponding teaching time of each teaching content type module. The client combines each screened teaching content type module to help teachers accurately match teaching content, make time allocation reasonable, and make the class smooth. The phased screening can flexibly arrange the teaching rhythm, thereby improving classroom efficiency and teaching effect;

[0004] However, when screening and combining actual teaching content type modules, screening teaching materials only through labels, teaching content type modules and corresponding teaching times may result in the screening of teaching content type modules with a relatively chaotic content complexity, making it more difficult for teachers to select, and making the final combination of teaching content not meet the requirements of teaching courseware; in addition, a large number of teaching content type modules screened out through screening teaching materials through labels, teaching content type modules and corresponding teaching times may cause teachers to spend more time and energy in selection and arrangement, increasing the burden of preparing teaching courseware and possibly affecting the efficiency of lesson preparation. Summary of the invention

[0005] The purpose of the present invention is to solve the above-mentioned problems and to provide a method, system, device and medium for generating teaching content based on artificial intelligence.

[0006] In a first aspect of the implementation of the present invention, a method for generating teaching content based on artificial intelligence is first proposed. The method is applied in a cloud server and comprises the following steps:

[0007] Receive a teaching outline sent by a client; the teaching outline includes a course tag and multiple content tags; the target content tag is any one of the multiple content tags, and the target content tag includes a type tag and a teaching time tag;

[0008] Filtering the modules that match the course label and the target content label in the database as the first target set, determining the number of words, pictures and videos in each module in the first target set to obtain the complexity level of each module;

[0009] Sending the complexity level of each module in the first target set to the client; so that the client receives the selection instruction and selects the second target set that meets the selection instruction from the first target set;

[0010] For each module in the second target set, its usage frequency, corresponding time and form type are obtained to obtain a priority index; so that the client can combine the modules with the largest priority index in the second target set of multiple content tags to generate the final teaching content.

[0011] Optionally, determining the number of words, pictures and videos in each module in the first target set to obtain the complexity level of each teaching material includes:

[0012] Obtain the number of different words and the total number of words in the text contained in each module, and divide the number of different words by the total number of words to obtain the lexical richness;

[0013] Obtain the total number of sentences contained in each module, and divide the total number of words by the total number of sentences to obtain the average sentence length;

[0014] The vocabulary richness and average sentence length of each module are normalized, and the text complexity of each module is obtained based on the normalized vocabulary richness and average sentence length;

[0015] The calculation formula is: In the formula, dc and df are the normalized vocabulary richness and average sentence length, Bn is the text complexity, a1 and a2 are the preset proportional coefficients of dc and df, and a1 and a2 are both greater than 0;

[0016] The complexity level of each module is obtained based on the text complexity, number of pictures and number of videos in each module.

[0017] Optionally, the complexity level of each module is obtained according to the text complexity, the number of pictures and the number of videos of each module, including:

[0018] Obtain the total number of images contained in each module, and decompose each image into different frequency sub-bands by wavelet transform using Daubechies wavelet function;

[0019] For each frequency subband, calculate the grayscale histogram H s (i), H represents the grayscale histogram, s represents a frequency subband obtained after wavelet decomposition, and i represents the grayscale level;

[0020] Calculate the information entropy of each subband, the calculation formula is:

[0021]

[0022] In the formula, represents the information entropy of the nth subband, P s (i)·log2(P s (i)) represents the contribution of each gray level i to the subband entropy value;

[0023] Calculate the information entropy index of the entire image. The calculation formula is: In the formula, EFG is the information entropy index of the entire image, and u is the total number of sub-bands;

[0024] Calculate the mean information entropy index of all images contained in each module as the image complexity of each module;

[0025] Get the number of videos contained in each module, extract frames from each video to obtain the image set of the corresponding video, and calculate the mean information entropy index of the images in the video image set as the complexity index of the corresponding video, and take the mean complexity index of all videos contained in each module as the video complexity of each module;

[0026] The complexity coefficient of each module is obtained according to the text complexity, image complexity and video complexity of each module, and the complexity level of each module is obtained according to the preset stage complexity coefficient threshold.

[0027] Optionally, obtaining a complexity coefficient of each module according to the text complexity, image complexity and video complexity of each module, and obtaining a complexity level of each module according to a preset stage complexity coefficient threshold includes:

[0028] The text complexity, image complexity and video complexity of each module are normalized, and the complexity coefficient of each module is obtained according to the normalized text complexity, image complexity and video complexity. The calculation formula is:

[0029] Ghc=b1×tgh+b2×yhl+b3×kmu

[0030] Where Ghc is the complexity coefficient, tgh, yhl, kmu are the normalized text complexity, image complexity, and video complexity, b1, b2, and b3 are the preset proportional coefficients of tgh, yhl, and kmu, respectively, and b1, b2, and b3 are all greater than 0;

[0031] Comparing the complexity coefficient of each module with a preset first complexity coefficient threshold and a preset second complexity coefficient threshold, and the preset first complexity coefficient threshold is less than the preset second complexity coefficient threshold;

[0032] When the complexity coefficient is not greater than a preset first complexity coefficient threshold, the complexity level of the corresponding module is marked as simple;

[0033] When the complexity coefficient is greater than a preset first complexity coefficient threshold but less than a preset second complexity coefficient threshold, the complexity level of the corresponding module is marked as medium;

[0034] When the complexity coefficient is not less than a preset complexity coefficient second threshold, the complexity level of the corresponding module is marked as difficult.

