Intelligent question answering method, system and equipment based on large model, medium and product

Abstract

The invention discloses an intelligent question answering method, system and device based on a large model, a medium and a product. The method comprises the steps of obtaining an original question; generating an answer to the original question and a plurality of derivative questions associated with the original question according to a large model; inputting the plurality of derivative questions into a preset multi-role dialogue model to obtain an answer of each derivative question output by the multi-role dialogue model; and outputting and displaying the answer of the original question and the answers of the plurality of derivative questions. The question input by the user is expanded into a plurality of questions by using the large model, and the corresponding answers are generated and output to the user at one time, so that the cost of multiple questions of the user is effectively reduced, the comprehensiveness and accuracy of questions and answers are improved, and the user experience is improved.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence technology, and in particular to an intelligent question-answering method, system, device, medium, and product based on a large model. Background Technology

[0002] A large model typically refers to a model with a large number of parameters and complex computational structures in the fields of machine learning and artificial intelligence. Large models have wide applications in many fields, including but not limited to Natural Language Processing (NLP), Computer Vision (CV), speech recognition, recommender systems, and question answering systems.

[0003] Current large-scale generative question-answering systems follow a multi-turn dialogue model. The user asks a question, and the large model provides a response. If the user wants to learn more, they need to ask questions in successive rounds, which is cumbersome. Furthermore, the input length of the large model is limited; if there are too many dialogue rounds, the large model may forget previous questions and responses, leading to inaccurate responses in new rounds. Additionally, the way current large-scale question-answering systems present answers to users through multiple rounds results in a fragmented and incomplete presentation of content. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide an intelligent question-answering method, system, device, medium and product based on a large model. The large model expands the user's input question into multiple questions and generates corresponding answers to output to the user at once, which effectively reduces the cost of users asking multiple questions, improves the comprehensiveness and accuracy of question and answer, and enhances the user experience.

[0005] To achieve the above objectives, embodiments of the present invention provide an intelligent question-answering method based on a large model, comprising:

[0006] Get the original question;

[0007] The answer to the original question is generated based on the large model, as well as multiple derivative questions associated with the original question;

[0008] The multiple derivative questions are input into a preset multi-role dialogue model to obtain the answer to each of the derivative questions output by the multi-role dialogue model.

[0009] The answer to the original question, as well as the answers to multiple derivative questions, will be output and displayed.

[0010] As an improvement to the above scheme, the step of generating the answer to the original question based on the large model, and multiple derivative questions associated with the original question, includes:

[0011] The answer to the original question is generated based on the large model;

[0012] Based on the large model, multiple knowledge points associated with the original problem are generated;

[0013] Based on the large model, multiple derivative questions are generated for each of the aforementioned knowledge points.

[0014] As an improvement to the above solution, the multi-role dialogue model includes a first-role dialogue model, a retrieval unit, and a second-role dialogue model; wherein, both the first-role dialogue model and the second-role dialogue model are built based on the large model.

[0015] The first role dialogue model is used to generate at least one primary extension question based on the derived question;

[0016] The search engine is used to retrieve relevant reference content based on the extended question;

[0017] The second role dialogue model is used to generate corresponding answers based on the extended questions.

[0018] As an improvement to the above solution, the step of inputting the multiple derivative questions into a preset multi-role dialogue model to obtain the answer to each of the derivative questions output by the multi-role dialogue model includes:

[0019] The multiple derivative questions are input into the multi-role dialogue model;

[0020] Based on the first character dialogue model and the preset first prompt words, generate extended questions associated with each of the derived questions;

[0021] The extended question is retrieved using the retrieval tool to obtain reference content for the extended question;

[0022] The answer to the extended question is generated based on the second role dialogue model, the reference content, and the preset second prompt words;

[0023] The answer to the extended question and the preset third prompt are input into the first character dialogue model for iterative looping until the preset number of loops or the loop stopping condition is reached, so as to obtain the answer to at least one extended question.

[0024] The answers to at least one extended question of each of the derived questions are aggregated to obtain the answer to each of the derived questions.

