Tutorial data generation method, device and equipment

Abstract

The invention provides a tutorial data generation method, device and equipment. The method comprises the following steps: acquiring target question information of a user; obtaining knowledge point data related to the user target question information according to the user target question information and a preset knowledge point database; according to the knowledge point data related to the user target question information, obtaining a tutorial outline of an answer about the user target question information; and obtaining target course data information about the user target question information according to the course outline and the preset knowledge point database. According to the scheme, accurate tutorial data can be generated through database comparison retrieval.

Description

Technical Field

[0001] The present invention relates to the field of computer information technology processing technology, and in particular to a tutorial data generation method, device and equipment. Background Art

[0002] Traditional manuals and tutorials need to be manually organized and collected, and then understood and output by experienced people. In this process, due to the lack of comprehensive information or differences in personnel capabilities, the quality of the output learning resources is often uneven, and some cannot even serve as a guide for teaching. Summary of the invention

[0003] The technical problem to be solved by the present invention is to provide a tutorial data generation method, device and equipment, which can generate accurate tutorial data through database comparison and retrieval.

[0004] In order to solve the above technical problems, the technical solution of the present invention is as follows:

[0005] A tutorial data generation method, comprising:

[0006] Obtain user target question information;

[0007] According to the user target question information and a preset knowledge point database, obtaining knowledge point data related to the user target question information;

[0008] Obtaining a tutorial outline of an answer to the user's target question information according to the knowledge point data related to the user's target question information;

[0009] According to the course outline and the preset knowledge point database, target course data information about the user target question information is obtained.

[0010] Optionally, obtaining knowledge point data related to the user target question information according to the user target question information and a preset knowledge point database includes:

[0011] Parsing the user target question information to obtain multiple keywords of the question information;

[0012] The multiple keywords are searched in the preset knowledge point database to obtain multiple pieces of knowledge point data about the question information.

[0013] Optionally, searching the plurality of keywords in the preset knowledge point database to obtain a plurality of knowledge point data related to the question information includes:

[0014] Vectorize the multiple keywords to obtain keyword vectors;

[0015] Comparing the keyword vector with a plurality of vectors in the knowledge point database to obtain a similarity ranking;

[0016] The top n vectors in the similarity sorting are output to obtain a plurality of knowledge point data about the question information, where n is a positive integer.

[0017] Optionally, obtaining a tutorial outline for an answer to the user's target question information according to the knowledge point data related to the user's target question information includes:

[0018] Determining prompt information about the course outline;

[0019] Taking the prompt information as a target task;

[0020] Taking a plurality of pieces of knowledge point data related to the user target question information as background knowledge;

[0021] A tutorial outline is obtained for answering the user's target question information based on the target task and background knowledge.

[0022] Optionally, obtaining target course data information about the user target question information according to the course outline and the preset knowledge point database includes:

[0023] Splitting the tutorial outline into multiple target sections;

[0024] Determine keyword information for each of the target paragraphs;

[0025] Compare the keyword information of each target paragraph in the knowledge point database respectively to obtain multiple knowledge point data of each target paragraph;

[0026] Determine the task information for each target paragraph;

[0027] Obtaining data information of each target paragraph according to the task information of each target paragraph and the plurality of knowledge point data;

[0028] The data information of each target paragraph is integrated to obtain the target tutorial data information about the user's target question information.

[0029] Optionally, the knowledge point database is obtained through the following process:

[0030] Obtain knowledge point data of multiple question-answer pairs;

[0031] Performing text splitting on the plurality of question-answer pair knowledge point data to obtain keyword information on each of the question-answer pair knowledge point data;

[0032] Vectorize the keyword information to obtain multiple knowledge point vectors;

[0033] The plurality of knowledge point vectors are stored in a database to obtain the knowledge point database.

[0034] Optionally, storing the plurality of knowledge point vectors in a database to obtain the knowledge point database includes:

[0035] Labels are added to the knowledge point vectors, and the labeled knowledge point vectors are stored in a database in the form of key-value pairs to obtain the knowledge point database.