[0035] Optionally, for each module in the second target set, obtaining its usage frequency, corresponding time and form type to obtain a preference index includes:

[0036] The total number of times each module in the second target set is used, the frequency of use, and the average value of the use effect score are obtained, and normalized processing is performed to obtain the use effect index of each module. The calculation formula is: hd = ln (c1 × rt + c2 × ed + c3 × ws + 1), where rt, ed, and ws are the total number of times, frequency of use, and average value of the use effect score after normalization processing; c1, c2, and c3 are the preset proportional coefficients of rt, ed, and ws, respectively, and c1, c2, and c3 are all greater than 0;

[0037] Obtain the actual duration of the teaching content corresponding to each module in the second target set, calculate the absolute difference between the actual duration and the preset standard time, and use the absolute difference as the time deviation index;

[0038] According to the use effect index, time deviation index and form type of each module in the second target set, the priority selection index of the corresponding module is obtained.

[0039] Optionally, obtaining a priority index of a corresponding module according to the use effect index, time deviation index and form type of each module in the second target set includes:

[0040] Obtain all form types contained in each module in the second target set, and obtain the usage count and preset interest weight value corresponding to each form type, and obtain the form diversity coefficient of each module. The calculation formula is: In the formula, TE is the form diversity index, e i is the number of times the i-th form type is used, ω i is the preset interest weight value corresponding to the i-th form type, and q is the total number of form types;

[0041] According to the use effect index, time deviation index and form diversity index of each module in the second target set, the priority selection index of the corresponding module is obtained.

[0042] Optionally, obtaining the priority index of the corresponding module according to the usage effect index, time deviation index and form diversity index of each module in the second target set includes:

[0043] The use effect index, time deviation index and form diversity index are normalized, and the priority index of the corresponding module is obtained according to the normalized use effect index, time deviation index and form diversity index. The calculation formula is:

[0044]

[0045] Wherein, sx, zp, and mp are respectively the normalized usage effect index, time deviation index, and form diversity index; f1, f2, and f3 are respectively the preset proportional coefficients of sx, zp, and mp, and f1, f2, and f3 are all greater than 0.

[0046] In a second aspect of the present invention, a system for generating teaching content based on artificial intelligence is provided, the system comprising a client and a cloud server, wherein the cloud server and the client are connected, wherein:

[0047] The client sends a teaching outline to the cloud server; the teaching outline includes a course tag and a plurality of content tags; the target content tag is any one of the plurality of content tags, and the target content tag includes a type tag and a teaching time tag;

[0048] The cloud server is used to select modules that match the course label and the target content label in the database as the first target set, determine the number of text words, the number of pictures and the number of videos in each module in the first target set to obtain the complexity level of each module; and send the complexity level of each module in the first target set to the client;

[0049] The client is also used to receive a selection instruction to filter out a second target set that meets the selection instruction from the first target set;

[0050] The cloud server is also used to obtain the usage frequency, corresponding time and form type of each module in the second target set to obtain a priority selection index;

[0051] The client is also used to prioritize the modules with the largest indexes from the second targets of the multiple content tags to combine and generate final teaching content.

[0052] In a third aspect of the present invention, a device for generating teaching content based on artificial intelligence is provided. The device is applied in a cloud server and comprises:

[0053] The first screening module: receiving a teaching outline sent by a client; the teaching outline includes a course tag and a plurality of content tags; the target content tag is any one of the plurality of content tags, and the target content tag includes a type tag and a teaching time tag;

[0054] Complexity level module: Filter the modules that match the course label and the target content label in the database as the first target set, determine the number of words, pictures and videos of each module in the first target set to obtain the complexity level of each module;

[0055] The second screening module: sends the complexity level of each module in the first target set to the client, so that the client receives the selection instruction and screens out the second target set that meets the selection instruction from the first target set;

[0056] Content generation module: for each module in the second target set, obtain its usage frequency, corresponding time and form type to obtain a priority index; so that the client can combine the modules with the largest priority index in the second target set of multiple content tags to generate the final teaching content.

[0057] Beneficial effects of the present invention:

[0058] The present invention proposes a method, system, device and medium for generating teaching content based on artificial intelligence. The method comprises the following steps: receiving a teaching syllabus sent by a client, filtering out modules that match a course label and a target content label in a database as a first target set, determining the number of text words, the number of pictures and the number of videos in each module in the first target set to obtain a complexity level of each module; and sending the complexity level of each module in the first target set to a client, so that the client receives a selection instruction and filters out a second target set that matches the selection instruction from the first target set; for each module in the second target set, obtaining its usage frequency, corresponding time and form type to obtain a priority selection index; and allowing the client to combine the modules with the largest priority selection index in the second target sets of multiple content labels to generate the final teaching content. In this way, it is more convenient for teachers to make selections, and the final combined teaching content meets the requirements of teaching courseware; at the same time, it reduces the time and energy of teachers in selection and arrangement, alleviates the burden of preparing teaching courseware, and reduces the impact on preparation efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] The present invention will be further described below in conjunction with the accompanying drawings.