[0025] As an improvement to the above solution, the step of inputting the answer to the extended question and a preset third prompt word into the first role dialogue model for iterative looping until a preset number of loops or a loop stopping condition is reached, to obtain an answer to at least one extended question, includes:

[0026] The answer to the extended question and the preset third prompt word are input into the first character dialogue model;

[0027] If the first character dialogue model generates a new extended question, then return to the step of retrieving the extended question according to the retrieval tool to obtain the reference content of the extended question, until the preset number of loop rounds or the loop stopping condition is reached, and obtain the answers to multiple extended questions;

[0028] If the first character's dialogue model does not generate a new extended question, the dialogue ends, and an answer to an extended question is obtained.

[0029] As an improvement to the above solution, the step of outputting and displaying the answer to the original question and the answers to multiple derivative questions includes:

[0030] The answer to the original question, the multiple derivative questions associated with the original question, and the answer to each of the derivative questions are output to the user terminal for visualization display according to a preset format.

[0031] As an improvement to the above solution, the step of outputting and displaying the answer to the original question and the answers to multiple derivative questions includes:

[0032] The answer to the original question, the multiple knowledge points associated with the original question, the multiple derivative questions corresponding to each knowledge point, and the answer to each derivative question are output to the user terminal for visualization display according to a preset format.

[0033] This invention also provides an intelligent question-answering system based on a large model, comprising:

[0034] The retrieval module is used to retrieve the original question;

[0035] A generation module is used to generate the answer to the original question and multiple derivative questions associated with the original question based on the large model;

[0036] The dialogue module is used to input the multiple derivative questions into a preset multi-role dialogue model and obtain the answer to each of the derivative questions output by the multi-role dialogue model.

[0037] The output module is used to output and display the answer to the original question and the answers to multiple derivative questions.

[0038] This invention also provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor executes the computer program to implement the intelligent question-answering method based on a large model as described above.

[0039] This invention also provides a computer-readable storage medium comprising a stored computer program, wherein the computer program, when running, controls the device where the computer-readable storage medium is located to execute the intelligent question-answering method based on a large model as described above.

[0040] This invention also provides a computer program product, which includes a computer program or computer instructions. When the computer program or computer instructions are executed by a processor, they implement the intelligent question-answering method based on a large model as described above.

[0041] Compared to existing technologies, the beneficial effects of the intelligent question-answering method, system, device, medium, and product based on a large model provided by this invention are as follows: First, an original question is obtained. Then, an answer to the original question and multiple derivative questions associated with the original question are generated according to the large model. Next, the multiple derivative questions are input into a preset multi-role dialogue model to obtain the answer to each of the derivative questions output by the multi-role dialogue model. Finally, the answer to the original question and the answers to the multiple derivative questions are output and displayed. This invention utilizes a large model to expand the user-input question into multiple questions and generate corresponding answers, outputting them to the user all at once. This effectively reduces the cost of multiple question-asking for the user, improves the comprehensiveness and accuracy of question-answering, and enhances the user experience. Attached Figure Description

[0042] Figure 1 This is a flowchart illustrating a preferred embodiment of an intelligent question-answering method based on a large model provided by the present invention;

[0043] Figure 2 This is a flowchart illustrating another preferred embodiment of an intelligent question-answering method based on a large model provided by the present invention;

[0044] Figure 3 This is a schematic diagram of a preferred embodiment of an intelligent question-answering system based on a large model provided by the present invention;

[0045] Figure 4 This is a schematic diagram of a preferred embodiment of a terminal device provided by the present invention. Detailed Implementation

[0046] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0047] Please see Figure 1 , Figure 1 This is a flowchart illustrating a preferred embodiment of a large-model-based intelligent question answering method provided by the present invention. The large-model-based intelligent question answering method includes:

[0048] S1, obtain the original problem;

[0049] S2, Generate the answer to the original question and multiple derivative questions associated with the original question based on the large model;

[0050] S3, input the multiple derivative questions into a preset multi-role dialogue model to obtain the answer to each of the derivative questions output by the multi-role dialogue model;

[0051] S4, output and display the answer to the original question and the answers to multiple derivative questions.

