Teaching process abstract generation method and device, equipment and storage medium

Abstract

The invention is suitable for the field of artificial intelligence, and provides a teaching process abstract generation method and device, equipment and a storage medium, and the method comprises the steps: carrying out the voice recognition of a voice record of a teaching process, and obtaining a text record; generating cue words based on the character record and processing prompt information, wherein the processing prompt information is used for guiding an output result of a pre-trained target large model for the character record; and inputting the cue word into the target large model to obtain an execution abstract of the teaching process output by the target large model. According to the scheme, time and energy of the user can be saved, and resource utilization efficiency is effectively improved.

Description

Technical Field

[0001] The present application belongs to the field of artificial intelligence, and in particular, relates to a method, device, equipment and storage medium for generating a summary of a teaching process. Background Art

[0002] Reviewing the classroom situation after class can effectively improve the teaching quality. By recording the behavior of teachers and students in the classroom through recording and broadcasting products, audio, video, etc. can be obtained, which is convenient for classroom review after class.

[0003] At present, classroom recordings are mainly composed of audio and video, with a large amount of data. If viewers want to review the class based on the classroom recordings, it will take a lot of time and energy, and the resource utilization efficiency is low. Summary of the invention

[0004] The embodiments of the present application provide a method, device, equipment and storage medium for generating a summary of a teaching process, so as to solve the problem in the prior art that classroom review based on classroom records requires much time and effort and results in low resource utilization efficiency.

[0005] A first aspect of an embodiment of the present application provides a method for generating a summary of a teaching process, comprising:

[0006] Perform speech recognition on the speech record of the teaching process to obtain a text record;

[0007] Generate prompt words based on the text record and processing prompt information, wherein the processing prompt information is used to guide the output result of the pre-trained target large model for the text record;

[0008] The prompt words are input into the target macro model to obtain an executive summary of the teaching process output by the target macro model.

[0009] A second aspect of an embodiment of the present application provides a summary generation device for a teaching process, including:

[0010] The speech conversion module is used to perform speech recognition on the speech record of the teaching process to obtain a text record;

[0011] A prompt generation module, used to generate prompt words based on the text record and processing prompt information, wherein the processing prompt information is used to guide the output result of the pre-trained target large model for the text record;

[0012] A summary generation module is used to input the prompt words into the target macro model to obtain an executive summary of the teaching process output by the target macro model.

[0013] A third aspect of an embodiment of the present application provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method described in the first aspect when executing the computer program.

[0014] A fourth aspect of an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method described in the first aspect are implemented.

[0015] The fifth aspect of the present application provides a computer program product, including a computer-readable code, or a non-volatile computer-readable storage medium carrying a computer-readable code. When the computer-readable code is executed in an electronic device, the processor in the electronic device executes the steps in the method described in the first aspect above.

[0016] The above scheme in the embodiment of the present application obtains a text record by performing voice recognition on the voice record of the teaching process, generates prompt words based on the text record and processing prompt information, and then inputs it into the target big model to obtain the executive summary of the teaching process output by the target big model. The introduction of the big model realizes rapid data analysis and processing of the teaching process, and finally generates executive summary information of the teaching process, so that viewers can quickly understand the classroom teaching content and process by directly viewing the executive summary, saving user time and energy, and effectively improving resource utilization efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present application. Moreover, the same reference numerals are used throughout the drawings to represent the same components. In the drawings:

[0018] Figure 1 This is the process of the summary generation method of the teaching process in some embodiments of this application Figure 1 ;

[0019] Figure 2 is an example diagram of the system framework of some embodiments of the present application;

[0020] Figure 3 This is the process of the summary generation method of the teaching process in some embodiments of this application Figure 2 ;

[0021] Figure 4 is an example diagram of a network structure of some embodiments of the present application;

[0022] Figure 5is a structural diagram of a summary generating device of a teaching process in some embodiments of the present application;

[0023] Figure 6 It is a structural diagram of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0024] The following embodiments of the technical solution of the present application are described in detail in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present application, and are therefore only used as examples, and cannot be used to limit the scope of protection of the present application.

[0025] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by technicians in the technical field to which this application belongs; the terms used herein are only for the purpose of describing specific embodiments and are not intended to limit this application; the terms "including" and "having" in the specification and claims of this application and the above-mentioned figure descriptions and any variations thereof are intended to cover non-exclusive inclusions.