[0036] The present invention also provides a tutorial data generating device, comprising:

[0037] The acquisition module is used to obtain the user's target question information;

[0038] A processing module is used to obtain knowledge point data related to the user target question information based on the user target question information and a preset knowledge point database; obtain a tutorial outline of the answer to the user target question information based on the knowledge point data related to the user target question information; and obtain target tutorial data information about the user target question information based on the tutorial outline and the preset knowledge point database.

[0039] The present invention also provides a computing device, comprising: a processor and a memory storing a computer program, wherein when the computer program is run by the processor, the method as described above is executed.

[0040] The present invention also provides a computer-readable storage medium storing instructions, which, when executed on a computer, enable the computer to execute the method described above.

[0041] The above solution of the present invention includes at least the following beneficial effects:

[0042] The above scheme of the present invention obtains user target question information; obtains knowledge point data related to the user target question information according to the user target question information and a preset knowledge point database; obtains a course outline of the answer to the user target question information according to the knowledge point data related to the user target question information; obtains target course data information about the user target question information according to the course outline and the preset knowledge point database. Accurate course data can be generated by database comparison and retrieval. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 is a flow chart of a method for generating tutorial data according to an embodiment of the present invention;

[0044] Figure 2 is a schematic diagram of tutorial data generation of a tutorial data generation method according to an embodiment of the present invention;

[0045] Figure 3 is a schematic diagram of vector database generation of a tutorial data generation method according to an embodiment of the present invention;

[0046] Figure 4 It is a structural diagram of a tutorial data generating device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0047] The exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present invention and to enable the scope of the present invention to be fully communicated to those skilled in the art.

[0048] like Figure 1 As shown, an embodiment of the present invention provides a tutorial data generation method, comprising:

[0049] Step 11, obtaining user target question information;

[0050] Step 12, obtaining knowledge point data related to the user target question information according to the user target question information and a preset knowledge point database;

[0051] Step 13, obtaining a tutorial outline of an answer to the user's target question information based on the knowledge point data related to the user's target question information;

[0052] Step 14: Obtain target tutorial data information about the user target question information based on the tutorial outline and the preset knowledge point database.

[0053] In this embodiment, the user target question information includes question information about generating tutorial data, such as "generate a tutorial on modeling". After obtaining the user's target question information, firstly, based on the preset knowledge point database, the knowledge point data related to the user's target question information is retrieved. Then, according to the knowledge point data related to the user's target question information and the large language model, a tutorial outline about the user's target question information is obtained. For example, when the user question information is "generate a tutorial on modeling", the tutorial outline is the outline of the modeling tutorial. Then, according to the tutorial outline, through the knowledge point database and the large language model, the detailed tutorial data of each part in the tutorial outline is obtained in turn. When each part in the tutorial outline is generated, a complete target tutorial data information is obtained. It should be noted that the large language model described in the present invention is an artificial intelligence model designed to understand and generate human language. It is generated by pre-training with large-scale corpus data, which is one of the ways of natural language processing technology. The large language model is a deep learning model trained on a huge data set to understand human language. Its core goal is to accurately learn and understand human language. The large language model enables the machine to interpret language like we humans interpret language. The large language model described in the present invention can be any existing large language model that has been trained and can be used directly.

[0054] In an optional embodiment of the present invention, step 12 may include:

[0055] Step 121, parsing the user target question information to obtain multiple keywords of the question information;

[0056] Step 122, searching the plurality of keywords in the preset knowledge point database to obtain a plurality of knowledge point data about the question information.

