Question and answer generation method, intelligent partner training method and related device

Abstract

The invention provides a question and answer generation method, an intelligent partner training method and a related device, and relates to the technical field of artificial intelligence. The question and answer generation method can comprise the following steps: receiving query information input by a user; wherein the query information is used for querying content related to partner training; according to the query information, matching a corresponding target corpus in target document content; wherein the target document content comprises required partner training content, and the target document content is stored in a knowledge base; and according to the query information and the target corpus, generating a partner training question and a question answer. The technical scheme provided by the invention is used for solving the problem of low efficiency of generation of questions and answers in intelligent partner training in the prior art.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, specifically to the field of intelligent training technology, and more particularly to a method for generating questions and answers, an intelligent training method and related devices. Background Art

[0002] With the development of society, users have higher and higher requirements for services. Therefore, improving service quality in the service industry has become a top priority for corporate development.

[0003] Among them, the improvement of service quality requires a lot of training and real-life scenario exercises, so it takes a lot of manpower and material resources to improve service quality. In order to improve this situation, smart training came into being. Smart training is a training tool that combines artificial intelligence technology. It interacts with users by simulating real scenarios to improve the learning efficiency and professional skills of individuals or corporate employees.

[0004] However, in related solutions, the training questions and answers in intelligent training are all set manually, which is time-consuming and labor-intensive. Summary of the invention

[0005] Based on the above-mentioned defects and shortcomings of the prior art, the present application proposes a method for generating questions and answers, an intelligent training method and related devices, which are used to solve the problem of low efficiency in generating questions and answers in intelligent training in the prior art.

[0006] According to a first aspect of the present application, a method for generating questions and answers is provided, the method comprising:

[0007] Receive query information input by a user; wherein the query information is used to query content related to sparring;

[0008] According to the query information, the corresponding target corpus is matched in the target document content; wherein the target document content includes the required sparring content and is stored in the knowledge base;

[0009] Generate training questions and answers based on the query information and the target corpus.

[0010] According to a second aspect of an embodiment of the present application, there is provided an intelligent training method, the method comprising:

[0011] Determine a target training question bank; wherein the target training question bank includes training questions and answers corresponding to the training questions, the training questions and the answers are generated in advance based on query information and target corpus matching the query information in target document content, the query information is used to query content related to training, the target document content includes the required training content, and is stored in the knowledge base;

[0012] Intelligent training is performed based on the training questions and answers in the training question bank.

[0013] According to a third aspect of the present application, a device for generating questions and answers is provided, the device comprising:

[0014] A first receiving module is used to receive query information input by a user; wherein the query information is used to query related to the training content;

[0015] A query module, used to match the corresponding target corpus in the target document content according to the query information; wherein the target document content includes the required sparring content and is stored in the knowledge base;

[0016] A generation module is used to generate training questions and answers based on the query information and the target corpus.

[0017] According to a fourth aspect of the embodiments of the present application, there is provided an intelligent training device, the device comprising:

[0018] A determination module is used to determine a target training question bank; wherein the target training question bank includes training questions and answers corresponding to the training questions, the training questions and the answers are generated in advance based on query information and a target corpus matching the query information in the target document content, the query information is used to query content related to training, the target document content includes the required training content, and is stored in the knowledge base;

[0019] The training module is used to perform intelligent training based on the training questions and answers in the training question bank.

[0020] According to a fifth aspect of the present application, there is provided an electronic device, comprising: a memory and a processor;

[0021] The memory is connected to the processor and is used to store programs;

[0022] The processor is used to implement the method for generating questions and answers as described in the first aspect, or to implement the intelligent training method as described in the second aspect, by running the program in the memory.

[0023] According to the sixth aspect of the present application, a storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the method for generating questions and answers as described in the first aspect is implemented, or the intelligent training method as described in the second aspect is implemented.

[0024] According to the seventh aspect of the present application, a computer program product or a computer program is provided, the computer program product includes the computer program, and when the processor of the computer device executes the computer program, the steps in the method for generating questions and answers as described in the first aspect are implemented, or the steps in the intelligent training method as described in the second aspect are implemented. Optionally, the computer program can be stored in a readable storage medium or cloud of the computer device; the processor of the computer device reads the computer program from the readable storage medium or cloud.

[0025] In the technical solution provided by the present application, the user can input query information into the knowledge base for the required training content, and the knowledge base can match the corresponding corpus according to the query information. Since the query information is equivalent to question-type information, and the matching corpus is equivalent to the answer-type information for the question, the corresponding training questions and answers can be generated after certain logical reasoning and summarization of the query information and the matching corpus. In this way, the user only needs to input the query information, and the system can automatically generate the training questions and answers based on the query information and the matched corpus, thereby improving the efficiency of generating questions and answers. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0027] Figure 1 A flowchart of a method for generating questions and answers provided in an embodiment of the present application;

[0028] Figure 2 A flowchart of an intelligent training method provided in an embodiment of the present application;

[0029] Figure 3 A block diagram of a device for generating questions and answers provided in an embodiment of the present application;

[0030] Figure 4 A block diagram of an intelligent training device provided in an embodiment of the present application;

[0031] Figure 5 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0032] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0033] Application Overview

[0034] With the rapid development of artificial intelligence technology, the intelligent training system, as a new type of training tool, has begun to be applied in many fields, such as sales, customer service, education, etc. The intelligent training system relies on artificial intelligence technologies such as machine learning, deep learning, natural language processing, etc. to achieve simulation of complex scenarios and behavior prediction. By simulating real work scenarios, it can provide users with an interactive learning environment to improve their specific skills and knowledge levels.

[0035] Although the intelligent training system has made some progress in simulating real conversations, the generation of questions and answers required for training is still an inefficient process in existing technologies. The generation of traditional questions and answers relies on manual editing, which is not only time-consuming and labor-intensive, but also difficult to quickly adapt to business changes and knowledge updates, affecting the widespread application and development of intelligent training systems in different fields.

[0036] In order to solve the aforementioned technical problems, the present application provides a technique for generating questions and answers. Users can input query information into the knowledge base for the required training content, and the knowledge base can match the corresponding corpus based on the query information. Since the query information is equivalent to question-type information, and the matching corpus is equivalent to the reply to the question, the corresponding training questions and answers can be generated after certain logical reasoning and summarization of the query information and the matching corpus. In this way, the user only needs to input the query information, and the system can automatically generate the training questions and answers based on the query information and the matched corpus, thereby improving the efficiency of generating questions and answers, quickly adapting to the needs of business changes and knowledge updates, and enabling the intelligent training system to be widely used and developed in different fields.