[0060] Figure 1 A flowchart of a method for generating teaching content based on artificial intelligence;

[0061] Figure 2 A framework diagram of a system for generating teaching content based on artificial intelligence;

[0062] Figure 3 A framework diagram of a device for generating teaching content based on artificial intelligence. DETAILED DESCRIPTION

[0063] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0064] Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in the field without making any creative work shall fall within the scope of protection of the present invention.

[0065] The embodiment of the present invention provides a method for generating teaching content based on artificial intelligence. Figure 1 , Figure 1 A flowchart of a method for generating teaching content based on artificial intelligence provided by an embodiment of the present invention. The method is applied in a cloud server and includes the following steps:

[0066] receiving a teaching syllabus sent by a client; the teaching syllabus includes a course tag and multiple content tags; the target content tag is any one of the multiple content tags, and the target content tag includes a type tag and a teaching time tag;

[0067] Modules matching the course labels and target content labels are selected from the database as the first target set, and the number of words, pictures and videos of each module in the first target set is determined to obtain the complexity level of each module;

[0068] Sending the complexity level of each module in the first target set to the client; so that the client receives the selection instruction and selects the second target set that meets the selection instruction from the first target set;

[0069] For each module in the second target set, its usage frequency, corresponding time and form type are obtained to obtain a priority index; so that the client can combine the modules with the largest priority index in the second target set of multiple content tags to generate the final teaching content.

[0070] Based on the method for generating teaching content based on artificial intelligence provided by the embodiment of the present invention, through the above-mentioned method, it is more convenient for teachers to make choices, so that the final combined teaching content meets the needs of teaching courseware; at the same time, it reduces the time and energy of teachers in selection and arrangement, alleviates the burden of preparing teaching courseware, and reduces the impact on preparation efficiency.

[0071] It should be noted that the teacher inputs the required syllabus on the client, and the syllabus includes a course tag and multiple content tags; the target content tag is any one of the multiple content tags, and the target content tag includes a type tag and a teaching time tag;

[0072] For example, a teacher inputs the required course label (such as "mathematics, Fourier function", etc.) on the client. For multiple content labels, suppose there are three content labels in total. One content label is "import", and the type label corresponding to "import" is "import module", and the corresponding teaching time label is "import module 15 minutes"; one content label is "explain", and the type label corresponding to "explain" is "explain module", and the corresponding teaching time label is "explain module 30 minutes"; one content label is "interaction", and the type label corresponding to "interaction" is "interaction module", and the corresponding teaching time label is "interaction module 15 minutes";

[0073] According to the above-mentioned teaching syllabus, the modules that match the course label and the target content label are selected in the database as the first target set; for example, the first target set of the "introduction" content label includes several introduction modules, and each introduction module lasts about 15 minutes; the first target set of the "explanation" content label includes several explanation modules, and each explanation module lasts about 30 minutes; the first target set of the "interaction" content label includes several interaction modules, and each interaction module lasts about 15 minutes;

[0074] In addition, the content corresponding to each module in the first target set of all content tags is "mathematics, Fourier function"; this preliminary screening is mainly completed through artificial intelligence algorithms and data analysis technology. The client automatically matches relevant teaching content modules from the huge teaching resource database based on the teaching syllabus input by the teacher. It can quickly and accurately screen out teaching materials that meet the needs of teachers, reduce the tedious operations of teachers in lesson preparation, and at the same time ensure the pertinence and systematicness of the teaching content, thereby improving the efficiency and quality of teaching preparation.

[0075] In one embodiment, determining the number of words, pictures, and videos in each module of the first target set to obtain the complexity level of each teaching material includes:

[0076] Obtain the number of different words and the total number of words in the text contained in each module, and divide the number of different words by the total number of words to obtain the lexical richness of each module;

[0077] Obtain the total number of sentences contained in each module, and divide the total number of words by the total number of sentences to obtain the average length of sentences in each module;

[0078] The vocabulary richness and average sentence length of each module are normalized, and the text complexity of each module is obtained based on the normalized vocabulary richness and average sentence length;

[0079] The calculation formula is: In the formula, dc and df are the normalized vocabulary richness and average sentence length, Bn is the text complexity, a1 and a2 are the preset proportional coefficients of dc and df, and a1 and a2 are both greater than 0;

[0080] The complexity level of each module is obtained based on the text complexity, number of pictures and number of videos in each module.

[0081] It should be noted that a1 and a2 are set by professionals according to actual circumstances. Generally, the sum of a1 and a2 is 1. For example, a1 and a2 can be 0.4, 0.6, or other numbers, without specific limitation. In addition, commonly used normalization methods include Min-Max normalization, Z-Score standardization, etc. The specific method is selected by professionals according to actual circumstances, and no specific limitation or elaboration is made.