[0057] In this embodiment, the user target question information is first parsed to obtain keyword information. For example, when the user target question is "Please provide tutorial materials for deploying a complete customer service consulting system", the extracted keywords are "customer service, consulting, system". The keyword extraction method in the present invention evaluates the importance of each candidate keyword by weighted summing up multiple attributes of each candidate keyword, thereby extracting key information in the text. Specifically, the score of each candidate keyword is obtained by weighted summing up the following five parts: ‌word betweenness density‌, which is used to measure the position and frequency of the word in the text; ‌the position score of the article where the word is located, which is used to quantify the influence of the position of the word in the article on the score; ‌word length score, which is used to evaluate the influence of the length of the word on the score; ‌word part of speech score, which is used to evaluate the influence of the word part of speech on the score; ‌word TFIDF value, which is used to measure the frequency and importance of the word in the document. The scores of the above five parts are weighted summed to obtain the comprehensive score of each candidate keyword, and the size of the comprehensive score is used to measure whether the candidate keyword is the final keyword.

[0058] After obtaining the keywords of the user's question information, the keywords are compared and searched in a preset knowledge point database to obtain a plurality of knowledge point data related to the keywords stored in the knowledge point database.

[0059] In an optional embodiment of the present invention, step 122 may include:

[0060] Step 1221, vectorizing the multiple keywords to obtain keyword vectors;

[0061] Step 1222, performing a similarity comparison between the keyword vector and a plurality of vectors in the knowledge point database to obtain a similarity ranking;

[0062] Step 1223, output the top n vectors in the similarity sorting to obtain multiple pieces of knowledge point data about the question information, where n is a positive integer.

[0063] In this embodiment, when searching for keywords in the knowledge point database, the method of word segmentation and vector comparison retrieval is adopted. When searching for the keywords "customer service, consultation, system", more services and components for designing customer service systems will be retrieved. The large language model organizes and generates the outline and composition framework of this tutorial based on the searched information. Specifically, the keywords are first vectorized, and the vectorization is to map human language symbols to a digital space that can be understood by the computer. Specifically, in the present invention, a neural network model that can convert words into vectors is trained, and the neural network model includes an input layer, a projection layer, a hidden layer and an output layer. The input layer receives a set of unique hot encodings of context words as input, and each word is mapped to a unique integer, which is used as input in the neural network. The vector representation of each word in the projection layer is taken out and averaged to obtain an average vector of the size of the vector dimension, and this average vector is used as the input of the hidden layer. In the hidden layer, the word vectors obtained by the projection layer are added to obtain the vector representation of the target word. This process is equivalent to performing multiple linear regressions on the input words, and the dimension of the parameter vector of each neuron is consistent with the dimension of the input vector. The vector of the hidden layer in the output layer is input to the output layer to obtain the vector representation of the target word.

[0064] Compare the keyword vector with the vectors stored in the knowledge point database to obtain n vectors with high similarity. Each vector in the knowledge point database represents a knowledge point, and output the n vectors with high similarity to obtain the relevant knowledge point data retrieved by the group of keywords.

[0065] In an optional embodiment of the present invention, step 13 may include:

[0066] Step 131, determining prompt information about the course outline;

[0067] Step 132, taking the prompt information as the target task;

[0068] Step 133, taking multiple pieces of knowledge point data related to the user's target question information as background knowledge;

[0069] Step 134, obtaining a tutorial outline of an answer to the user's target question information based on the target task and background knowledge.

[0070] In this embodiment, according to the user target question information, a new prompt information for generating a tutorial outline is determined. For example, when the user target question information is "Please provide tutorial materials for deploying a complete customer service consulting system", the prompt information about the tutorial outline can be "Please combine the following materials to provide a tutorial outline for deploying a complete customer service consulting system". The prompt information is used as the target task, and the multiple knowledge point data retrieved from the knowledge point database are used as background knowledge, that is, context. The target task and the background knowledge are input into the large language model together, and a tutorial outline for the answer to the user target question information can be obtained. Since the large language model has been trained and can be put into use, when the large language model receives the target task and background knowledge input by the user, it can understand the user's language intention and finally output a reasonable answer, that is, the tutorial outline, through the model's own network architecture and calculation process. This process is similar to the large language model that can communicate with humans in the existing artificial intelligence technology, both of which receive the user's input questions and then use their own natural language processing capabilities to give answers. I will not go into details here.