[0037] Exemplary Methods

[0038] The embodiment of the present application provides a method for generating questions and answers, and the execution subject may be an electronic device, such as a terminal device, a server, etc. The terminal device mentioned here may include but is not limited to at least one of the following: a mobile phone, a tablet computer, a vehicle-mounted terminal, a personal digital assistant device, etc.

[0039] The method is described in detail below through some embodiments. The following embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0040] like Figure 1 As shown, the method for generating questions and answers may include steps 101 to 103, which are specifically described as follows:

[0041] Step 101: Receive query information input by a user.

[0042] The query information described here is used to query content related to coaching. For example, for car sales coaching, the query information can be used to query content related to car sales training. For example, the query information can be what performance parameters of car A include, what are the characteristics of car A, etc. It can be seen that the query information described here can be understood as question-type information.

[0043] The query described in the embodiment of the present application refers to querying a knowledge base. The knowledge base described here can be an enterprise knowledge base (i.e., a knowledge base for storing internal knowledge assets of an enterprise) or a public knowledge base (i.e., a knowledge base open to the public, such as digital resources of a public library).

[0044] In an embodiment of the present application, the information required for training can be stored in the knowledge base in advance, for example, the documents related to automobile sales (such as vehicle manuals) can be stored in the knowledge base in the form of electronic information to facilitate the required inquiries.

[0045] Step 102: According to the query information, the corresponding target corpus is matched in the target document content.

[0046] The target document content described here is stored in the knowledge base in the form of electronic information, which is related to the training and can also be understood as the file information used to generate the answer. The target document content includes the required training content.

[0047] Since the knowledge base stores a lot of information, and there is a lot of information that is not related to the training, in order to improve the query efficiency, the query range can be pre-specified. For example, you can pre-set which documents to query. For example, for automobile sales training, you can specify to query in documents such as vehicle manuals, vehicle market analysis reports, and vehicle technical manuals, thereby narrowing the query range and improving query efficiency. At the same time, it can also improve the accuracy of the query results and reduce redundant information. In the embodiment of the present application, the target document content is the specified query range.

[0048] The target corpus mentioned here refers to the original content in the target document content that matches the query information, which can be understood as answer-type information. For example, if the query information is "the advantages of electric car A", then the target corpus is the original document content that matches the advantages of electric car A found in the target document content. For example, the original document content that matches the advantages of electric car A found in the vehicle market analysis report is "electric car sales" and "second in sales in December", and the original text content that matches the advantages of electric car A found in the vehicle technical manual is "800 kilometers of cruising range", "voice-controlled windows", "the vehicle has a lane keeping assist function", etc. It should be noted that this is only an example, and it is not a specific limitation on the embodiments of the present application.

[0049] Step 103: Generate training questions and answers based on the query information and the target corpus.

[0050] Since the query information is equivalent to question-type information and the target corpus is equivalent to answer-type information for the question, the corresponding practice questions and answers can be generated after certain logical reasoning and summarization of the query information and the target corpus.

[0051] Continuing with the above example, if the query information is "the advantages of electric car A", then after logical reasoning and summarizing, the corresponding training question can be obtained as "What are the advantages of electric car A". The target corpus includes: "electric car sales", "second in sales in December", "800 kilometers of cruising range", "voice-controlled windows", "the vehicle has lane keeping assist function", etc. Then, after logical reasoning and summarizing, the answer to the question corresponding to the above training question can be "First, electric car A has good sales, ranking second in electric car sales in December; secondly, it has a long cruising range of 800 kilometers, and supports functions such as window voice control and lane keeping assist". It should be noted that this is only an example, and it is not a specific limitation of the embodiments of the present application.

[0052] Among them, before generating training questions and answers based on query information and target corpus, labels can be added to the query information and target corpus respectively, such as adding query input labels to the query information and adding query output labels to the target corpus, so as to distinguish between the two, thereby more accurately generating corresponding training questions based on the query information, and more accurately generating corresponding answers based on the target corpus.

[0053] In the embodiment of the present application, the user only needs to input the query information, and the system can automatically generate the training questions and answers based on the query information and the matched corpus, thereby improving the generation efficiency of the questions and answers. In addition, the training questions generated by this method are determined by the query information input by the user, that is, the training questions are determined by the user, so that the training content can be better grasped. And the questions and answers generated by this method, because the accuracy and range of machine query are far better than manual query, therefore, the obtained answers to the questions are more accurate and comprehensive, which can help students better understand the knowledge and skills they need to master, and help improve the training effect.

[0054] In some optional embodiments, since the automatically generated training questions and answers may deviate from the fluency and logic of the sentences, or deviate from the user's expectations (such as in terms of comprehensiveness and conciseness), after generating the training questions and answers, the method may further include:

[0055] The generated training questions and answers are displayed, and the training questions and answers are set to editable mode.

[0056] In this way, users can check whether the automatically generated training questions and answers meet expectations, whether the sentences are fluent, whether the logic is accurate, etc. Setting the two to editable mode allows users to adjust and modify the training questions and answers, thereby achieving the effect of correcting the questions and answers.

[0057] In some optional embodiments, a large language model may automatically generate training questions and answers based on query information and target corpus, as described below.

[0058] Step 103: Generate training questions and answers based on the query information and the target corpus, which may include:

[0059] The query information and target corpus are input into the large language model, and the large language model is used to generate practice questions and answers.

[0060] The large language model has powerful natural language processing capabilities and can understand and generate complex language expressions. Therefore, in the embodiment of the present application, the query information and the target corpus can be input into the large language model, and the large language model generates training questions and answers after performing certain logical reasoning and summarization on the query information and the target corpus. When the large language model is connected to the network, further queries can be performed based on the input query information, and answers to questions can be generated based on its own query results and the target corpus, so that the answers to the questions are more accurate and complete.

[0061] Among them, label information is added to the query information and target corpus input into the large language model. For example, the query information is added with a label of the query input, and the target corpus is added with a label of the query output, so that the large language model can distinguish the two based on the labels, thereby more accurately generating corresponding sparring questions based on the query information and corresponding question answers based on the target corpus.

[0062] Optionally, the large language model can also combine context information to generate practice questions and answers.