[0082] It should be noted that the text content of each module can be analyzed using natural language processing (NLP) technology to extract the number of different words, the total number of words, and the total number of sentences; this can be achieved using tools such as word segmentation algorithms and word frequency statistics, or other methods can be used for extraction, which will not be limited or elaborated on in detail;

[0083] It should be noted that the greater the text complexity of each module, the more complex the teaching content in the corresponding module. The reason is that the increase in text complexity usually means the use of more unique words and longer sentences, which indicates that the ideas and concepts expressed are richer and deeper. Therefore, when choosing teaching modules, teachers can judge their suitability based on the text complexity of the module to ensure that the selected content meets the teacher's requirements.

[0084] In one embodiment, the complexity level of each module is obtained according to the text complexity, the number of pictures and the number of videos of each module, including:

[0085] Obtain the total number of images contained in each module, and decompose each image into different frequency sub-bands by wavelet transform using Daubechies wavelet function;

[0086] For each frequency subband, calculate the grayscale histogram H s (i), H represents the grayscale histogram, s represents a frequency subband obtained after wavelet decomposition, and i represents the grayscale level;

[0087] Calculate the information entropy of each subband, the calculation formula is:

[0088]

[0089] In the formula, represents the information entropy of the nth subband, P s (i)·log2(P s (i)) represents the contribution of each gray level i to the subband entropy value;

[0090] Calculate the information entropy index of the entire image. The calculation formula is: In the formula, EFG is the information entropy index of the entire image, and u is the total number of sub-bands;

[0091] Calculate the mean information entropy index of all images contained in each module as the image complexity of each module;

[0092] Get the number of videos contained in each module, extract frames from each video to obtain the image set of the corresponding video, and calculate the mean information entropy index of the images in the video image set as the complexity index of the corresponding video, and take the mean complexity index of all videos contained in each module as the video complexity of each module;

[0093] The complexity coefficient of each module is obtained according to the text complexity, image complexity and video complexity of each module, and the complexity level of each module is obtained according to the preset stage complexity coefficient threshold.

[0094] It should be noted that the frame extraction of each video can be achieved through a video processing library (such as OpenCV, FFmpeg, etc.). These tools can easily read video files and extract specific frames. Other methods may also be used, which are not limited or elaborated on in detail.

[0095] It should be noted that the information entropy index of an image refers to the complexity of the image content. The larger the information entropy index of an image, the more complex the content conveyed by the image. The reason is that information entropy measures the diversity and uniformity of the grayscale distribution in the image. Specifically, a higher information entropy value means that the image contains rich details and features, and the grayscale distribution is more extensive, reflecting the complexity of the image content, such as the combination of different objects, textures and shapes. A lower information entropy value indicates that the image content is relatively simple, the grayscale level is concentrated, and it may only contain a single color or shape. Therefore, by analyzing information entropy, we can effectively evaluate the expressiveness of the image in each module and the richness of the information conveyed, so as to more accurately judge its complexity level, and ultimately help teachers make more appropriate decisions in the selection of teaching content.

[0096] In one implementation, the Daubechies wavelet function is used for wavelet transform to decompose the image into different frequency sub-bands, which can effectively capture the multi-scale features and detail information of the image. The advantage of wavelet transform is that it has good time-frequency localization ability and can provide both spatial and frequency information of the image, which is especially important for analyzing complex images. By calculating the information entropy for each frequency sub-band, the complexity and information content of different frequency components can be quantified, thereby revealing the structure and texture characteristics of the image; this method not only improves the accuracy of the evaluation of the complexity of the image content, but also helps to select more expressive images in the generation of teaching content, thereby enhancing the teaching effect.

[0097] In one embodiment, the complexity coefficient of each module is obtained according to the text complexity, image complexity and video complexity of each module, and the complexity level of each module is obtained according to the preset stage complexity coefficient threshold, including:

[0098] The text complexity, image complexity and video complexity of each module are normalized, and the complexity coefficient of each module is obtained according to the normalized text complexity, image complexity and video complexity. The calculation formula is:

[0099] Ghc=b1×tgh+b2×yhl+b3×kmu

[0100] Where Ghc is the complexity coefficient, tgh, yhl, kmu are the normalized text complexity, image complexity, and video complexity, b1, b2, and b3 are the preset proportional coefficients of tgh, yhl, and kmu, respectively, and b1, b2, and b3 are all greater than 0;

[0101] Comparing the complexity coefficient of each module with a preset first complexity coefficient threshold and a preset second complexity coefficient threshold, and the preset first complexity coefficient threshold is less than the preset second complexity coefficient threshold;

[0102] When the complexity coefficient is not greater than the preset first threshold value of the complexity coefficient, the complexity level of the teaching content of the corresponding module is marked as simple;

[0103] When the complexity coefficient is greater than a preset first complexity coefficient threshold but less than a preset second complexity coefficient threshold, the complexity level of the teaching content of the corresponding module is marked as medium;

[0104] When the complexity coefficient is not less than a preset second threshold value of the complexity coefficient, the complexity level of the teaching content of the corresponding module is marked as difficult.