[0071] By inputting the background knowledge retrieved from the knowledge point database into the large language model, the model's ability to understand and answer the target question information can be enhanced. Thus, external knowledge is integrated into the generation process, making the generated text content richer, more accurate and in line with user needs. The method of combining information retrieval and the large language model in the present invention can use the information retrieved from the knowledge point database to guide the generation of text, thereby improving the quality and accuracy of text generation.

[0072] In an optional embodiment of the present invention, step 14 may include:

[0073] Step 141, splitting the course outline into multiple target paragraphs;

[0074] Step 142, determining keyword information of each of the target paragraphs;

[0075] Step 143, comparing the keyword information of each target paragraph in the knowledge point database to obtain multiple pieces of knowledge point data for each target paragraph;

[0076] Step 144, determining the task information of each target paragraph;

[0077] Step 145, obtaining data information of each target paragraph according to the task information of each target paragraph and the plurality of knowledge point data;

[0078] Step 146, integrating the data information of each target paragraph to obtain target tutorial data information about the user's target question information.

[0079] In this embodiment, Figure 2 As shown, the tutorial outline includes target paragraphs such as introduction, system composition, framework diagram, components and descriptions, deployment method of each component, relevant precautions, and frequently asked questions and answers. Keywords are extracted for each target paragraph, and after the keywords are extracted, they are searched in the knowledge point database to obtain the relevant knowledge point data for each target paragraph. The retrieved relevant knowledge point data is used as background knowledge, and the background knowledge and prompt information are input into the large language model together to obtain detailed content information for each target paragraph. It should be noted that the target tutorial data information can be tutorial data in multiple formats determined according to user needs, such as PPT, DOC, and layout documents.

[0080] In an optional embodiment of the present invention, the knowledge point database is obtained through the following process:

[0081] Step 21, obtaining multiple question-answer pair knowledge point data;

[0082] Step 22, performing text splitting on the plurality of question-answer pair knowledge point data to obtain keyword information of each of the question-answer pair knowledge point data;

[0083] Step 23, vectorizing the keyword information to obtain multiple knowledge point vectors;

[0084] Step 24, storing the plurality of knowledge point vectors in a database to obtain the knowledge point database.

[0085] In this embodiment, Figure 3 As shown, the knowledge point database is preferably a vector database, which stores multiple vectorized knowledge point data. The knowledge point data is some fragmented knowledge points in the form of one question and one answer. For example, the following question and answer pair is a piece of knowledge point data:

[0086] Question: What is the notification mechanism of Nacos (a platform for dynamic service discovery, configuration management, and service management)?

[0087] Answer: When a service instance no longer provides services, it can initiate an offline request through the Nacos client. The client will actively notify the Nacos server that it has been offline and is no longer available. This notification process includes destroying the service, stopping the heartbeat, and removing the service instance request from the Nacos service registry. After receiving the notification, the server will update the service registry and trigger a notification event to all clients listening to the service to update their locally cached service list. This design ensures that even if the service is offline, other parts of the system can be updated in time to prevent continuous calls to offline services.

[0088] Specifically, the establishment of the knowledge point database first requires the content segmentation and extraction of the questions and answers of the question-answer pair according to a certain granularity, and the extraction of keywords. For example, in the question-answer pair mentioned above, the keywords are "Nacos", "notification mechanism", "registration", "notification", etc.

[0089] The segmented text is then vectorized, that is, converted into vectors for subsequent retrieval. The vectorization model is responsible for converting text into vectors, capturing semantic relationships, so that similar sentences can be identified by vector differences. The vectorization model finds word vectors to represent the meaning of sentences or text.

[0090] The vectorized data is stored in the database, and the vectors generated by all knowledge points are stored here to meet the needs of rapid retrieval of massive knowledge vectors. When searching, the user's question information (such as the deployment method of nacos) is also segmented into text, and input into the vector database as the search condition for retrieval and recall. Then all knowledge points related to nacos can be obtained, such as the working principle, deployment method, problem location, etc. of nacos.

[0091] In an optional embodiment of the present invention, step 24 may include:

[0092] Step 241 , adding a label to the knowledge point vector, and storing the labeled knowledge point vector in a database in the form of a key-value pair to obtain the knowledge point database.