[0063] The context information described here may include but is not limited to at least one of the following: historically input query information, historically input target corpus, historically output training questions, historically output answers to questions, etc. For example, the query information currently input is: What are the advantages of electric car A compared to electric car B, the historically output training question a is: What are the advantages of electric car A, and the historically output training question b is: What are the advantages of electric car B. At this time, the large language model can combine the answers to the training questions a and b and the target corpus matching the current query information to generate the answer to this training question, which can make the answer to the question more accurate.

[0064] Optionally, for the function of generating practice questions and answers, the large language model can be trained in advance. For example, different query information and matching corpora are used as training data to train the large language model, and corresponding true value data is set at the same time, that is, the practice questions and answers expected by the user. By comparing the model output results with the true value data, the model parameters are continuously adjusted until the error between the model output results and the true value data is reduced to within the target range.

[0065] Optionally, the type of the large language model described here can be selected according to actual needs, for example, it can be a large language model based on the Transformer mechanism, a codec large language model, a universal large language model or a multimodal large language model, etc. Preferably, the large language model described here can be a multimodal large language model, so that the user can input query information of different data types (such as text, image, audio, etc.), thereby generating different types of accompanying training questions, such as text questions, picture questions, text and picture combination questions, sound questions, etc., to improve the richness of the accompanying training questions.

[0066] In some optional embodiments, the generated training questions and answers or the manually corrected training questions and answers can be added to the training question bank for subsequent training processes. This event can occur during the question bank establishment process, that is, before the question bank is open for use; it can also occur during the question bank maintenance process, that is, after the question bank is open for use, the process of updating the training content in the question bank.

[0067] Optionally, in the embodiment of the present application, a training question bank can be set for different positions, and the corresponding training question bank includes training questions and answers related to the training of the position, so as to provide personalized training services for trainees. For example, for a car sales position, a car sales training question bank is set, and questions and answers about car sales training are stored in the training question bank; for another example, for a car after-sales position, a car after-sales training question bank is set, and questions and answers about car after-sales training are stored in the training question bank.

[0068] On this basis, when the query information is used to query document content related to the target position training, after generating the training questions and answers through step 103, the method can also include: storing the training questions and answers in the training question bank of the target position.

[0069] The target position described here can be any position supported by the intelligent training function. The training question bank of the target position described here is pre-established. For a training system, multiple training question banks can be included.

[0070] Optionally, since different students need different training contents, in order to achieve personalized training, the student login information can be pre-associated with the training question bank of the target position, so that after the student logs in, the training system can directly provide the student with the training question bank of the target position, so that the student can perform the training process based on the training question bank of the target position, eliminating the operation of the student selecting the question bank by himself. The specific implementation method can be described as follows:

[0071] Receive the login information of at least one student, select a training question bank for a target position from multiple training question banks; and establish an association relationship between the login information of the at least one student and the training question bank for the target position.

[0072] The multiple training question banks mentioned here refer to training question banks for different positions.

[0073] In the embodiment of the present application, a setting interface may be provided, in which the student's login information may be input, and a drop-down list of accompanying training question banks may be provided in the setting interface, through which the desired accompanying training question bank may be selected. A confirmation button may also be provided in the setting interface, and after the confirmation button is triggered, an association relationship between the input student login information and the selected accompanying training question bank may be established.

[0074] In some optional embodiments, after step 101: generating a training question and a question answer, the method for generating the question and the answer may further include:

[0075] Receive the scoring criteria set for the training questions; and establish a correlation between the scoring criteria and the training questions.

[0076] After the practice questions and answers are generated, the administrator can set the scoring criteria for each question based on the question type to facilitate a more accurate and fair evaluation of the students' performance.

[0077] Among them, the question types can be divided into objective questions and subjective questions. Objective questions refer to questions with clear and fixed answers, which can usually be presented in the form of multiple choice, fill-in-the-blank or judgment. Subjective questions refer to questions that require candidates to answer based on their own understanding, analysis and expression abilities. The answers are not unique and more emphasis is placed on the candidates' personal insights and creativity.

[0078] For objective questions, the scoring criteria can be: determine whether the answer entered by the student is the same as the pre-set answer to the question. If they are the same, then score; if they are different, then no score. Scoring criteria can be set uniformly for objective questions. For example, in the scoring criteria setting interface, select all objective questions in the current field or position, and set the scoring criteria uniformly, thereby eliminating repeated and tedious operations. Among them, the answers to objective questions can be understood as standard answers.

[0079] For subjective questions, at least one keyword can be set according to the answer to the question, and the corresponding scoring criteria can be: in the answer entered by the student, keyword matching is performed, and the score is determined according to the number and similarity of the matched keywords. For example, the more matched keywords, the higher the score; the higher the similarity between the matched keywords and the preset keywords, the higher the score; and the scores can be respectively based on the number of matched keywords and the similarity of the matched keywords, and then the comprehensive score is obtained according to the preset weight values ​​of the number and similarity, as well as the respective scores. Among them, the weight values ​​of quantity and similarity can be set according to actual needs, and the embodiments of the present application do not specifically limit this. Among them, the answers to subjective questions can be understood as reference answers. The aforementioned scoring criteria refer to scoring based on accuracy.

[0080] For subjective questions, in addition to scoring based on the number and similarity of matched keywords, you can also score from other angles, such as completeness, logic, emotion, standardization (whether sensitive words appear, etc., which is a deduction item), etc., and then combine the scores of multiple angles and the preset weight values ​​of multiple angles to get the final score.

[0081] The scoring criteria for subjective questions can be set according to different training needs. For example, for training with pronunciation requirements, scoring can be based on accuracy, fluency, and other aspects.

[0082] Optionally, the keywords in the subjective questions may be manually set based on the training questions and the answers to the questions, or may be determined by the large language model based on the training questions and the answers to the questions. For the latter, the process of determining the keywords in the scoring criteria may include:

[0083] The generated training questions and answers are input into the large language model, and the keywords corresponding to the training questions are generated through the large language model and the task instruction information in the large language model.

[0084] The task indication information described here is also called prompt, which is used to indicate the determination requirements of keywords. The task indication information can clearly inform the model of the tasks to be performed, and can improve the accuracy of the model output results.