[0026] In the description of the embodiments of the present application, the technical terms "first", "second", etc. are only used to distinguish different objects, and cannot be understood as indicating or implying relative importance or implicitly indicating the number, specific order or primary and secondary relationship of the indicated technical features. In the description of the embodiments of the present application, the meaning of "multiple" is more than two, unless otherwise clearly and specifically defined.

[0027] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0028] In the description of the embodiments of the present application, the term "and / or" is only a description of the association relationship of the associated objects, indicating that there may be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0029] In order to allow teachers to review the classroom situation after class, reflect on their teaching, and improve the teaching resources such as courseware and teaching design needed for teaching, so as to improve teaching effectiveness, recording products can be used to record the behavior of teachers and students in the classroom, including recording audio, video, etc.

[0030] By watching classroom recordings, teachers can observe and analyze their own teaching behaviors and students' learning behaviors, and then improve the teaching resources such as courseware and teaching designs needed for teaching, so as to improve teaching results.

[0031] At present, after recording the entire course, classroom recordings will generate audio and video of up to 45 minutes, for example. As the number of recorded classes increases, the overall data volume will be larger. This means that when teachers review the classroom situation by watching the audio and video of classroom recordings, they will need to spend more time and energy, and the resource utilization efficiency is low, which puts a burden on the normal teaching activities of teachers, resulting in some situations where they can only choose to watch or not watch them, and a large number of valuable classroom recording resources are not fully utilized.

[0032] In an embodiment of the present application, a summary generation method for a teaching process is proposed, which obtains a text record by performing voice recognition on the voice record of the teaching process, generates prompt words based on the text record and processing prompt information, and then inputs it into a target large model to obtain an executive summary of the teaching process output by the target large model. The teaching process can be quickly analyzed and processed to generate executive summary information of the teaching process, so that viewers can quickly understand the classroom teaching content and process by directly viewing the executive summary, thereby effectively improving resource utilization efficiency.

[0033] In order to illustrate the technical solution described in this application, a specific embodiment is provided below for illustration.

[0034] Combination Figure 1 As shown, in some embodiments, a method for generating a summary of a teaching process is proposed, comprising:

[0035] Step 101, perform voice recognition on the voice record of the teaching process to obtain a text record.

[0036] Here, the voice recording refers to the voice data obtained by recording the actual teaching process.

[0037] Since the text transcript is converted from the voice transcript, the text transcript is also considered to be text data that records the real teaching process.

[0038] Step 102, generating prompt words based on the text record and processing prompt information.

[0039] Here, the generated prompt word includes both text record and processing prompt information. Among them, the processing instruction information can be information input by the user.

[0040] The processing prompt information is used to guide the pre-trained target large model to output results for the text transcript.

[0041] The output here is an executive summary of the teaching process corresponding to the text transcript.

[0042] Optionally, the processing prompt information can guide the pre-trained target large model to output an executive summary of the teaching process based on the text transcript, such as what language organization form, what data format, what descriptive content needs to be included, etc.

[0043] In one example, the voice record is obtained by recording the teaching process of the text "Runtu" in a Chinese class. Based on the voice record, voice recognition is performed on it to obtain a corresponding text record.

[0044] The information input by the user is, for example, "Please help me generate an executive summary of the teaching process for the text "Run Tu" based on the text transcript." Then, the information input by the user here is used as processing prompt information to guide the pre-trained target large model to generate an executive summary of the teaching process of the text "Run Tu" based on the text transcript.

[0045] Step 103, input the prompt words into the target macro model to obtain the executive summary of the teaching process output by the target macro model.

[0046] Here, the executive summary of the teaching process refers to the summary data in classroom teaching, which is led by the teacher and based on the students, and includes the execution process and relevant main contents of the entire teaching process.

[0047] In this process, voice recognition is performed on the voice record of the teaching process to obtain a text record, and prompt words are generated based on the text record and processing prompt information, which are then input into the target large model to obtain the executive summary of the teaching process output by the target large model. The introduction of the large model enables rapid data analysis and processing of the teaching process, and finally generates executive summary information of the teaching process, so that viewers can quickly understand the classroom teaching content and process by directly viewing the executive summary, saving users' time and energy and effectively improving resource utilization efficiency.