[0105] It should be noted that b1, b2, and b3 are set by professionals according to actual circumstances. Generally, the sum of b1, b2, and b3 is 1. For example, b1, b2, and b3 can be 0.3, 0.3, and 0.4, respectively, or other numbers without specific limitation. In addition, commonly used normalization methods include Min-Max normalization, Z-Score standardization, etc. The specific method is selected by professionals according to actual circumstances, and is not limited or elaborated on.

[0106] It should be noted that the preset first threshold value of the complexity coefficient and the preset second threshold value of the complexity coefficient are set by professionals according to actual conditions, and are not specifically limited or elaborated;

[0107] It should be noted that the complexity level of each module in the first target set is sent to the client, so that the client receives the selection instruction and selects the second target set that meets the selection instruction from the first target set; for example, for the topic of "Fourier function", if the teacher wants to select some basic Fourier transform concepts and simple examples in "Import" to ensure that all students can understand the basic background knowledge, the teacher can select the module with the complexity level of "simple" in the first target set of the "Import" content label as the selection instruction, and the client uses all the modules with the complexity level of "simple" in the first target set of the "Import" content label as the second target set;

[0108] In the explanation module, if the teacher wants to introduce slightly complex formula derivation and application examples to teach students in a slightly deeper way, the teacher can select the module with a complexity level of "medium" in the first target set of the "Explanation" content label as the selection instruction, and the client will use all the modules with a complexity level of "medium" in the first target set of the "Explanation" content label as the second target set;

[0109] Finally, in the interactive module, assuming that the teacher wants to design some challenging exercises or discussion questions to test the teaching effect, the teacher can select the module with a complexity level of "difficult" in the first target set of the "interactive" content label as the selection instruction, and the client will use all the modules with a complexity level of "difficult" in the first target set of the "interactive" content label as the second target set;

[0110] In this way, teachers can automatically select the complexity levels of different content labels, which can effectively improve the pertinence and flexibility of teaching. The systematic module selection also saves time for teachers and simplifies the lesson preparation process, allowing them to focus more on teaching itself and interaction with students, thereby improving the overall quality of the class.

[0111] In one embodiment, for each module in the second target set, obtaining its usage frequency, corresponding time, and form type to obtain a preference index includes:

[0112] The total number of times each module in the second target set is used, the frequency of use, and the average value of the use effect score are obtained, and normalized processing is performed to obtain the use effect index of each module. The calculation formula is: hd = ln (c1 × rt + c2 × ed + c3 × ws + 1), where rt, ed, and ws are the total number of times, frequency of use, and average value of the use effect score after normalization processing; c1, c2, and c3 are the preset proportional coefficients of rt, ed, and ws, respectively, and c1, c2, and c3 are all greater than 0;

[0113] Obtain the actual duration of the teaching content corresponding to each module in the second target set, calculate the absolute difference between the actual duration and the preset standard time, and use the absolute difference as the time deviation index;

[0114] According to the use effect index, time deviation index and form type of each module in the second target set, the priority selection index of the corresponding module is obtained.

[0115] It should be noted that c1, c2, and c3 are set by professionals according to actual circumstances. Generally, the sum of c1, c2, and c3 is 1. For example, c1, c2, and c3 can be 0.4, 0.3, and 0.3, respectively, or other numbers without specific limitation. In addition, commonly used normalization methods include Min-Max normalization, Z-Score standardization, etc. The specific method is selected by professionals according to actual circumstances, and no specific limitation or elaboration is made. In addition, the preset standard time is set by the teacher according to actual circumstances. For example, the preset standard time for the import module is 15 minutes, the preset standard time for the explanation module is 30 minutes, and the preset standard time for the interactive module is 15 minutes, etc. It depends on the actual situation and no limitation or elaboration is made.

[0116] It should be noted that the total number of times each module is used, the frequency of use, and the use effect score can be recorded through the background data collection function of the client management system (LMS). The actual teaching duration of each module can be obtained through the client's recording system, or other acquisition methods, which are not limited or elaborated in detail;

[0117] It should be noted that the better the usage effect index of each module in the second target set, the greater the priority index of the corresponding module, and the greater the priority as the final teaching content. The reason is that the better the usage effect index, the better the actual application performance of the module in the teaching process, which is specifically reflected in the frequent use, high frequency of use, and good student feedback scores. These factors together reflect the effectiveness and popularity of the module, thereby improving its priority index in the teaching content; therefore, the improvement of the usage effect index not only shows that the module can better meet the teaching objectives and student needs, but also helps teachers to be more targeted when selecting teaching content, ensuring that the final teaching content has a higher priority, thereby improving the overall teaching effect.

[0118] It should be noted that the time deviation index refers to the degree of deviation between the actual duration of the teaching content corresponding to each module in the second target set and the preset standard time; if the degree of deviation is greater, the corresponding module priority index is smaller, and the priority as the final teaching content is smaller. The reason is that a larger degree of deviation may indicate that there are problems with the design or implementation of the module, resulting in reduced teaching efficiency; therefore, an increase in the time deviation index will lead to a decrease in the priority index of the corresponding module, thereby reducing its priority in the final teaching content. This mechanism encourages teachers to grasp the teaching time more accurately, ensuring that each module can effectively convey knowledge within the scheduled time, thereby improving the overall teaching effect.