[0093] In this embodiment, the knowledge point vectors are stored in the database in the form of key-value pairs, where the question of the question-answer pair corresponds to the key and the answer to the question-answer pair corresponds to the value. An index is created for the key-value pair, and the question vector is stored in the index structure in association with the answer. The vector representation of the user query is compared with the question vector in the index to retrieve the most matching question-answer pair.

[0094] The implementation process of the above method is further described below with a specific embodiment:

[0095] Embodiment 1,

[0096] A user asked to write a "deployment document about nacos middleware". After receiving this request, the system will split the question information into several keywords such as nacos, middleware, deployment, and document. Then these keywords are input into the vectorized database, and all knowledge points matching these keywords can be retrieved.

[0097] Through the keywords of middleware, deployment, and document, you can find some outline templates for such tutorials, such as operation and maintenance, deployment, etc. Through searching, you can get a standard tutorial step, which is convenient for users to learn and understand. The content of this outline is as follows:

[0098] Introduction;

[0099] Background related technology;

[0100] Data or information that needs to be prepared;

[0101] Deployment steps;

[0102] Check the dependency environment;

[0103] Download the installation package;

[0104] Installation and deployment;

[0105] verify;

[0106] Some work after installation;

[0107] Possible problems and answers;

[0108] According to the above outline, each paragraph needs to be searched and recalled and organized into a large model, and then detailed content is generated. In this example, you can search by several keywords. For example, in the first step of the outline deployment step in this example, the section of checking the dependent environment can be searched using the keywords nacos+dependent environment, which will list all the dependent environments and system requirements required by nacos.

[0109] With all the knowledge point information retrieved above, a prompt word input large model is constructed for reorganization. For example, this prompt word can be "Please combine the context information and organize this information into a copy suitable for teaching use. The content must be detailed and the description must be clear. It is best to organize the content in a list format." In this way, the large language model will output a paragraph of easy-to-understand description text, and generate a description copy suitable for teaching for the system and operating environment that nacos relies on.

[0110] Fill the generated text into the corresponding position of the outline, and generate the content of the next item in the outline in sequence until the entire lesson plan document is completely generated.

[0111] The above method of the present invention is implemented based on a large number of independent knowledge points. It relies on an existing or independently established knowledge question and answer system and is connected according to the characteristics of the knowledge point source system. The ultimate goal is to organize the content into knowledge point data in the form of question and answer pairs. In order to facilitate retrieval, the collected knowledge points need to be processed, such as using various cleaning scripts or model methods to extract content, label and vectorize them, and store them in a vector database.

[0112] The above method of the present invention mainly utilizes the characteristics of vector retrieval and large language models, and uses a process similar to the traditional manual writing of tutorials or instructions. The entire process from outline writing to content generation is completed by two steps of search and generation. Therefore, the characteristics of conditional retrieval and large language models can be fully utilized. According to the needs and scenarios of users, the prompt words and search conditions of each process are customized to achieve the generation of tutorial documents that best meet the needs of users. With the continuous enrichment and supplementation of the knowledge point database, the generated instructions and tutorials can be updated to improve and enrich the tutorial content.

[0113] like Figure 4 As shown, an embodiment of the present invention further provides a tutorial data generating device 40, comprising:

[0114] The acquisition module 41 is used to acquire the user's target question information;

[0115] Processing module 42 is used to obtain knowledge point data related to the user target question information based on the user target question information and a preset knowledge point database; obtain a tutorial outline of the answer to the user target question information based on the knowledge point data related to the user target question information; and obtain target tutorial data information about the user target question information based on the tutorial outline and the preset knowledge point database.

[0116] Optionally, obtaining knowledge point data related to the user target question information according to the user target question information and a preset knowledge point database includes:

[0117] Parsing the user target question information to obtain multiple keywords of the question information;

[0118] The multiple keywords are searched in the preset knowledge point database to obtain multiple pieces of knowledge point data about the question information.