[0048] In an optional embodiment, the executive summary output by the target large model includes:

[0049] At least one teaching link information, teaching implementation description information of each teaching link, and teaching suggestion information for different teaching links.

[0050] Correspondingly, optionally, the executive summary of the teaching process outputted in step 103 includes:

[0051] The target big model identifies at least one teaching link information from the text record; obtains the teaching implementation description information extracted from the text record based on each teaching link information by the target big model; and obtains the teaching suggestion information output by the target big model based on different teaching link information.

[0052] Among them, the recognition of teaching link information can be that the big model recognizes the link based on key words detected from the text transcript, or the big model recognizes at least one teaching link information contained in the text transcript based on the characteristic information, time proportion and other standardized requirement information of each teaching link in the learned tutorial design specifications.

[0053] The teaching implementation description information is specifically information describing the implementation actions in the teaching process.

[0054] Optionally, the teaching process may include, for example: introducing a new lesson, teaching a new lesson, consolidating a new lesson, etc.

[0055] In some embodiments, the teaching implementation description information of the new course is introduced, for example, the teacher can arouse the students' interest in learning and curiosity through situation introduction, problem introduction, etc., to lay the foundation for the teaching of the new course.

[0056] In some embodiments, the teaching implementation description information in the new lesson segment, for example, the teacher uses a variety of methods such as explanation and demonstration to impart new knowledge and skills to help students master the core content of the class.

[0057] In some embodiments, the teaching implementation description information in the new lesson consolidation phase is used, for example, for teachers to consolidate the learned knowledge through classroom exercises, discussions, etc., to deepen students' understanding and memory of knowledge points.

[0058] This process specifically uses at least one teaching link information, the teaching implementation description information under each teaching link information and the teaching suggestion information output under different teaching link information as the content included in the executive summary of the teaching process, ensuring that effective summary information for actual teaching review needs is obtained through the large model, so that users can quickly understand the classroom teaching content and process, facilitate classroom review, and meet actual needs.

[0059] In one embodiment, the summary generation method of the teaching process is applied to the summary generation system of the teaching process. Figure 2 As shown, as an optional example, the summary generation system of the teaching process includes a speech recognition module, a large model fine-tuning module, a prompt word input module and an executive summary output module.

[0060] During the application process, the input information in the speech recognition module is the teaching speech recorded during the teaching process, and the text information is output after being processed by the speech recognition algorithm. In the large model fine-tuning module, the initial large model is fine-tuned mainly with text information such as teaching materials and teaching standards as the core, so as to facilitate the construction of the teaching large model (target large model). The main function of the prompt word input module is to receive the processing prompt information input by the user and generate prompt words based on the text record and processing prompt information, and then pass these prompt words to the fine-tuned large model. The executive summary output module is mainly used to feedback the executive summary of the teaching process generated by the fine-tuning large model. In order to facilitate user personalized modifications, the executive summary output module allows users to make personalized modifications to the generated executive summary.

[0061] In an alternative embodiment, in combination Figure 3 As shown in the figure, the pre-training process of the target large model includes:

[0062] Step 301: construct multiple prompt word samples as model input samples.

[0063] Step 302: construct multiple teaching process execution summary samples as model output samples.

[0064] The prompt word samples and the teaching process executive summary samples can be selected from the standard library as relevant data as samples, or can be configured and input into the system by experienced configuration personnel to realize sample construction, etc. It can be implemented according to the actual situation, and the specific construction method is not limited here.

[0065] Step 303: extract context information from the text information of various tutorial materials as a preset prefix.

[0066] Among them, the various course materials include, for example: teaching materials, teaching standards, teaching courseware, etc.

[0067] Based on these tutorial materials, context information is extracted, which can be specifically entity recognition, keyword extraction, dependency syntax analysis, etc., on the serialized tutorial text information to extract context information that is useful for model understanding and processing tasks. The extracted context information may include relevant concepts, definitions, principles, methods, application scenarios, and other information in the text information of the tutorial materials. This information helps the big model better understand and process the recorded data of the teaching process.

[0068] Contextual information extracted from serialized textbooks, teaching materials, teaching standards and other text information is added as a preset prefix to the input data of the large model to be fine-tuned. The purpose of this is to provide the model with additional context or background knowledge to help the model better understand and process the input prompt word samples and teaching process executive summary samples.

