Question and answer consultation method, system and equipment based on large language model and medium

Abstract

The invention discloses a consultation question and answer method, system and device based on a large language model and a medium, and belongs to the technical field of large language models. The method comprises the following steps: generating a large-model SDK development kit, integrating the large-model SDK development kit into a project background, and docking the project background with a service system; generating a server push event SSE request according to the consultation question proposed by the user; according to the SSE request, a consultation question proposed by the user is answered, according to the SSE streaming response, an answering result is packaged into a consultation question and answer result, and the consultation question and answer result is fed back to a user side; and performing de-encapsulation processing on the consultation question-answering result, generating a consultation question-answering result of the consultation question proposed by the user, and displaying the consultation question-answering result to the user. According to the invention, the connection between the large language model and the business system can be realized, so that the user question can be efficiently answered based on the business system.

Description

Technical Field

[0001] The present invention relates to the technical field of large language models, and more specifically, to a consultation and question - answering method, system, device, and medium based on large language models. Background Art

[0002] Large models are reshaping all industries. The large model technology will surely bring deep - level technological changes to numerous industries, reconstruct the original product forms, service models, and work processes of each industry, promote the innovative intelligent upgrading of each industry, make the entire business chain move towards high - efficiency, automation, and intelligence, and the large model market pattern will also be reshuffled.

[0003] With the development of social economy, the demand for the use of large models is also becoming wider. Large models are the key to realizing the high - efficiency of consultation and question - answering. By applying large model technology, a large amount of data and information can be processed, and at the same time, automated classification, induction, reasoning, and analysis can be carried out, so as to quickly provide accurate support and solutions for enterprises and individuals. At the same time, by applying large model technology, the intelligence of question - answering can be realized, automatically answering simple questions, recommending relevant materials, giving risk warnings, etc., improving the intelligence level of question - answering and enhancing the customer experience.

[0004] How to quickly connect to large models in different fields, such as quickly accessing large models to realize the intelligent empowerment of existing businesses, has become an urgent need for more and more software development companies and products. Summary of the Invention

[0005] In view of the above problems, the present invention proposes a consultation and question - answering method based on large language models, including:

[0006] Encapsulate the large language model to generate a large model SDK development toolkit, integrate the large model SDK development toolkit into the project background, and dock the project background with the business system;

[0007] On the user side, when a user asks a consultation question to the business system, through the business system, according to the consultation question proposed by the user, generate a Server - Sent Events (SSE) request, and forward the SSE request to the project background;

[0008] Through the project background, according to the SSE request, answer the consultation question proposed by the user, and return an SSE streaming response to the business system based on the answer result. Through the business system, according to the SSE streaming response, encapsulate the answer result into a consultation and question - answering result, and feedback the consultation and question - answering result to the user side;

[0009] On the user side, the consultation Q&A result is unpacked to generate the consultation Q&A result of the consultation question raised by the user, and the consultation Q&A result is presented to the user.

[0010] Optionally, the large model SDK development toolkit includes:

[0011] An SDK development toolkit with SSE call capabilities.

[0012] Optionally, integrating the large model SDK development toolkit into the project background includes:

[0013] Using Java to develop a Spring Boot background project, and based on the developed Spring Boot background project, integrating the large model SDK development toolkit into the project background.

[0014] Optionally, docking the project background with the business system includes:

[0015] In the project background, write an SSE asynchronous call interface and dock with the business system through the SSE asynchronous call interface.

[0016] Optionally, after the project background and the business system are docked, the large model SDK development toolkit in the project background is called through the business system. After the large model SDK development toolkit gives a consultation Q&A result to the consultation question raised by the user, based on the SSE asynchronous call interface, an SSE streaming response is generated according to the consultation Q&A result.

[0017] Optionally, the method further includes:

[0018] Establish a front-end project for the business system, and based on the front-end project, build a dialogue interaction box and a Q&A interface on the user side;

[0019] Among them, the consultation question of the user is received through the dialogue interaction box, and the consultation Q&A result is presented to the user through the Q&A interface.