[0052] Specifically, this invention provides an intelligent question-answering method based on a large model. First, the original question input by the user is obtained. Then, a large model is invoked to generate an answer to the original question, as well as multiple derivative questions associated with the original question. For example, the large model can be used to expand and associate based on the keywords and themes of the original question. For instance, semantic analysis is performed on the original question to extract the main entities and concepts, and then the large model is used to generate other derivative questions related to these entities and concepts. Subsequently, multiple derivative questions are input into a preset multi-role dialogue model. The model performs multiple rounds of dialogue for each derivative question to obtain an answer for each derivative question. It should be noted that multiple threads can be created in the multi-role dialogue model, with each thread performing multiple rounds of dialogue for each derivative question to obtain an answer for each derivative question. Multiple threads can operate in parallel, allowing for the simultaneous generation of answers to multiple derivative questions. Finally, the answer to the original question and the answers to the multiple derivative questions are compiled and summarized, and output to the user terminal for visualization according to a preset format.

[0053] This invention utilizes a large model to expand the user's input question into multiple questions and generates corresponding answers to output to the user all at once. This effectively reduces the cost of users asking multiple questions, improves the comprehensiveness and accuracy of the question and answer system, and enhances the user experience.

[0054] In another preferred embodiment, step S1 generates the answer to the original question and a plurality of derivative questions associated with the original question based on a large model, including:

[0055] S101, Generate the answer to the original question based on the large model;

[0056] S102, Generate multiple knowledge points associated with the original problem based on the large model;

[0057] S103, Generate multiple derivative questions corresponding to each knowledge point based on the large model.

[0058] Specifically, in this embodiment of the invention, when generating the answer to the original question and multiple derivative questions associated with the original question based on the large model, the direct answer to the original question is first generated based on the large model. Then, multiple knowledge points associated with the original question are generated based on the large model and preset prompts. For each knowledge point, multiple derivative questions corresponding to each knowledge point are generated based on the large model and preset prompts, thus obtaining multiple derivative questions corresponding to multiple knowledge points. For example, based on the large model and preset prompts, such as "You are an expert in the domain of question x, please generate n relevant knowledge points about question x", n knowledge points associated with the original question are generated, namely knowledge point 1, knowledge point 2, ..., knowledge point n. Then, n threads are started, each thread creating a large model instance i. Based on the preset prompts, such as "You are an expert in the domain of 'knowledge point i', please help me generate m questions about 'knowledge point i'", large model instance i generates m derivative questions for each knowledge point i, and finally, the large model generates a total of n*m derivative questions.

[0059] In another preferred embodiment, the multi-role dialogue model includes a first role dialogue model, a retrieval device, and a second role dialogue model; wherein, both the first role dialogue model and the second role dialogue model are built based on the large model;

[0060] The first role dialogue model is used to generate at least one primary extension question based on the derived question;

[0061] The search engine is used to retrieve relevant reference content based on the extended question;

[0062] The second role dialogue model is used to generate corresponding answers based on the extended questions.

[0063] Specifically, the multi-role dialogue model in this embodiment of the invention includes a first-role dialogue model, a retrieval unit, and a second-role dialogue model. Both the first-role and second-role dialogue models are built upon a larger model. The first-role dialogue model simulates a novice, i.e., a novice dialogue model, generating at least one basic extension question based on the derived question. This extension question has relatively low complexity. The retrieval unit retrieves relevant reference content based on the extension question, providing external knowledge support for the dialogue. The second-role dialogue model simulates an expert, i.e., an expert dialogue model, generating professional answers based on the extension question.

[0064] In another preferred embodiment, step S3 involves inputting the plurality of derivative questions into a preset multi-role dialogue model to obtain an answer to each of the derivative questions output by the multi-role dialogue model, including:

[0065] S301, input the multiple derivative questions into the multi-role dialogue model;

[0066] S302, Based on the first role dialogue model and the preset first prompt words, generate extended questions associated with each of the derived questions;

[0067] S303, the extended question is retrieved using the retrieval tool to obtain reference content for the extended question;

[0068] S304, Generate the answer to the extended question based on the second role dialogue model, the reference content, and the preset second prompt words;

[0069] S305, the answer to the extended question and the preset third prompt are input into the first role dialogue model for iterative looping until the preset number of loops or the loop stopping condition is reached, and the answer to at least one extended question is obtained;

[0070] S306, aggregate the answers to at least one extended question of each of the derived questions to obtain the answer to each of the derived questions.