[0119] In one embodiment, obtaining the priority index of the corresponding module according to the use effect index, time deviation index and form type of each module in the second target set includes:

[0120] Obtain all form types contained in each module in the second target set, and obtain the usage count and preset interest weight value corresponding to each form type, and obtain the form diversity coefficient of each module. The calculation formula is: In the formula, TE is the form diversity index, e i is the number of times the i-th form type is used, ω i is the preset interest weight value corresponding to the i-th form type, and q is the total number of form types;

[0121] According to the use effect index, time deviation index and form diversity index of each module in the second target set, the priority selection index of the corresponding module is obtained.

[0122] It should be noted that the preset interest weight value is set by professionals based on the actual situation, and no specific limitation or elaboration is made; in addition, the form type can be, for example, video lectures, mini-program game interactions, etc. The specific form type depends on the actual situation and no limitation or elaboration is made.

[0123] It should be noted that the form types included in each module and the number of times each form type is used and the preset interest weight value can be recorded through the background data collection function of the client management system (LMS), or other methods can be used, which are not limited or elaborated in detail;

[0124] It should be noted that the form diversity index refers to the degree of form diversity of the teaching content corresponding to each module in the second goal set; if the degree of form diversity is greater, the corresponding module priority index is greater, and the priority as the final teaching content is greater. The reason is that the form diversity index reflects the richness of the teaching content in the design and implementation process. The diversity of forms can meet the learning styles and interests of different students, promote more active participation and higher learning motivation. When a module contains multiple form types, such as video lectures, mini-program game interactions, etc., students can access and understand knowledge in different ways, which not only improves the fun of learning, but also enhances the absorption and memory of knowledge; therefore, the greater the degree of form diversity, the higher the module priority index, and the teaching content obtains greater priority in the teaching process, thereby improving the overall teaching quality.

[0125] In one embodiment, obtaining the priority index of the corresponding module according to the usage effect index, time deviation index and form diversity index of each module in the second target set includes:

[0126] The use effect index, time deviation index and form diversity index are normalized, and the priority index of the corresponding module is obtained according to the normalized use effect index, time deviation index and form diversity index. The calculation formula is:

[0127]

[0128] Wherein, sx, zp, and mp are respectively the normalized usage effect index, time deviation index, and form diversity index; f1, f2, and f3 are respectively the preset proportional coefficients of sx, zp, and mp, and f1, f2, and f3 are all greater than 0.

[0129] It should be noted that f1, f2, and f3 are set by professionals according to actual conditions. Generally, the sum of f1, f2, and f3 is 1. For example, f1, f2, and f3 can be 0.27, 0.41, and 0.32, respectively, or other numbers without specific limitation. In addition, commonly used normalization processing methods include Min-Max normalization, Z-Score standardization, etc. The specific method is selected by professionals according to actual conditions, and no specific limitation or elaboration is made.

[0130] It should be noted that the teacher, on the client side, prioritizes the modules with the largest indexes in the second target focus of multiple content tags to combine and generate the final teaching content.

[0131] For example, assuming that the module with the largest preference index in the second target set of the "introduction" content label is A, the module with the largest preference index in the second target set of the "explain" content label is H, and the module with the largest preference index in the second target set of the "interaction" content label is U, then the teacher first selects the teaching content corresponding to module A as the introduction content, then selects the teaching content corresponding to module H for in-depth knowledge explanation, and finally selects the teaching content corresponding to module U to stimulate students' participation and interactivity. Such a combination not only ensures the continuity and logic of the teaching content, but also takes into account the connection between different modules, thus forming a complete and systematic teaching plan; in this way, teachers can achieve teaching goals more effectively, improve students' learning effects and classroom participation; at the same time, it reduces the time and energy of teachers in selection and arrangement, reduces the burden of teaching courseware preparation, and reduces the impact on lesson preparation efficiency.

[0132] Based on the same inventive concept, the embodiment of the present invention also provides a system for generating teaching content based on artificial intelligence. Figure 2 , Figure 2 A framework diagram of a device for generating teaching content based on artificial intelligence provided by an embodiment of the present invention, the system includes a client and a cloud server, the cloud server and the client are connected, wherein:

[0133] The client sends a teaching outline to the cloud server; the teaching outline includes a course tag and multiple content tags; the target content tag is any one of the multiple content tags, and the target content tag includes a type tag and a teaching time tag;

[0134] The cloud server is used to select modules that match the course label and the target content label in the database as the first target set, determine the number of text words, the number of pictures and the number of videos in each module in the first target set to obtain the complexity level of each module; and send the complexity level of each module in the first target set to the client;

[0135] The client is also used to receive the selection instruction and filter out the second target set that meets the selection instruction from the first target set;

[0136] The cloud server is also used to obtain the usage frequency, corresponding time and form type of each module in the second target set to obtain a priority selection index;

[0137] The client is also used to prioritize the modules with the largest indexes in the second target focus of multiple content tags to combine and generate final teaching content.