[0069] Step 304: Based on the model input samples, the model output samples and the preset prefix, the initial large model is trained to obtain the target large model.

[0070] Optionally, the initial large model is fine-tuned using the large model fine-tuning module in the summary generation system of the teaching process to obtain the target large model.

[0071] In an example, let's first introduce the principle of the big model:

[0072] Assuming a large model The input x is the context, and the output y is a tag sequence. Here x refers to the model input, and y refers to the model output. Based on the input context x, it can process and output the tag sequence y corresponding to x.

[0073] The activation vector at the i-th time step is in is the union of all activation layers at this time step, represented by z i And the function of past activation in its left context is calculated as:

[0074]

[0075] h i The final layer of is used to compute the distribution of the next token:

[0076] is a matrix used to convert The number of pairs mapped to the vocabulary.

[0077] In the large model fine-tuning module, the fine-tuning method combines serialized teaching materials, teaching courseware, and teaching standards and other text information, and uses the context information PREFIX extracted from them as a preset prefix. Based on this, in the embodiment of the present application, the model input used by the fine-tuning model is z = [PREFIX; x; y], and the PREFIX prefix index sequence is P idx , the prefix length is |P idx |.

[0078] Here, PREFIX refers to contextual information extracted from serialized textbooks, courseware, and teaching standards, which is added to the model input as a preset prefix. This provides additional context or background knowledge to the model to help it better understand and process input data x and y.

[0079] The prefix index activation is initialized as a free parameter with dimension |P idx |×dim(h i ) θTo this end, the activation vector h at the i-th time step is i It can be expressed as:

[0080]

[0081] The training objective function is:

[0082]

[0083] During model fine-tuning training, the set of trainable parameters may change, but the language model parameters is fixed, and the prefix parameter θ is the only trainable parameter.

[0084] To optimize stability, P θ Mainly through a large feedforward neural network (MLP θ ) to form a small matrix P θ ′ Parameterized, expressed as:

[0085] P θ [i,:]=MLP θ (P θ ′ [i,:]);

[0086] Among them, P θ and P θ ′ The number of rows is the same, that is, the prefix length is the same, but the number of columns is different.

[0087] Finally, fine-tuning the large model is equivalent to training P θ ′ and MLP θ After training, only the prefix P is retained. θ .

[0088] The above implementation method, by adding contextual information extracted from text information of various tutorial materials as a preset prefix to the training input of the model when constructing model fine-tuning training samples, can provide additional background knowledge for the model, so that the large model can effectively learn the teaching process requirements under the teaching scenario, so as to help the model better understand and handle related tasks, and ensure that the teaching process executive summary finally output by the trained large model is effective and reliable.

[0089] In an optional implementation, the sample of the teaching process execution summary is constructed based on the prompt word sample.

[0090] That is, the construction of the teaching process execution summary sample needs to be realized based on the already constructed prompt word sample. In this way, the constructed teaching process execution summary sample will establish a construction relationship with the prompt word sample.

[0091] In an optional implementation, step 304 includes:

[0092] Based on the model input samples, model output samples and preset prefixes, multiple data combinations are obtained as model training samples.

[0093] Among them, each data combination includes a target model input sample, a target model output sample having a construction relationship with the target model input sample, and a preset prefix; based on the model training sample, the initial large model is trained to obtain the target large model.

[0094] In this way, based on the construction relationship established between the teaching process execution summary sample and the prompt word sample, the association division between the sample data is realized, which can make the input and output in the constructed model training sample have effective correlation, ensure the effectiveness of the constructed model training sample, and ensure the final model fine-tuning training effect.

[0095] In an optional implementation, step 101 includes:

[0096] The speech record of the teaching process is input into the speech recognition model to obtain the text record output by the speech recognition model.

[0097] Among them, the speech recognition model includes a DCNN (Deep Convolutional Neural Network) model and a long short-term memory network LSTM (Long Short-Term Memory) model followed by the DCNN model.

[0098] The DCNN model includes a plurality of convolutional layers arranged in sequence, and each convolutional layer is connected to a normalization layer.

[0099] As an optional example, the process is implemented through a speech recognition module.

[0100] In the speech recognition module, the DCNN-LSTM-CTC model is specifically used to recognize teaching speech.