[0085] For example, the task instruction information may include the following content:

[0086] Please combine the practice questions and extract key words from the answers as scoring points. Specific requirements include:

[0087] n1. The number of characters in a single keyword shall not exceed 8

[0088] n2. Output in array format, strictly follow the following format: ["keyword 1", "keyword 2"], no other content

[0089] n3. Keep it as simple as possible and only keep the core content

[0090] n4. The number does not exceed 5

[0091] n5. Don’t use pure nouns, try to use descriptive words instead.

[0092] For example, the training topic is: What are the methods for handling customer complaints to improve customer satisfaction?

[0093] The answer to the question is:

[0094] Ways to handle customer complaints to improve customer satisfaction include:

[0095] 1. Timely response: Respond to customer complaints quickly to make customers feel valued.

[0096] 2. Listen carefully: Listen patiently to the customer’s demands and understand the problem.

[0097] 3. Sincere Apology: Express sincere apologies for the customer’s problem and eliminate the customer’s dissatisfaction.

[0098] 4. Actively solve problems: Provide reasonable solutions to customer problems and ensure that the problems are properly handled.

[0099] 5. Follow-up feedback: After the problem is solved, actively communicate with the customer to understand customer satisfaction and continuously improve service quality.

[0100] The output of the large language model can be:

[0101] ["respond promptly", "listen carefully", "apology sincerely", "proactive solution", "follow up feedback"]

[0102] The output results of "timely response, careful listening, sincere apology, active resolution, and follow-up feedback" can be used as keywords in the scoring criteria for the above-mentioned practice questions.

[0103] It is understandable that the output results of the model can be displayed for the administrator to view, and the displayed results can be set to an editable mode so that the administrator can correct the output results.

[0104] In the embodiment of the present application, the large language model can combine the training questions and task instruction information to automatically determine the keywords in the answers to the questions, thereby improving the efficiency of keyword positioning.

[0105] In some optional embodiments, before step 101: receiving query information input by a user, the method for generating questions and answers may further include:

[0106] Step A1: Obtain document content in the target document.

[0107] In an embodiment of the present application, a user can upload documents required for training to an electronic device. These documents can be various forms of text materials such as operating manuals and documents. Accordingly, the electronic device can receive document materials related to training uploaded by the user. Afterwards, the electronic device can process the uploaded document materials to obtain the document content therein. For example, when the target document is an image document (such as a PDF format document, a document photo, etc.), the document content in the target document can be identified by optical character recognition (OCR) technology to obtain the document content.

[0108] OCR technology refers to the process in which an electronic device (such as a scanner or digital camera) examines characters printed on paper, determines their shapes by detecting dark and light patterns, and then uses character recognition methods to translate the shapes into computer text; that is, for printed characters, optical methods are used to convert the text in paper documents into black and white dot matrix image files, and the text in the image is converted into text format through recognition software for further editing and processing by word processing software. OCR technology can convert the content of paper documents into editable electronic text, achieving the purpose of convenient storage, retrieval and sharing.

[0109] Prior to this, electronic devices can pre-process uploaded documents, such as removing irrelevant information, converting them into a file format that can be processed by OCR technology, etc., to ensure the accuracy and efficiency of subsequent processing.

[0110] Step A2: Segment the obtained document content according to preset rules to obtain document content blocks.

[0111] The preset rule described here is to segment according to the number of characters, or the preset rule is to segment according to paragraphs.

[0112] To facilitate processing and retrieval, electronic devices can divide long document content into several document content chunks according to certain rules (such as character length, paragraph structure, etc.).

[0113] Step A3: Convert each document content block into a vector, and store the converted vector in a vector database in the knowledge base.

[0114] Next, electronic devices can use vectorization technology (Embedding) to convert these document content chunks into vector data. Among them, vectorization technology is a technology that converts text into numerical vector representation, which can capture the semantic information of the text so that computers can better understand and process the text. Commonly used vectorization methods can include bag-of-words model, TF-IDF, Word2Vec, BERT, etc.

[0115] The vectorized vector data will be stored in a dedicated vector database. A vector database is a special database used to store and process vector data, and can provide efficient similarity search and retrieval functions.

[0116] When a user enters a query, the electronic device also vectorizes the query text and performs a similarity search in the vector database to find vectors similar to the query.

[0117] The electronic device can package the query information input by the user and the similarity corpus retrieved from the vector database and submit them to the large language model, which will generate practice questions and answers.

[0118] Among them, steps A1 to A3 can be specifically executed by a knowledge base engine.

[0119] In summary, in the method for generating questions and answers provided in the embodiment of the present application, the user can input query information to the knowledge base for the required accompaniment content, and the knowledge base can match the corresponding corpus according to the query information. Since the query information is equivalent to the question type information, and the matching corpus is equivalent to the answer type information for the question, the corresponding accompaniment questions and answers can be generated after certain logical reasoning and summarization of the query information and the matching corpus. In this way, the user only needs to input the query information, and the system can automatically generate the accompaniment questions and answers based on the query information and the matched corpus, thereby improving the generation efficiency of the questions and answers. In addition, the accompaniment questions generated by the method are determined by the query information input by the user, that is, the accompaniment questions are determined by the user, so as to better grasp the accompaniment content. And the questions and answers generated by the method, because the accuracy and range of the machine query are far better than the manual query, therefore, the obtained questions and answers are more accurate and comprehensive, which can help students better understand the knowledge and skills required to master, and help to improve the accompaniment effect.

[0120] The embodiment of the present application also provides a smart training method, the execution subject can be an electronic device, such as a terminal device, and the terminal device described here can include but is not limited to at least one of the following: a mobile phone, a tablet computer, a vehicle terminal, a personal digital assistant device, etc. The terminal device supports the smart training function. The smart training function can be called out through a mini program or a dedicated smart training APP.

[0121] The method is described in detail below through some embodiments. The following embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0122] like Figure 2 As shown, the intelligent training method may include steps 201 to 203, which are specifically described as follows:

[0123] Step 201: Determine the target training question bank.

[0124] The target training question bank mentioned here refers to the training question bank required by the trainees using the training system. For example, when the trainees are in the position of car sales, the training question bank they need is the car sales training question bank; for another example, when the trainees are in the position of car after-sales, the training question bank they need is the car after-sales training question bank. Setting different training question banks for different training needs helps to achieve targeted training.

[0125] Among them, the target training question bank includes training questions and answers corresponding to the training questions. The training questions and answers in the target training question bank are pre-generated based on the query information and the target corpus matching the query information in the target document content.