[0138] A system for generating teaching content based on artificial intelligence provided by an embodiment of the present invention makes it more convenient for teachers to make choices through the above-mentioned method, so that the final combined teaching content meets the needs of teaching courseware; at the same time, it reduces the time and energy of teachers in selection and arrangement, alleviates the burden of preparing teaching courseware, and reduces the impact on preparation efficiency.

[0139] Based on the same inventive concept, the embodiment of the present invention also provides a device for generating teaching content based on artificial intelligence. Figure 3 , Figure 3 A framework diagram of a device for generating teaching content based on artificial intelligence provided by an embodiment of the present invention, the device is applied in a cloud server, and the device includes:

[0140] The first screening module: receiving the teaching syllabus sent by the client; the teaching syllabus includes a course tag and multiple content tags; the target content tag is any one of the multiple content tags, and the target content tag includes a type tag and a teaching time tag;

[0141] Complexity level module: Modules matching the course label and target content label are selected from the database as the first target set, and the number of words, pictures and videos in each module in the first target set is determined to obtain the complexity level of each module;

[0142] The second screening module: sends the complexity level of each module in the first target set to the client, so that the client receives the selection instruction and screens out the second target set that meets the selection instruction from the first target set;

[0143] Content generation module: for each module in the second target set, obtain its usage frequency, corresponding time and form type to obtain a priority index; so that the client can combine the modules with the largest priority index in the second target set of multiple content tags to generate the final teaching content.

[0144] Based on the device for generating teaching content based on artificial intelligence provided by the embodiment of the present invention, through the above method, it is more convenient for teachers to make choices, so that the final combined teaching content meets the needs of teaching courseware; at the same time, it reduces the time and energy of teachers in selection and arrangement, alleviates the burden of preparing teaching courseware, and reduces the impact on preparation efficiency.

[0145] In another embodiment provided by the present invention, a computer-readable storage medium is also provided, in which a computer program is stored. When the computer program is executed by a processor, the steps of any one of the above-mentioned methods for generating teaching content based on artificial intelligence are implemented.

[0146] In another embodiment provided by the present invention, a computer program product including instructions is also provided, which, when executed on a computer, enables the computer to execute any one of the methods for generating teaching content based on artificial intelligence in the above embodiments.

[0147] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A method for generating teaching content based on artificial intelligence, characterized in that: The method is applied in a cloud server and includes the following steps: Receive a teaching outline sent by a client; the teaching outline includes a course tag and multiple content tags; the target content tag is any one of the multiple content tags, and the target content tag includes a type tag and a teaching time tag; Filtering the modules that match the course label and the target content label in the database as the first target set, determining the number of words, pictures and videos in each module in the first target set to obtain the complexity level of each module; Sending the complexity level of each module in the first target set to the client; so that the client receives the selection instruction and filters out the second target set that meets the selection instruction from the first target set; For each module in the second target set, its usage frequency, corresponding time and form type are obtained to obtain a priority index; so that the client can combine the modules with the largest priority index in the second target set of multiple content tags to generate the final teaching content.

2. The method for generating teaching content based on artificial intelligence according to claim 1, characterized in that: Determining the number of words, pictures and videos in each module of the first target set to obtain the complexity level of each teaching material includes: Obtain the number of different words and the total number of words in the text contained in each module, and divide the number of different words by the total number of words to obtain the lexical richness; Obtain the total number of sentences contained in each module, and divide the total number of words by the total number of sentences to obtain the average sentence length; The vocabulary richness and average sentence length of each module are normalized, and the text complexity of each module is obtained based on the normalized vocabulary richness and average sentence length; The complexity level of each module is obtained based on the text complexity, number of pictures and number of videos in each module.

3. The method for generating teaching content based on artificial intelligence according to claim 2, characterized in that: The complexity level of each module is obtained based on the text complexity, number of pictures and number of videos in each module, including: Obtain the total number of images contained in each module, and decompose each image into different frequency sub-bands by wavelet transform using Daubechies wavelet function; For each frequency sub-band, the information entropy of each sub-band is calculated, and the sum of the information entropy of each sub-band is used as the information entropy index of the corresponding image; Calculate the mean information entropy index of all images contained in each module as the image complexity of each module; Get the number of videos contained in each module, extract frames from each video to obtain the image set of the corresponding video, and calculate the mean information entropy index of the images in the video image set as the complexity index of the corresponding video, and take the mean complexity index of all videos contained in each module as the video complexity of each module; The complexity coefficient of each module is obtained according to the text complexity, image complexity and video complexity of each module, and the complexity level of each module is obtained according to the preset stage complexity coefficient threshold.