[0101] The model structure is as follows Figure 4 As shown in the figure, the model is based on a DCNN consisting of 20 convolutional layers, 20 normalization layers, and 10 maximum pooling layers, with ReLU as the activation function of the convolutional layer and CTC (Connectionist Temporal Classification) algorithm as the loss function of the model.

[0102] Among them, in order to enhance the relevance of the previous and next contexts in Chinese speech recognition, the DCNN-LSTM-CTC model adds a batch normalization layer after each convolution operation of DCNN and inherits LSTM after DCNN.

[0103] After DCNN-LSTM-CTC processes the input data, the pinyin sequence classified by the Softmax layer is the output result.

[0104] In DCNN, assuming that there are m neurons in the l-1th layer and n neurons in the lth layer, the linear coefficients w of the lth layer form an n×m matrix w l , the bias b of the lth layer forms an n×l vector b l , the output a of the l-1th layer forms an m×l vector a l-1 , the inactive linear output z of the lth layer forms an n×l vector z 1 , the output a of the lth layer forms an n×l vector a 1 . Using matrix method, the output of the lth layer is:

[0105] a 1 =σ(z 1 )=σ(w 1 a 1-1 +b 1 );

[0106] Since DCNN consists of 20 convolutional layers, 20 normalization layers and 10 maximum pooling layers, the specific process of spectrum map feature extraction is convolution-normalization-convolution-normalization-pooling.

[0107] LSTM adopts a recursive neural network structure and solves the problems of gradient vanishing and gradient exploding in traditional recursive neural networks by adding forget gates, input gates, and output gates. Figure 4 As shown, x t is the input information of the unit, f t 、i t and t Represent the outputs of the forget gate, input gate, and output gate respectively. f 、w i 、w o 、w c and b f , b i , b o , b c Represent the weight matrix and bias coefficient of the forget gate, input gate, output gate and unit respectively. The specific calculation process is as follows:

[0108] f t =σ(w f [x t ,ht-1 ])+b f ;

[0109] i t =σ(w i [x t ,h t-1 ])+b i ;

[0110] o t =σ(w o [x t ,h t-1 ])+b o ;

[0111] C t =f t ×C t-1 +i t ×tanh(x t ,h t-1 )+b c ;

[0112] h t =o t ×tanh(C t ).

[0113] By implementing the recognition and text conversion of speech transcripts through a speech recognition model with the above-mentioned model structure characteristics, the relevance of the previous and next contexts in Chinese speech recognition can be effectively enhanced, thereby improving the data processing effect.

[0114] In one embodiment, step 102 includes:

[0115] Display at least one prompt word template; obtain user input information; determine a target template matching the user input information from at least one prompt word template; parse the user input information based on the text structure of the target template to obtain processing prompt information.

[0116] As an optional example, the process is implemented through a prompt word input module.

[0117] The prompt word input module is mainly responsible for receiving the processing prompt information input by the user, and passing the processing prompt information together with the text record to the fine-tuned large model. In order to facilitate users to quickly input prompt words, prompt word templates are designed in the prompt word input module for users to refer to and quickly select.

[0118] Different prompt word targets may have their own text structure. In one example, the text structure of the prompt word template is, for example:

[0119] "Subject: Purpose + Detailed description of content + Special requirements".

[0120] Other forms of prompt word templates for text composition structures may also be set.

[0121] After obtaining the user's input information, it is necessary to determine the target prompt word template that matches the user's input information, and then parse the user's input information with the text structure of the target prompt word template to obtain processing prompt information, and then generate prompt words based on the processing prompt information and the text record.

[0122] This process improves the user experience by providing prompt word templates for users to refer to and quickly select, and assists in parsing the user input data with the help of the text structure in the prompt word template, thereby improving data processing efficiency.

[0123] As an optional embodiment, after step 103, the method further includes:

[0124] Obtain identification information of the text transcript and / or voice transcript; associate the identification information with the executive summary and output it.

[0125] Here, the executive summary of the teaching process obtained with the help of the big model will be associated with the text transcript and / or voice transcript and output, so that the text transcript and / or voice transcript can be annotated with the executive summary of the teaching process, thereby improving the user's utilization of the text transcript and / or voice transcript.