[0119] Optionally, searching the plurality of keywords in the preset knowledge point database to obtain a plurality of knowledge point data related to the question information includes:

[0120] Vectorize the multiple keywords to obtain keyword vectors;

[0121] Comparing the keyword vector with a plurality of vectors in the knowledge point database to obtain a similarity ranking;

[0122] The top n vectors in the similarity sorting are output to obtain a plurality of knowledge point data about the question information, where n is a positive integer.

[0123] Optionally, obtaining a tutorial outline for an answer to the user's target question information according to the knowledge point data related to the user's target question information includes:

[0124] Determining prompt information about the course outline;

[0125] Taking the prompt information as a target task;

[0126] Taking a plurality of pieces of knowledge point data related to the user target question information as background knowledge;

[0127] A tutorial outline is obtained for answering the user's target question information based on the target task and background knowledge.

[0128] Optionally, obtaining target course data information about the user target question information according to the course outline and the preset knowledge point database includes:

[0129] Splitting the tutorial outline into multiple target sections;

[0130] Determine keyword information for each of the target paragraphs;

[0131] Compare the keyword information of each target paragraph in the knowledge point database respectively to obtain multiple knowledge point data of each target paragraph;

[0132] Determine the task information for each target paragraph;

[0133] Obtaining data information of each target paragraph according to the task information of each target paragraph and the plurality of knowledge point data;

[0134] The data information of each target paragraph is integrated to obtain the target tutorial data information about the user's target question information.

[0135] Optionally, the knowledge point database is obtained through the following process:

[0136] Obtain knowledge point data of multiple question-answer pairs;

[0137] Performing text splitting on the plurality of question-answer pair knowledge point data to obtain keyword information on each of the question-answer pair knowledge point data;

[0138] Vectorize the keyword information to obtain multiple knowledge point vectors;

[0139] The plurality of knowledge point vectors are stored in a database to obtain the knowledge point database.

[0140] Optionally, storing the plurality of knowledge point vectors in a database to obtain the knowledge point database includes:

[0141] Labels are added to the knowledge point vectors, and the labeled knowledge point vectors are stored in a database in the form of key-value pairs to obtain the knowledge point database.

[0142] It should be noted that the device is a device corresponding to the above method, and all implementation methods in the above method embodiments are applicable to the embodiments of the device and can achieve the same technical effects.

[0143] The embodiment of the present invention further provides a computing device, comprising: a processor, a memory storing a computer program, wherein when the computer program is executed by the processor, the method described above is executed. All implementations in the above method embodiment are applicable to this embodiment and can achieve the same technical effect.

[0144] The embodiment of the present invention also provides a computer-readable storage medium storing instructions, which, when executed on a computer, enable the computer to execute the method described above. All implementations in the above method embodiment are applicable to this embodiment and can achieve the same technical effect.

[0145] Those skilled 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 the present invention.

[0146] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0147] In the embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the 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 devices or units, which can be electrical, mechanical or other forms.

[0148] 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.

[0149] In addition, each functional unit in each embodiment of the present invention 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.

[0150] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, ROM, RAM, magnetic disks, or optical disks.

[0151] In addition, it should be pointed out that in the apparatus and method of the present invention, it is obvious that each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations should be regarded as equivalent schemes of the present invention. Moreover, the steps of performing the above series of processing can naturally be performed in chronological order according to the order of description, but it is not necessary to perform them in chronological order, and some steps can be performed in parallel or independently of each other. For those of ordinary skill in the art, it is understandable that all or any steps or components of the method and apparatus of the present invention can be implemented in hardware, firmware, software or a combination thereof in any computing device (including processors, storage media, etc.) or a network of computing devices, which can be achieved by those of ordinary skill in the art using their basic programming skills after reading the description of the present invention.