[0126] The query information described here is used to query content related to coaching. For example, for car sales coaching, the query information can be used to query content related to car sales training. For example, the query information can be what performance parameters of car A include, what are the characteristics of car A, etc. It can be seen that the query information described here can be understood as question-type information.

[0127] The query described in the embodiment of the present application refers to querying a knowledge base. The knowledge base described here can be an enterprise knowledge base (i.e., a knowledge base for storing internal knowledge assets of an enterprise) or a public knowledge base (i.e., a knowledge base open to the public, such as digital resources of a public library).

[0128] In an embodiment of the present application, the information required for training can be stored in the knowledge base in advance, for example, the documents related to automobile sales (such as vehicle manuals) can be stored in the knowledge base in the form of electronic information to facilitate the required inquiries.

[0129] The target document content described here is stored in the knowledge base in the form of electronic information, which is related to the training and can also be understood as the file information used to generate the answer. The target document content includes the required training content.

[0130] Since the knowledge base stores a lot of information, and there is a lot of information that is not related to the training, in order to improve the query efficiency, the query scope can be pre-specified. For example, you can pre-set which documents to query. For example, for automobile sales training, you can specify to query in documents such as vehicle manuals, vehicle market analysis reports, and vehicle technical manuals, thereby narrowing the query scope and improving query efficiency. At the same time, it can also improve the accuracy of the query results and reduce redundant information. In the embodiment of the present application, the target document content is the specified query scope.

[0131] The target corpus mentioned here refers to the original content in the target document that matches the query information, which can be understood as answer-type information.

[0132] Since the query information is equivalent to question-type information and the target corpus is equivalent to answer-type information for the question, the corresponding practice questions and answers can be generated after certain logical reasoning and summarization of the query information and the target corpus.

[0133] Among them, before generating training questions and answers based on query information and target corpus, labels can be added to the query information and target corpus respectively, such as adding a query input label to the query information and adding a query output label to the target corpus, so that the device or system used to implement the method can distinguish between the two, thereby more accurately generating corresponding training questions based on the query information, and more accurately generating corresponding answers to the questions based on the target corpus.

[0134] In the embodiment of the present application, the user only needs to input the query information, and the system can automatically generate practice questions and answers based on the query information and the matched corpus, thereby improving the efficiency of generating questions and answers.

[0135] Optionally, the large language model can automatically generate training questions and answers based on the query information and the target corpus. The details can be seen in the previous embodiment, which will not be repeated here.

[0136] Step 202: Perform intelligent training based on the training questions and answers in the target training question bank.

[0137] Since the training questions in the training question bank are determined by the query information input by the user, that is, the training questions are determined by the user, which can better grasp the training content. The answers to the questions generated by this method are more accurate and comprehensive, because the accuracy and scope of machine query are far better than those of manual query, which can help students better understand the knowledge and skills they need to master, and help improve the training effect.

[0138] In some optional embodiments, step 201: determining a target training question bank may include:

[0139] Obtain login information for the training system; determine a target training question bank that matches the login information.

[0140] Therein, a correlation is pre-established between the login information and the target training question bank.

[0141] Since different students require different training content, in order to achieve personalized training, the student login information can be pre-associated with the training question bank of the target position. In this way, after the student logs in, the training question bank of the target position can be directly provided to the student, so that the student can perform the training process based on the training question bank of the target position, eliminating the need for the student to select the question bank by himself.

[0142] In some optional embodiments, the student's response to the training question needs to be scored based on a preset scoring standard.

[0143] For objective questions, the scoring criteria may be: determine whether the answer entered by the student is the same as the pre-set answer to the question; if so, score; if not, no score.

[0144] For subjective questions, at least one keyword can be set according to the answer to the question, and the corresponding scoring criteria can be: in the answer entered by the student, the keyword matching is performed, and the score is determined according to the number and similarity of the matched keywords. For example, the more the number of matched keywords, the higher the score; the higher the similarity between the matched keywords and the preset keywords, the higher the score; and the scores can be respectively based on the number of matched keywords and the similarity of the matched keywords, and then the weight values ​​preset for the number and similarity, as well as the respective scores, are used to obtain a comprehensive score. Among them, the weight values ​​of the number and similarity can be set according to actual needs, and the embodiments of the present application do not specifically limit this.

[0145] When scoring subjective questions, the tutoring system can collect audio data of students' answers and upload it to the tutoring system backend in real time. The system backend can input the audio file into the speech transcription engine to convert the student's voice answer into text. After that, the word segmentation engine is called to perform word segmentation on the transcription record; then, the similarity matching engine is called for the obtained words to perform similarity matching with the pre-set scoring keywords. For each pre-set scoring keyword, the word with the highest similarity is used as the matched keyword.

[0146] For subjective questions, in addition to scoring the accuracy based on the number and similarity of matched keywords, you can also score from other angles, such as completeness, logic, emotion, and standardization (whether sensitive words appear, etc.), and then combine the scores of multiple angles and the preset weight values ​​of multiple angles to get the final score.

[0147] The scoring criteria for subjective questions can be set according to different training needs. For example, for training with pronunciation requirements, scoring can also be conducted from the perspectives of accuracy, fluency, etc.

[0148] In some optional embodiments, step 201: performing intelligent training based on the training questions in the training question bank may include:

[0149] Step B1: upon receiving an instruction that the answer to the target training question is complete, determining whether the student's answer to the target training question is complete based on the target training question.

[0150] Step B2: When it is determined that the reply content is incomplete, output a prompt message.

[0151] The target training questions described here are subjective questions in the target training question bank.

[0152] The prompts described here are used to prompt students to provide complete responses.

[0153] In the embodiment of the present application, for subjective questions, when the student inputs a reply completion instruction to the training system, such as outputting "reply completed" by voice, the training system can determine whether the student's reply is complete. If it is incomplete, the student is prompted to complete the reply to improve the training effect.

[0154] For example, you can check whether the student's answer covers all the key points required by the question to determine whether the answer is complete. Suppose the training question is: Please explain the service provided to customers with purchasing intentions from three aspects. If based on the student's answer, it is detected that the student only explained from one or two aspects, it is obvious that the answer is incomplete, so a prompt message can be output at this time, asking the student to provide a complete answer. For example, the prompt message can be "This question needs to be explained from three aspects, and it has been explained from two aspects. It needs to be explained from another aspect."