[0071] In another preferred embodiment, the step of inputting the answer to the extended question and a preset third prompt word into the first role dialogue model for iterative looping until a preset number of loop rounds or a loop stopping condition is reached, to obtain an answer to at least one extended question, includes:

[0072] The answer to the extended question and the preset third prompt word are input into the first character dialogue model;

[0073] If the first character dialogue model generates a new extended question, then return to the step of retrieving the extended question according to the retrieval tool to obtain the reference content of the extended question, until the preset number of loop rounds or the loop stopping condition is reached, and obtain the answers to multiple extended questions;

[0074] If the first character's dialogue model does not generate a new extended question, the dialogue ends, and an answer to an extended question is obtained.

[0075] For details, please refer to Figure 2 , Figure 2 This is a flowchart illustrating another preferred embodiment of the intelligent question-answering method based on a large model provided by the present invention. In this embodiment, n*m derivative questions are input into a multi-role dialogue model. Within the multi-role dialogue model, n*m threads are simultaneously activated, each thread performing multiple rounds of dialogue for each derivative question. Within each thread, a large model instance i, i.e., a novice dialogue model, is created to simulate a novice and generates extended questions associated with each derivative question based on a preset first prompt. Then, a retrieval tool is invoked to search for the extended questions on the internet, obtaining reference content for the extended questions. Next, another large model instance j, i.e., an expert dialogue model, is created to simulate an expert and generates answers to the extended questions based on the retrieval tool's reference content and a preset second prompt. The answers to the extended questions and a preset third prompt are then input into the novice dialogue model for iterative looping until a preset number of loops or a loop termination condition is reached, resulting in an answer to at least one extended question. Finally, the answers to at least one extended question for each derivative question are compiled and summarized to obtain the answer to each derivative question.

[0076] For example, the specific workflow of each thread is as follows:

[0077] 1> Create a large model instance i, which is a novice dialogue model, to simulate a beginner. Based on a preset first prompt, it generates extended questions related to the derived questions. The prompt is as follows: You are a beginner about "question i" and are very eager to learn about this. Do you have any questions for the experts? It should be noted that the prompt needs to make the model understand that it is a novice in the knowledge domain i, lacks knowledge in this area, and wants to master the knowledge in this domain through multiple rounds of questioning. The system calls the large model instance i via an HTTP request, with the above prompt in the request body. Then, it waits for the large model to respond, parses the message returned by the large model, and extracts the generated beginner question from it. Let's assume that the beginner question generated by the large model is user_question_i_1.

[0078] 2> Call the search engine to retrieve reference content about user_question_i_1 from the Internet, and store the search results and source URLs in the database.

[0079] 3> Create another large model instance j, namely the expert dialogue model, to simulate an expert and generate answers to extended questions based on the retrieval tool's reference content and preset second prompt words. The prompt words are as follows: You are an expert in the field of "user_question_i". Please summarize the content I entered about this question. The prompt needs to make the model understand that it is an expert in the field, possesses rich knowledge in the field, and can give the questioner an accurate answer. The system calls the large model instance j via an HTTP request. The request body contains the above prompt words and the reference content retrieved from the Internet in step 2>. Then, it waits for the large model to respond, parses the message returned by the large model, extracts the expert-generated answer from it, assuming the large model outputs the answer "expert_answer_i_1", and writes the answer to the database.

[0080] 4> Call the expert's output, expert_answer_i_1, into the novice dialogue model with the following prompt: "This is the expert's answer to your question. Is there anything else you would like to know?" If no, the prompt should be "No, thank you." If the novice dialogue model continues to output the question user_question_i_2, repeat steps 2>3>4>.

[0081] By constructing a multi-turn dialogue loop through the above 4 steps, you can set the number of loop turns to k. The dialogue ends after reaching k turns, or if the novice dialogue model outputs "No more, thank you", it is also considered that the dialogue has ended.

[0082] Then, all expert responses (expert_answer) for "Question i" are read from the database, and the large model is invoked to summarize them, marking the source of each sentence. Let's assume the response is Answer 1. The prompt is as follows: You are very proficient in Question i. Below are some answers related to Question i. Please summarize them, avoiding repetition but also ensuring no information is omitted. The system invokes the large model instance via an HTTP request. The request body contains the above prompt and all expert responses (expert_answer_i_1-expert_answer_i_k) stored in the database for Question i. After waiting for the large model's response, the system parses the returned message, extracts the final expert summary, and caches this answer in the database.