[0126] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0127] Based on the same inventive concept, the embodiment of the present application also provides a summary generation device for a teaching process. The summary generation device for a teaching process provided in the embodiment of the present application can implement each process of the embodiment of the summary generation method for the teaching process described above, and can achieve the same technical effect. Therefore, the specific limitations in the embodiments of the summary generation device for one or more teaching processes provided below can refer to the limitations of the summary generation method for the teaching process described above. To avoid repetition, they will not be repeated here.

[0128] In this embodiment, the computing side can be divided into functional modules according to the above method. For example, each function can be divided into functional modules, or two or more functions can be integrated into one processing module.

[0129] See also Figure 5 , Figure 5 This is a structural diagram of a summary generation device for a teaching process provided in an embodiment of the present application. For ease of explanation, only the parts related to the embodiment of the present application are shown.

[0130] The summary generation device 500 of the teaching process includes:

[0131] The speech conversion module 501 is used to perform speech recognition on the speech record of the teaching process to obtain a text record;

[0132] A prompt generating module 502, for generating prompt words based on the text record and processing prompt information, wherein the processing prompt information is used to guide the output result of the pre-trained target large model for the text record;

[0133] The summary generation module 503 is used to input the prompt words into the target macro model to obtain the executive summary of the teaching process output by the target macro model.

[0134] Optionally, the summary generation module 503 is specifically configured to:

[0135] Obtaining at least one teaching link information recognized by the target macro model from the text record;

[0136] Obtaining the teaching implementation description information extracted from the text record by the target macro model based on each of the teaching link information; and

[0137] The teaching suggestion information output by the target macro model based on the different teaching link information is obtained.

[0138] Optionally, the teaching process summary generation device 500 further includes:

[0139] Model fine-tuning module for:

[0140] Construct multiple prompt word samples as model input samples;

[0141] Construct multiple teaching process executive summary samples as model output samples;

[0142] Extracting context information from text information of various tutorial materials as a preset prefix;

[0143] Based on the model input samples, the model output samples and the preset prefix, the initial large model is trained to obtain the target large model.

[0144] Optionally, the teaching process execution summary sample is constructed based on the prompt word sample;

[0145] The model fine-tuning module is specifically used for:

[0146] Based on the model input sample, the model output sample and the preset prefix, a plurality of data combinations are obtained as model training samples; wherein each of the data combinations includes a target model input sample, a target model output sample having a construction relationship with the target model input sample and the preset prefix;

[0147] Based on the model training samples, the initial large model is trained to obtain the target large model.

[0148] Optionally, the voice conversion module 501 is specifically used for:

[0149] Inputting the speech record of the teaching process into a speech recognition model to obtain the text record output by the speech recognition model;

[0150] The speech recognition model includes a deep convolutional neural network (DCNN) model and a long short-term memory network (LSTM) model connected to the DCNN model.

[0151] The DCNN model includes a plurality of convolutional layers arranged in sequence, and each of the convolutional layers is connected to a normalization layer.

[0152] Optionally, the prompt generating module 502 is specifically configured to:

[0153] Display at least one prompt word template;

[0154] Get user input information;

[0155] Determining a target template matching the user input information from the at least one prompt word template;

[0156] Based on the text structure of the target template, the user input information is parsed to obtain the processing prompt information.

[0157] Optionally, the teaching process summary generation device 500 further includes:

[0158] Summary output module, specifically used for:

[0159] Obtaining identification information of the text transcript and / or the voice transcript;

[0160] The identification information is associated with the execution summary and outputted.

[0161] The above integrated modules can be implemented in the form of hardware. It should be noted that the division of modules in this embodiment is schematic and is only a logical function division. There may be other division methods in actual implementation.

[0162] It should be noted that the relevant contents of each step involved in the above method embodiment can all be referred to the functional description of the corresponding functional module, which will not be repeated here.

[0163] Each module in the summary generation device of the teaching process can be implemented in whole or in part by software, hardware and a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute the operations corresponding to each module.

[0164] In one embodiment, Figure 6 As shown, a computer device is provided. The computer device 6 of this embodiment includes: at least one processor 600 ( Figure 6 Only one is shown in the figure), a memory 601, and a computer program 602 stored in the memory 601 and executable on the at least one processor 600, wherein the processor 600 implements the steps of any of the above-mentioned method embodiments when executing the computer program 602.