[0155] Optionally, in the embodiment of the present application, the large language model can be used to determine whether the student's answer to the target training question is complete, and the specific implementation method is as follows:

[0156] The target training questions and the students' responses are input into the large language model. The students' responses are analyzed through the large language model and the task instruction information in the large language model to determine whether the responses are complete.

[0157] The task instruction information described here is used to instruct the large language model to judge the completeness of the answer content. For example, the task instruction information can be "1. Each question needs to be asked separately, and the next one needs to be asked after the previous one is answered; 2. If my answer is not clear, you need to ask follow-up questions to ensure that the question is answered; 3. Asking questions means directly selecting a question from the target training question bank, and asking follow-up questions means asking follow-up questions about the content that is not fully answered in the question."

[0158] In an embodiment of the present application, the large language model can use a natural language processing algorithm to semantically understand and evaluate the student's response content, and use the reference answer (i.e., the pre-set question answer) as a basis to analyze the completeness of the response content, thereby determining whether the response content is complete.

[0159] It should be noted that when the scoring criteria include scoring the completeness of the response content, the processing method is the same as the aforementioned scheme type, with an additional scoring step compared to the aforementioned scheme.

[0160] Optionally, the large language model can also determine the output of the next training question based on the historical training questions. For example, the training question a outputted last time was: What are the advantages of electric car A? The training question b outputted last time was: What are the advantages of electric car B? Based on the two historical training questions, the large language model can output the training question: What are the advantages of electric car A compared to electric car B? as the current training question.

[0161] In some optional embodiments, in order to enhance the training effect and improve the user experience, the training system in the embodiment of the present application also integrates virtual human generation technology. By inputting the selected training topic into the virtual human generation engine, an interactive virtual character video can be created. These virtual characters can communicate with the trainees in a natural and smooth way, providing the latter with a learning experience that is closer to the actual work scene. This method not only improves the fun and participation of the training, but also effectively improves the efficiency and quality of information transmission.

[0162] Accordingly, step 201: performing intelligent training based on the training questions and answers in the training question bank may include:

[0163] Output target practice questions through preset virtual characters.

[0164] In the embodiment of the present application, a variety of virtual characters and different voices can be provided for students to choose from to meet the personalized needs of students.

[0165] After the practice is completed, the system can store the audio data and answers of the practice persistently, and generate a practice report to facilitate subsequent review.

[0166] In summary, in the technical solution provided by the embodiment of the present application, since the training questions in the training question bank are determined by the query information input by the user, that is, the training questions are determined by the user, so that the training content can be better grasped. And the answers to the questions generated by this method, because the accuracy and range of machine query are far better than human query, therefore, the answers to the questions obtained are more accurate and comprehensive, which can help students better understand the knowledge and skills required to master, and help improve the training effect. In addition, for subjective questions, when the student inputs the reply completion instruction to the training system, the training system can determine whether the student's reply content is complete, and in the case of incompleteness, prompt the student to answer completely, so as to improve the training effect. Further, the embodiment of the present application can also provide the latter with a learning experience that is closer to the actual work scene through virtual character generation technology, so as to improve the fun and participation of training.

[0167] Exemplary Devices

[0168] Accordingly, the embodiment of the present application also provides a device for generating questions and answers, and the execution subject may be an electronic device, such as a terminal device, a server, etc. The terminal device mentioned here may include but is not limited to at least one of the following: a mobile phone, a tablet computer, a vehicle-mounted terminal, a personal digital assistant device, etc.

[0169] like Figure 3 As shown, the device may include:

[0170] The first receiving module 301 is used to receive query information input by a user; wherein the query information is used to query content related to sparring.

[0171] The query module 302 is used to match the corresponding target corpus in the target document content according to the query information; wherein the target document content includes the required sparring content and is stored in the knowledge base.

[0172] The generation module 303 is used to generate training questions and answers based on the query information and the target corpus.

[0173] In some optional embodiments, the generating module 303 may include:

[0174] A generating unit is used to input the query information and the target corpus into a large language model, and generate the training questions and the answers to the questions through the large language model.

[0175] In some optional embodiments, the query information is used to query document content related to target job training.

[0176] The device may also include:

[0177] A storage module is used to store the generated training questions and the answers to the questions in a training question bank for the target position.

[0178] In some optional embodiments, the device may further include:

[0179] The second receiving module is used to receive the login information of at least one student.

[0180] The first association establishing module is used to establish an association relationship between the login information of the at least one student and the training question bank of the target position.

[0181] In some optional embodiments, the device may further include:

[0182] The third receiving module is used to receive the scoring criteria set for the training questions.

[0183] The second association establishing module is used to establish an association relationship between the scoring criteria and the training questions.

[0184] In some optional embodiments, when the training topic is a subjective topic, the scoring criteria include at least one keyword.

[0185] The device may also include:

[0186] A keyword generation module is used to input the generated training questions and the answers to the questions into a large language model, and generate keywords corresponding to the training questions through the large language model and the task instruction information in the large language model; wherein the task instruction information is used to indicate the determination requirements of the keywords.

[0187] The device for generating questions and answers provided in this embodiment belongs to the same application concept as the method for generating questions and answers provided in the above embodiments of this application, and can execute the method for generating questions and answers provided in any of the above embodiments of this application, and has the corresponding functional modules and beneficial effects of the execution method. For technical details not fully described in this embodiment, please refer to the specific processing content of the method for generating questions and answers provided in the above embodiments of this application, and will not be repeated here.

[0188] Accordingly, the embodiment of the present application also provides a smart training device, the execution subject can be an electronic device, such as a terminal device, and the terminal device described here can include but is not limited to at least one of the following: a mobile phone, a tablet computer, a vehicle terminal, a personal digital assistant device, etc. The terminal device supports the smart training function. The smart training function can be called out through a mini program or a dedicated smart training APP.

[0189] like Figure 4 As shown, the device may include:

[0190] The determination module 401 is used to determine the target training question bank.

[0191] Among them, the target training question bank includes training questions and answers corresponding to the training questions, and the training questions and answers are pre-generated based on query information and target corpus matching the query information in the target document content, and the query information is used to query content related to training, and the target document content includes the required training content and is stored in the knowledge base.

[0192] The training module 402 is used to perform intelligent training based on the training questions and answers in the training question bank.

[0193] In some optional embodiments, the determining module 401 may include:

[0194] The acquisition unit is used to obtain login information for the training system.