[0083] As a preferred embodiment, the step of outputting and displaying the answer to the original question and the answers to multiple derivative questions includes:

[0084] The answer to the original question, the multiple derivative questions associated with the original question, and the answer to each of the derivative questions are output to the user terminal for visualization display according to a preset format.

[0085] Specifically, in this embodiment of the invention, after all dialogue processes have ended, the answers to the original questions, the derived questions, and the answers to the derived questions are organized and output to the user terminal in a preset format, such as a hierarchical format, so that the user can intuitively understand and use this information.

[0086] As a preferred embodiment, the step of outputting and displaying the answer to the original question and the answers to multiple derivative questions includes:

[0087] The answer to the original question, the multiple knowledge points associated with the original question, the multiple derivative questions corresponding to each knowledge point, and the answer to each derivative question are output to the user terminal for visualization display according to a preset format.

[0088] Specifically, in this embodiment of the invention, after all dialogue processes have concluded, the answer to the original question, knowledge points, derivative questions, and answers to derivative questions are organized and output to the user terminal according to a preset format, such as a hierarchical structure, so that the user can intuitively understand and use this information. For example, the output format is as follows:

[0089] User Issues

[0090] Direct answer

[0091] Related knowledge compilation:

[0092] Knowledge Point 1

[0093] Question 1

[0094] Answer 1

[0095] Question 2

[0096] Answer 2 ......

[0098] Question m

[0099] Answer m

[0100] Knowledge Point 2

[0101] Question 1

[0102] Answer 1 ......

[0104] Accordingly, the present invention also provides an intelligent question-answering system based on a large model, which can implement all the processes of the intelligent question-answering method based on a large model in the above embodiments.

[0105] Please see Figure 3 , Figure 3 This is a schematic diagram of a preferred embodiment of an intelligent question-answering system based on a large model provided by the present invention. The intelligent question-answering system based on a large model includes:

[0106] Module 301 is used to retrieve the original problem;

[0107] The generation module 302 is used to generate the answer to the original question and multiple derivative questions associated with the original question based on the large model.

[0108] Dialogue module 303 is used to input the multiple derivative questions into a preset multi-role dialogue model to obtain the answer to each of the derivative questions output by the multi-role dialogue model;

[0109] The output module 304 is used to output and display the answer to the original question and the answers to multiple derivative questions.

[0110] Preferably, the generation module 302 is specifically used for:

[0111] The answer to the original question is generated based on the large model;

[0112] Based on the large model, multiple knowledge points associated with the original problem are generated;

[0113] Based on the large model, multiple derivative questions are generated for each of the aforementioned knowledge points.

[0114] Preferably, the multi-role dialogue model includes a first-role dialogue model, a retrieval system, and a second-role dialogue model; wherein the first-role dialogue model and the second-role dialogue model are both built based on the large model;

[0115] The first role dialogue model is used to generate at least one primary extension question based on the derived question;

[0116] The search engine is used to retrieve relevant reference content based on the extended question;

[0117] The second role dialogue model is used to generate corresponding answers based on the extended questions.

[0118] Preferably, the dialogue module 303 is specifically used for:

[0119] The multiple derivative questions are input into the multi-role dialogue model;

[0120] Based on the first character dialogue model and the preset first prompt words, generate extended questions associated with each of the derived questions;

[0121] The extended question is retrieved using the retrieval tool to obtain reference content for the extended question;

[0122] The answer to the extended question is generated based on the second role dialogue model, the reference content, and the preset second prompt words;

[0123] The answer to the extended question and the preset third prompt are input into the first character dialogue model for iterative looping until the preset number of loops or the loop stopping condition is reached, so as to obtain the answer to at least one extended question.

[0124] The answers to at least one extended question of each of the derived questions are aggregated to obtain the answer to each of the derived questions.