[0165] The computer device 6 may be a computing device such as a desktop computer, a notebook, a palm computer, etc. The computer device 6 may include, but is not limited to, a processor 600 and a memory 601. Those skilled in the art will understand that Figure 6 It is only an example of the computer device 6 and does not constitute a limitation of the computer device 6. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the computer device may also include input and output devices, network access devices, buses, etc.

[0166] The processor 600 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0167] The memory 601 may be an internal storage unit of the computer device 6, such as a hard disk or memory of the computer device 6. The memory 601 may also be an external storage device of the computer device 6, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the computer device 6. Further, the memory 601 may also include both an internal storage unit and an external storage device of the computer device 6. The memory 601 is used to store the computer program and other programs and data required by the computer device. The memory 601 may also be used to temporarily store data that has been output or is to be output.

[0168] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0169] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0170] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0171] In the embodiments provided in the present application, it should be understood that the disclosed apparatus / computer equipment and methods can be implemented in other ways. For example, the apparatus / computer equipment embodiments described above are merely schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of the apparatus or unit, which can be electrical, mechanical or other forms.

[0172] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0173] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0174] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0175] The present application implements all or part of the processes in the above-mentioned embodiment methods, and may also be implemented through a computer program product. When the computer program product runs on a computer device, the computer device can implement the steps in the above-mentioned method embodiments when executing the computer program product.

[0176] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A summary generation method for a teaching process, characterized in that: include: Perform speech recognition on the speech record of the teaching process to obtain a text record; Generate prompt words based on the text record and processing prompt information, wherein the processing prompt information is used to guide the pre-trained target large model to output the text record; The prompt words are input into the target macro model to obtain an executive summary of the teaching process output by the target macro model.

2. The method according to claim 1, characterized in that The executive summary of the teaching process for obtaining the target large model output includes: Obtaining at least one teaching link information recognized by the target macro model from the text record; Obtaining the teaching implementation description information extracted from the text record by the target macro model based on each of the teaching link information; and The teaching suggestion information output by the target macro model based on the different teaching link information is obtained.

3. The method according to claim 1, characterized in that The pre-training process of the target large model includes: Construct multiple prompt word samples as model input samples; Construct multiple teaching process executive summary samples as model output samples; Extracting context information from text information of various tutorial materials as a preset prefix; Based on the model input samples, the model output samples and the preset prefix, the initial large model is trained to obtain the target large model.

4. The method according to claim 3, characterized in that The teaching process execution summary sample is constructed based on the prompt word sample; The training of the initial large model based on the model input sample, the model output sample and the preset prefix to obtain the target large model includes: Based on the model input sample, the model output sample and the preset prefix, a plurality of data combinations are obtained as model training samples; wherein each of the data combinations includes a target model input sample, a target model output sample having a construction relationship with the target model input sample and the preset prefix; Based on the model training samples, the initial large model is trained to obtain the target large model.

5. The method according to claim 1, characterized in that The method of performing speech recognition on the speech record of the teaching process to obtain a text record includes: Inputting the speech record of the teaching process into a speech recognition model to obtain the text record output by the speech recognition model; The speech recognition model includes a deep convolutional neural network (DCNN) model and a long short-term memory network (LSTM) model connected to the DCNN model. The DCNN model includes a plurality of convolutional layers arranged in sequence, and each of the convolutional layers is connected to a normalization layer.

6. The method according to claim 1, characterized in that The generating of prompt words based on the text record and processing prompt information includes: Display at least one prompt word template; Get user input information; Determining a target template matching the user input information from the at least one prompt word template; Based on the text structure of the target template, the user input information is parsed to obtain the processing prompt information.

7. The method according to any one of claims 1 to 6, characterized in that: After inputting the prompt words into the target macro model and obtaining the execution summary of the teaching process output by the target macro model, the method further includes: Obtaining identification information of the text transcript and / or the voice transcript; The identification information is associated with the execution summary and outputted.

8. A summary generation device for a teaching process, characterized in that: include: The speech conversion module is used to perform speech recognition on the speech record of the teaching process to obtain a text record; A prompt generation module, used to generate prompt words based on the text record and processing prompt information, wherein the processing prompt information is used to guide the output result of the pre-trained target large model for the text record; A summary generation module is used to input the prompt words into the target macro model to obtain an executive summary of the teaching process output by the target macro model.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images