[0195] The first determination unit is used to determine the target training question bank that matches the login information; wherein an association relationship is pre-established between the login information and the target training question bank.

[0196] In some optional embodiments, the training module 402 may include:

[0197] The second determination unit is used to determine whether the student's answer to the target training question is complete based on the target training question when receiving an instruction to complete the answer to the target training question; wherein the target training question is a subjective question in the target training question bank.

[0198] The prompt unit is used to output a prompt message when it is determined that the reply content is incomplete; wherein the prompt message is used to prompt the learner to provide a complete reply.

[0199] In some optional embodiments, the training module 402 may include:

[0200] The output unit is used to output a target training question through a preset virtual character; wherein the target training question is a training question in the training question bank.

[0201] The intelligent training device provided in this embodiment belongs to the same application concept as the intelligent training method provided in the above embodiments of this application, and can execute the intelligent training method provided in any of the above embodiments of this application, and has the corresponding functional modules and beneficial effects of the execution method. For technical details not fully described in this embodiment, please refer to the specific processing content of the intelligent training method provided in the above embodiments of this application, and will not be repeated here.

[0202] It should be understood that the above-mentioned question and answer generation device and the modules in the intelligent training device can be implemented in the form of a processor calling software. For example, the device includes a processor, the processor is connected to a memory, and instructions are stored in the memory. The processor calls the instructions stored in the memory to implement any of the above methods or realize the functions of each unit of the device, wherein the processor can be a general-purpose processor, such as a CPU or a microprocessor, etc., and the memory can be a memory in the device or a memory outside the device. Alternatively, the unit in the device can be implemented in the form of a hardware circuit, and the functions of some or all units can be realized by designing the hardware circuit. The hardware circuit can be understood as one or more processors; for example, in one implementation, the hardware circuit is an ASIC, and the functions of some or all of the above units are realized by designing the logical relationship of the components in the circuit; for another example, in another implementation, the hardware circuit can be implemented by PLD, taking FPGA as an example, which can include a large number of logic gate circuits, and the connection relationship between the logic gate circuits is configured by a configuration file, so as to realize the functions of some or all of the above units. All units of the above device can be implemented in the form of a processor calling software, or in the form of a hardware circuit, or in the form of a processor calling software, and the remaining part can be implemented in the form of a hardware circuit.

[0203] In an embodiment of the present application, a processor is a circuit with the ability to process signals. In one implementation, the processor may be a circuit with the ability to read and run instructions, such as a CPU, a microprocessor, a GPU, or a DSP; in another implementation, the processor may implement certain functions through the logical relationship of a hardware circuit, and the logical relationship of the hardware circuit is fixed or reconfigurable, such as a hardware circuit implemented by an ASIC or PLD, such as an FPGA. In a reconfigurable hardware circuit, the process of the processor loading a configuration document to implement the hardware circuit configuration can be understood as the process of the processor loading instructions to implement the functions of some or all of the above units. In addition, it can also be a hardware circuit designed for artificial intelligence, which can be understood as an ASIC, such as an NPU, TPU, DPU, etc.

[0204] It can be seen that each unit in the above device can be one or more processors (or processing circuits) configured to implement the above method, such as: CPU, GPU, NPU, TPU, DPU, microprocessor, DSP, ASIC, FPGA, or a combination of at least two of these processor forms.

[0205] In addition, all or part of the units in the above device can be integrated together, or can be implemented independently. In one implementation, these units are integrated together and implemented in the form of a SOC. The SOC may include at least one processor for implementing any of the above methods or implementing the functions of each unit of the device. The type of the at least one processor may be different, for example, including a CPU and an FPGA, a CPU and an artificial intelligence processor, a CPU and a GPU, etc.

[0206] Exemplary Electronic Devices

[0207] The present application also provides an electronic device, such as Figure 5 As shown, the electronic device includes: a memory 500 and a processor 510 .

[0208] The memory 500 is connected to the processor 510 and is used to store programs.

[0209] The processor 510 is used to implement the method for generating questions and answers or the intelligent training method in the above-mentioned embodiment by running the program stored in the memory 500.

[0210] Specifically, the electronic device may further include: a communication interface 520 , an input device 530 , an output device 540 and a bus 550 .

[0211] The processor 510, the memory 500, the communication interface 520, the input device 530 and the output device 540 are connected to each other via a bus.

[0212] Bus 550 may include a pathway for transferring information between the various components of the computer system.

[0213] The processor 510 may be a general-purpose processor, such as a general-purpose central processing unit (CPU), a microprocessor, etc., or an application-specific integrated circuit (ASIC), or one or more integrated circuits for controlling the execution of the program of the scheme of the present invention. It may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0214] The processor 510 may include a main processor, and may also include a baseband chip, a modem, and the like.

[0215] The memory 500 stores a program for executing the technical solution of the present invention, and may also store an operating system and other key services. Specifically, the program may include a program code, and the program code includes a computer operation instruction. More specifically, the memory 500 may include a read-only memory (ROM), other types of static storage devices that can store static information and instructions, a random access memory (RAM), other types of dynamic storage devices that can store information and instructions, a disk storage, a flash, and the like.

[0216] The input device 530 may include a device for receiving data and information input by a user, such as a keyboard, a mouse, a camera, a scanner, a light pen, a voice input device, a touch screen, a pedometer, or a gravity sensor.

[0217] Output device 540 may include devices that allow information to be output to a user, such as a display screen, printer, speaker, etc.

[0218] The communication interface 520 may include any transceiver or the like to communicate with other devices or communication networks, such as Ethernet, Radio Access Network (RAN), Wireless Local Area Network (WLAN), etc.

[0219] The processor 510 executes the program stored in the memory 500 and calls other devices, which can be used to implement the various steps of the method for generating questions and answers or the intelligent training method provided in the above embodiments of the present application.

[0220] Exemplary computer program products and storage media

[0221] In addition to the above-mentioned methods and devices, an embodiment of the present application may also be a computer program product, which includes computer program instructions, which, when executed by a processor, enable the processor to execute the steps in the method for generating questions and answers or the intelligent training method described in the embodiments of the present application.

[0222] The computer program product may be implemented in hardware, software or a combination thereof. In one optional embodiment, the computer program product is implemented as a computer storage medium. In another optional embodiment, the computer program product is implemented as a software product, such as a software development kit (SDK).