4. The method for generating teaching content based on artificial intelligence according to claim 3, characterized in that: The complexity coefficient of each module is obtained according to the text complexity, image complexity and video complexity of each module, and the complexity level of each module is obtained according to the preset stage complexity coefficient threshold, including: The text complexity, image complexity and video complexity of each module are normalized, and the complexity coefficient of each module is obtained according to the normalized text complexity, image complexity and video complexity. The calculation formula is: Ghc=b1×tgh+b2×yhl+b3×kmu Where Ghc is the complexity coefficient, tgh, yhl, kmu are the normalized text complexity, image complexity, and video complexity, b1, b2, and b3 are the preset proportional coefficients of tgh, yhl, and kmu, respectively, and b1, b2, and b3 are all greater than 0; Comparing the complexity coefficient of each module with a preset first complexity coefficient threshold and a preset second complexity coefficient threshold, and the preset first complexity coefficient threshold is less than the preset second complexity coefficient threshold; When the complexity coefficient is not greater than a preset first complexity coefficient threshold, the complexity level of the corresponding module is marked as simple; When the complexity coefficient is greater than a preset first complexity coefficient threshold but less than a preset second complexity coefficient threshold, the complexity level of the corresponding module is marked as medium; When the complexity coefficient is not less than a preset complexity coefficient second threshold, the complexity level of the corresponding module is marked as difficult.

5. The method for generating teaching content based on artificial intelligence according to claim 1, characterized in that: For each module in the second target set, obtaining its usage frequency, corresponding time and form type to obtain a preference index includes: The total number of times each module in the second target set is used, the frequency of use, and the average value of the use effect score are obtained, and normalized processing is performed to obtain the use effect index of each module. The calculation formula is: hd = ln (c1 × rt + c2 × ed + c3 × ws + 1), where rt, ed, and ws are the total number of times, frequency of use, and average value of the use effect score after normalization processing; c1, c2, and c3 are the preset proportional coefficients of rt, ed, and ws, respectively, and c1, c2, and c3 are all greater than 0; Obtain the actual duration of the teaching content corresponding to each module in the second target set, calculate the absolute difference between the actual duration and the preset standard time, and use the absolute difference as the time deviation index; According to the use effect index, time deviation index and form type of each module in the second target set, the priority selection index of the corresponding module is obtained.

6. The method for generating teaching content based on artificial intelligence according to claim 5, characterized in that: According to the usage effect index, time deviation index and form type of each module in the second target set, the priority index of the corresponding module is obtained, including: Obtain all form types contained in each module in the second target set, and obtain the usage count and preset interest weight value corresponding to each form type, and obtain the form diversity coefficient of each module. The calculation formula is: In the formula, TE is the form diversity index, e i is the number of times the i-th form type is used, ω i is the preset interest weight value corresponding to the i-th form type, and q is the total number of form types; According to the use effect index, time deviation index and form diversity index of each module in the second target set, the priority selection index of the corresponding module is obtained.

7. The method for generating teaching content based on artificial intelligence according to claim 6, characterized in that: According to the usage effect index, time deviation index and form diversity index of each module in the second target set, the priority selection index of the corresponding module is obtained, including: The use effect index, time deviation index and form diversity index are normalized, and the priority index of the corresponding module is obtained according to the normalized use effect index, time deviation index and form diversity index. The calculation formula is: Wherein, sx, zp, and mp are respectively the normalized usage effect index, time deviation index, and form diversity index; f1, f2, and f3 are respectively the preset proportional coefficients of sx, zp, and mp, and f1, f2, and f3 are all greater than 0.

8. A system for generating teaching content based on artificial intelligence, characterized in that: The system includes a client and a cloud server, wherein the cloud server and the client are connected, wherein: the client sends a teaching outline to the cloud server; the teaching outline includes a course tag and a plurality of content tags; the target content tag is any one of the plurality of content tags, and the target content tag includes a type tag and a teaching time tag; The cloud server is used to select modules that match the course label and the target content label in the database as the first target set, determine the number of text words, the number of pictures and the number of videos in each module in the first target set to obtain the complexity level of each module; and send the complexity level of each module in the first target set to the client; The client is also used to receive the selection instruction to filter out a second target set that meets the selection instruction from the first target set; The cloud server is also used to obtain the usage frequency, corresponding time and form type of each module in the second target set to obtain a priority selection index; The client is also used to prioritize the modules with the largest indexes from the second targets of the multiple content tags to combine and generate final teaching content.

9. A device for generating teaching content based on artificial intelligence, used to implement a method for generating teaching content based on artificial intelligence as described in any one of claims 1 to 7, characterized in that: The device is applied in a cloud server, and the device includes: The first screening module: receiving a teaching outline sent by a client; the teaching outline includes a course tag and a plurality of content tags; the target content tag is any one of the plurality of content tags, and the target content tag includes a type tag and a teaching time tag; Complexity level module: Filter the modules that match the course label and the target content label in the database as the first target set, determine the number of words, pictures and videos of each module in the first target set to obtain the complexity level of each module; The second screening module: sends the complexity level of each module in the first target set to the client, so that the client receives the selection instruction and screens out the second target set that meets the selection instruction from the first target set; Content generation module: for each module in the second target set, obtain its usage frequency, corresponding time and form type to obtain a priority index; so that the client can combine the modules with the largest priority index in the second target set of multiple content tags to generate the final teaching content.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed, the method according to any one of claims 1 to 7 is implemented.

Images