[0125] Preferably, the step of inputting the answer to the extended question and a preset third prompt word into the first role dialogue model for iterative looping until a preset number of loops or a loop stopping condition is reached, to obtain an answer to at least one extended question, includes:

[0126] The answer to the extended question and the preset third prompt word are input into the first character dialogue model;

[0127] If the first character dialogue model generates a new extended question, then return to the step of retrieving the extended question according to the retrieval tool to obtain the reference content of the extended question, until the preset number of loop rounds or the loop stopping condition is reached, and obtain the answers to multiple extended questions;

[0128] If the first character's dialogue model does not generate a new extended question, the dialogue ends, and an answer to an extended question is obtained.

[0129] Preferably, the output module 304 is specifically used for:

[0130] The answer to the original question, the multiple derivative questions associated with the original question, and the answer to each of the derivative questions are output to the user terminal for visualization display according to a preset format.

[0131] Preferably, the output module 304 is specifically used for:

[0132] The answer to the original question, the multiple knowledge points associated with the original question, the multiple derivative questions corresponding to each knowledge point, and the answer to each derivative question are output to the user terminal for visualization display according to a preset format.

[0133] In specific implementation, the working principle, control process and technical effects of the intelligent question answering system based on a large model provided in this embodiment of the invention are the same as those of the intelligent question answering method based on a large model in the above embodiments, and will not be repeated here.

[0134] Please see Figure 4 , Figure 4 This is a schematic diagram of a preferred embodiment of a terminal device provided by the present invention. The terminal device includes a processor 401, a memory 402, and a computer program stored in the memory 402 and configured to be executed by the processor 401. When the processor 401 executes the computer program, it implements the intelligent question-answering method based on a large model as described in any of the above embodiments.

[0135] Preferably, the computer program can be divided into one or more modules / units (such as computer program 1, computer program 2, ...), and the one or more modules / units are stored in the memory 402 and executed by the processor 401 to complete the present invention. The one or more modules / units can be a series of computer program instruction segments capable of performing specific functions, and the instruction segments are used to describe the execution process of the computer program in the terminal device.

[0136] The processor 401 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor, or the processor 401 may be any conventional processor. The processor 401 is the control center of the terminal device, connecting various parts of the terminal device through various interfaces and lines.

[0137] The memory 402 mainly includes a program storage area and a data storage area. The program storage area can store the operating system, applications required for at least one function, etc., while the data storage area can store related data, etc. Furthermore, the memory 402 can be a high-speed random access memory, or a non-volatile memory, such as a plug-in hard disk, a smart media card (SMC), a secure digital card (SD), and a flash card, or it can be other volatile solid-state storage devices.

[0138] It should be noted that the aforementioned terminal devices may include, but are not limited to, processors and memory, as will be understood by those skilled in the art. Figure 4 The structural diagram is merely an example of the terminal device described above and does not constitute a limitation on the terminal device described above. It may include more or fewer components than shown in the diagram, or combine certain components, or use different components.

[0139] This invention also provides a computer-readable storage medium comprising a stored computer program, wherein the computer program, when running, controls the device where the computer-readable storage medium is located to execute the intelligent question-answering method based on a large model as described in any of the above embodiments.

[0140] This invention also provides a computer program product, which includes a computer program or computer instructions. When the computer program or computer instructions are executed by a processor, they implement the intelligent question-answering method based on a large model as described in any of the above embodiments.

[0141] This invention provides an intelligent question-answering method, system, device, medium, and product based on a large model. The method involves: acquiring an original question; generating an answer to the original question and multiple derivative questions associated with the original question based on the large model; inputting the multiple derivative questions into a preset multi-role dialogue model to obtain the answer to each of the derivative questions output by the multi-role dialogue model; and outputting and displaying the answer to the original question and the answers to the multiple derivative questions. This invention utilizes a large model to expand a user-input question into multiple questions and generate corresponding answers, outputting them to the user all at once. This effectively reduces the cost of multiple question-asking for the user, improves the comprehensiveness and accuracy of question-answering, and enhances the user experience.

[0142] It should be noted that the system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the system embodiments provided by this invention, the connection relationships between modules indicate that they have communication connections, which can be specifically implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.