[0223] The computer program product may be written in any combination of one or more programming languages ​​to write program codes for performing the operations of the embodiments of the present application, including object-oriented programming languages, such as Java, C++, etc., and conventional procedural programming languages, such as "C" language or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as an independent software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0224] In addition, an embodiment of the present application may also be a storage medium on which a computer program is stored, and the computer program is executed by a processor to perform the steps of the method for generating questions and answers or the intelligent training method described in the embodiment of the present application.

[0225] For the aforementioned method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the order of the actions described, because according to the present application, some steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present application.

[0226] It should be noted that each embodiment in this specification is described in a progressive manner, and each embodiment focuses on the differences from other embodiments, and the same or similar parts between the embodiments can be referred to each other. For the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0227] The steps in the methods of each embodiment of the present application can be adjusted in order, combined and deleted according to actual needs, and the technical features recorded in each embodiment can be replaced or combined.

[0228] The modules and sub-modules in the devices and terminals in the various embodiments of the present application can be combined, divided and deleted according to actual needs.

[0229] In the several embodiments provided in the present application, it should be understood that the disclosed terminals, devices and methods can be implemented in other ways. For example, the terminal embodiments described above are only schematic, for example, the division of modules or submodules is only a logical function division, and there may be other division methods in actual implementation, for example, multiple submodules or modules can be combined or integrated into another module, 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 an indirect coupling or communication connection through some interfaces, devices or modules, which can be electrical, mechanical or other forms.

[0230] The modules or submodules described as separate components may or may not be physically separated, and the components of the modules or submodules may or may not be physical modules or submodules, that is, they may be located in one place, or they may be distributed on multiple network modules or submodules. Some or all of the modules or submodules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0231] In addition, each functional module or submodule in each embodiment of the present application may be integrated into one processing module, or each module or submodule may exist physically separately, or two or more modules or submodules may be integrated into one module. The above-mentioned integrated modules or submodules may be implemented in the form of hardware or in the form of software functional modules or submodules.

[0232] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may 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.

[0233] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented directly by hardware, software units executed by a processor, or a combination of the two. The software units may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0234] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise one..." do not exclude the existence of other identical elements in the process, method, article or device including the elements. The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for generating questions and answers, characterized in that: The method comprises: Receive query information input by a user; wherein the query information is used to query content related to sparring; According to the query information, the corresponding target corpus is matched in the target document content; wherein the target document content includes the required sparring content and is stored in the knowledge base; Generate training questions and answers based on the query information and the target corpus.

2. The method for generating questions and answers according to claim 1, characterized in that: Generating a training question and an answer to the question according to the query information and the target corpus includes: The query information and the target corpus are input into a large language model, and the training questions and the answers are generated by the large language model.

3. The method for generating questions and answers according to claim 1, characterized in that: The query information is used to query the document content related to the target position training; After generating the training questions and the answers to the questions, the method further includes: The training questions and the answers to the questions are stored in the training question bank of the target position.

4. The method for generating questions and answers according to claim 3, characterized in that: The method further comprises: receiving input of login information of at least one learner; Establish an association between the login information of the at least one student and the training question bank for the target position.

5. The method for generating questions and answers according to any one of claims 1 to 4, characterized in that: After generating the training questions and the answers to the questions, the method further includes: Receiving a scoring standard set for the training topic; Establish a correlation between the scoring criteria and the training questions.

6. The method for generating questions and answers according to any one of claim 5, characterized in that: In the case where the training topic is a subjective topic, the scoring criteria include at least one keyword; The process of determining the keywords in the scoring criteria includes: The generated training questions and the answers to the questions are input into a large language model, and keywords corresponding to the training questions are generated through the large language model and the task instruction information in the large language model; wherein the task instruction information is used to indicate the determination requirements of the keywords.

7. An intelligent training method, characterized in that: include: Determine a target training question bank; wherein the target training question bank includes training questions and answers corresponding to the training questions, the training questions and the answers are generated in advance based on query information and target corpus matching the query information in target document content, the query information is used to query content related to training, the target document content includes required training content and is stored in a knowledge base; Intelligent training is performed based on the training questions and answers in the training question bank.

8. The intelligent training method according to claim 7, characterized in that: Determining the target training question bank includes: Get login information for the sparring system; Determine the target training question bank that matches the login information; wherein an association is pre-established between the login information and the target training question bank.

9. The intelligent training method according to claim 7 or 8, characterized in that: The intelligent training based on the training questions and the answers in the training question bank includes: In the case of receiving a reply completion instruction for a target training question, determining whether the student's reply to the target training question is complete based on the target training question; wherein the target training question is a subjective question in the target training question bank; When it is determined that the reply is incomplete, a prompt message is output; wherein the prompt message is used to prompt the learner to provide a complete reply.

10. The intelligent training method according to claim 7 or 8, characterized in that: The intelligent training based on the training questions in the training question bank includes: Output a target training question through a preset virtual character; wherein the target training question is a training question in the training question bank.

11. A device for generating questions and answers, characterized in that: The device comprises: A first receiving module is used to receive query information input by a user; wherein the query information is used to query related to the training content; A query module, used to match the corresponding target corpus in the target document content according to the query information; wherein the target document content includes the required sparring content and is stored in the knowledge base; A generation module is used to generate training questions and answers based on the query information and the target corpus.

12. An intelligent training device, characterized in that: The device comprises: A determination module is used to determine a target training question bank; wherein the target training question bank includes training questions and answers corresponding to the training questions, the training questions and the answers are generated in advance based on query information and a target corpus matching the query information in the target document content, the query information is used to query content related to the training, the target document content includes the required training content, and is stored in the knowledge base; The training module is used to perform intelligent training based on the training questions and answers in the training question bank.

13. An electronic device, characterized in that: include: Memory and processor; The memory is connected to the processor and is used to store programs; The processor is used to implement the method for generating questions and answers as described in any one of claims 1 to 6 or the intelligent training method as described in any one of claims 7 to 10 by running the program in the memory.

14. A storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by the processor, it implements the method for generating questions and answers as described in any one of claims 1 to 6 or the intelligent training method as described in any one of claims 7 to 10.

15. A computer program product, characterized in that The computer program product includes a computer program, and when the computer program is executed by a processor, it implements the method for generating questions and answers as described in any one of claims 1 to 6 or the intelligent training method as described in any one of claims 7 to 10.

Images