[0152] Therefore, the purpose of the present invention can also be achieved by running a program or a group of programs on any computing device. The computing device can be a well-known general device. Therefore, the purpose of the present invention can also be achieved by simply providing a program product containing a program code that implements the method or device. That is to say, such a program product also constitutes the present invention, and the storage medium storing such a program product also constitutes the present invention. Obviously, the storage medium can be any well-known storage medium or any storage medium developed in the future. It should also be pointed out that in the device and method of the present invention, it is obvious that each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations should be regarded as equivalent schemes of the present invention. In addition, the steps of performing the above-mentioned series of processing can naturally be performed in chronological order according to the order of description, but it is not necessary to perform them in chronological order. Some steps can be performed in parallel or independently of each other.

[0153] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A tutorial data generation method, characterized in that: include: Obtain user target question information; According to the user target question information and a preset knowledge point database, obtaining knowledge point data related to the user target question information; Obtaining a tutorial outline of an answer to the user's target question information according to the knowledge point data related to the user's target question information; According to the course outline and the preset knowledge point database, target course data information about the user target question information is obtained.

2. The tutorial data generation method according to claim 1, characterized in that: According to the user target question information and a preset knowledge point database, knowledge point data related to the user target question information is obtained, including: Parsing the user target question information to obtain multiple keywords of the question information; The multiple keywords are searched in the preset knowledge point database to obtain multiple pieces of knowledge point data about the question information.

3. The tutorial data generation method according to claim 2, characterized in that: The multiple keywords are searched in the preset knowledge point database to obtain multiple pieces of knowledge point data about the question information, including: Vectorize the multiple keywords to obtain keyword vectors; Comparing the keyword vector with a plurality of vectors in the knowledge point database to obtain a similarity ranking; The top n vectors in the similarity sorting are output to obtain a plurality of knowledge point data about the question information, where n is a positive integer.

4. The tutorial data generation method according to claim 1, characterized in that: Obtaining a tutorial outline for an answer to the user's target question information according to the knowledge point data related to the user's target question information includes: Determining prompt information about the course outline; Taking the prompt information as a target task; Taking a plurality of pieces of knowledge point data related to the user target question information as background knowledge; A tutorial outline is obtained for answering the user's target question information based on the target task and background knowledge.

5. The tutorial data generation method according to claim 1, characterized in that: According to the course outline and the preset knowledge point database, target course data information about the user target question information is obtained, including: Splitting the tutorial outline into multiple target sections; Determine keyword information for each of the target paragraphs; Compare the keyword information of each target paragraph in the knowledge point database respectively to obtain multiple knowledge point data of each target paragraph; Determine the task information for each target paragraph; Obtaining data information of each target paragraph according to the task information of each target paragraph and the plurality of knowledge point data; The data information of each target paragraph is integrated to obtain the target tutorial data information about the user's target question information.

6. The tutorial data generation method according to any one of claims 1 to 5, characterized in that: The knowledge point database is obtained through the following process: Obtain knowledge point data of multiple question-answer pairs; Performing text splitting on the plurality of question-answer pair knowledge point data to obtain keyword information on each of the question-answer pair knowledge point data; Vectorize the keyword information to obtain multiple knowledge point vectors; The plurality of knowledge point vectors are stored in a database to obtain the knowledge point database.

7. The tutorial data generation method according to claim 6, characterized in that: Storing the plurality of knowledge point vectors in a database to obtain the knowledge point database includes: Labels are added to the knowledge point vectors, and the labeled knowledge point vectors are stored in a database in the form of key-value pairs to obtain the knowledge point database.

8. A tutorial data generating device, characterized in that: include: The acquisition module is used to obtain the user's target question information; A processing module, used for obtaining knowledge point data related to the user target question information according to the user target question information and a preset knowledge point database; Obtaining a tutorial outline of an answer to the user's target question information according to the knowledge point data related to the user's target question information; According to the course outline and the preset knowledge point database, target course data information about the user target question information is obtained.

9. A computing device, characterized in that include: A processor and a memory storing a computer program, wherein when the computer program is executed by the processor, the method according to any one of claims 1 to 7 is performed.

10. A computer-readable storage medium, characterized in that: Instructions are stored, and when the instructions are executed on a computer, the computer is caused to execute the method according to any one of claims 1 to 7.

Images