[0143] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A smart question-answering method based on a large model, characterized in that, include: Get the original question; The answer to the original question is generated based on the large model, as well as multiple derivative questions associated with the original question; The multiple derivative questions are input into a preset multi-role dialogue model to obtain the answer to each of the derivative questions output by the multi-role dialogue model. The answer to the original question, as well as the answers to multiple derivative questions, will be output and displayed.

2. The intelligent question-answering method based on a large model as described in claim 1, characterized in that, The process of generating the answer to the original question based on the large model, and multiple derivative questions associated with the original question, includes: The answer to the original question is generated based on the large model; Based on the large model, multiple knowledge points associated with the original problem are generated; Based on the large model, multiple derivative questions are generated for each of the aforementioned knowledge points.

3. The intelligent question-answering method based on a large model as described in claim 1, characterized in that, The multi-role dialogue model includes a first-role dialogue model, a retrieval system, and a second-role dialogue model; wherein, both the first-role dialogue model and the second-role dialogue model are built based on the larger model. The first role dialogue model is used to generate at least one primary extension question based on the derived question; The search engine is used to retrieve relevant reference content based on the extended question; The second role dialogue model is used to generate corresponding answers based on the extended questions.

4. The intelligent question-answering method based on a large model as described in claim 3, characterized in that, The step of inputting the multiple derivative questions into a preset multi-role dialogue model to obtain the answer to each of the derivative questions output by the multi-role dialogue model includes: The multiple derivative questions are input into the multi-role dialogue model; Based on the first character dialogue model and the preset first prompt words, generate extended questions associated with each of the derived questions; The extended question is retrieved using the retrieval tool to obtain reference content for the extended question; The answer to the extended question is generated based on the second role dialogue model, the reference content, and the preset second prompt words; The answer to the extended question and the preset third prompt are input into the first character dialogue model for iterative looping until the preset number of loops or the loop stopping condition is reached, so as to obtain the answer to at least one extended question. The answers to at least one extended question of each of the derived questions are aggregated to obtain the answer to each of the derived questions.

5. The intelligent question-answering method based on a large model as described in claim 4, characterized in that, The step of inputting the answer to the extended question and a preset third prompt word into the first character dialogue model for iterative looping until a preset number of loop rounds or a loop stopping condition is reached, to obtain an answer to at least one extended question, includes: The answer to the extended question and the preset third prompt word are input into the first character dialogue model; If the first character dialogue model generates a new extended question, then return to the step of retrieving the extended question according to the retrieval tool to obtain the reference content of the extended question, until the preset number of loop rounds or the loop stopping condition is reached, and obtain the answers to multiple extended questions; If the first character's dialogue model does not generate a new extended question, the dialogue ends, and an answer to an extended question is obtained.

6. The intelligent question-answering method based on a large model as described in claim 1, characterized in that, The step of outputting and displaying the answer to the original question and the answers to multiple derivative questions includes: The answer to the original question, the multiple derivative questions associated with the original question, and the answer to each of the derivative questions are output to the user terminal for visualization display according to a preset format.

7. The intelligent question-answering method based on a large model as described in claim 2, characterized in that, The step of outputting and displaying the answer to the original question and the answers to multiple derivative questions includes: The answer to the original question, the multiple knowledge points associated with the original question, the multiple derivative questions corresponding to each knowledge point, and the answer to each derivative question are output to the user terminal for visualization display according to a preset format.

8. An intelligent question-answering system based on a large model, characterized in that, include: The retrieval module is used to retrieve the original question; A generation module is used to generate the answer to the original question and multiple derivative questions associated with the original question based on the large model; The dialogue module is used to input the multiple derivative questions into a preset multi-role dialogue model and obtain the answer to each of the derivative questions output by the multi-role dialogue model. The output module is used to output and display the answer to the original question and the answers to multiple derivative questions.

9. A terminal device, characterized in that, The system includes a processor and a memory, wherein the memory stores a computer program and the computer program is configured to be executed by the processor, wherein the processor, when executing the computer program, implements the intelligent question-answering method based on a large model as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, wherein when the device containing the computer-readable storage medium executes the computer program, it implements the intelligent question-answering method based on a large model as described in any one of claims 1 to 7.

11. A computer program product, characterized in that, The computer program product includes a computer program or computer instructions, which, when executed by a processor, implement the intelligent question-answering method based on a large model as described in any one of claims 1 to 7.

Images