[0020] Optionally, packaging the answer result into a consultation Q&A result according to the SSE streaming response includes:

[0021] According to the SSE streaming response, package the answer result into a consultation Q&A result in json message format and markdown format;

[0022] The consultation Q&A result in json message format is presented as a response message;

[0023] The consultation Q&A result in markdown format is presented as the text of the answer result.

[0024] Optionally, a typewriter effect and a streaming response effect are used to display the text of the answer result.

[0025] Optionally, the consultation and answer results are fed back to the user in a batch asynchronous manner.

[0026] On the other hand, the present invention also proposes a consultation and answer system based on a large language model, including:

[0027] A docking unit for encapsulating the large language model to generate a large model SDK development toolkit, integrating the large model SDK development toolkit into the project background, and docking the project background with the business system;

[0028] A request unit for, on the user side, when the user asks a consultation question to the business system, generating a Server-Sent Events (SSE) request according to the consultation question asked by the user through the business system, and forwarding the SSE request to the project background;

[0029] A question and answer unit for, through the project background, answering the consultation question asked by the user according to the SSE request, and returning an SSE streaming response to the business system based on the answer result, and encapsulating the answer result as a consultation and answer result through the business system according to the SSE streaming response, and feeding back the consultation and answer result to the user side;

[0030] An output unit for, on the user side, performing a de-encapsulation process on the consultation and answer result, generating a consultation and answer result for the consultation question asked by the user, and displaying the consultation and answer result to the user.

[0031] Optionally, the large model SDK development toolkit includes:

[0032] An SDK development toolkit with SSE call capabilities.

[0033] Optionally, integrating the large model SDK development toolkit into the project background includes:

[0034] Using Java to develop a Spring Boot background project, and integrating the large model SDK development toolkit into the project background based on the developed Spring Boot background project.

[0035] Optionally, docking the project background with the business system includes:

[0036] Writing an SSE asynchronous call interface in the project background, and docking with the business system through the SSE asynchronous call interface.

[0037] Optionally, after the project background is connected to the business system, the large model SDK development toolkit in the project background is called through the business system. After the large model SDK development toolkit gives an advisory Q&A result to the consultation question raised by the user, based on the SSE asynchronous call interface, a SSE streaming response is generated according to the consultation Q&A result.

[0038] Optionally, the system further includes: an interface unit for establishing a front-end project of the business system. Based on the front-end project, a dialogue interaction box on the user side is built in the request unit, and a Q&A interface on the user side is built in the output unit;

[0039] Among them, the consultation question of the user is received through the dialogue interaction box, and the consultation Q&A result is displayed to the user through the Q&A interface.

[0040] Optionally, according to the SSE streaming response, the solution result is encapsulated into a consultation Q&A result, including:

[0041] According to the SSE streaming response, the solution result is encapsulated into a consultation Q&A result in json message format and markdown format;

[0042] The consultation Q&A result in json message format is displayed as a response message;

[0043] The consultation Q&A result in markdown format is displayed as the text of the solution result.

[0044] Optionally, a typewriter effect and a streaming response effect are used to display the text of the solution result.

[0045] Optionally, the consultation Q&A result is fed back to the user in a batch asynchronous manner.

[0046] On the other hand, the present invention also provides a computing device, including: one or more processors;

[0047] The processor is used to execute one or more programs;

[0048] When the one or more programs are executed by the one or more processors, the method as described above is implemented.

[0049] On the other hand, the present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed, the method as described above is implemented.

[0050] Compared with the prior art, the beneficial effects of the present invention are:

[0051] The present invention provides a consultation and question-answering method based on a large language model, including: encapsulating the large language model to generate a large model SDK development toolkit, integrating the large model SDK development toolkit into the project background, and docking the project background with the business system; on the user side, when the user asks a consultation question to the business system, through the business system, according to the consultation question raised by the user, generate a Server-Sent Events (SSE) request, and forward the SSE request to the project background; through the project background, according to the SSE request, answer the consultation question raised by the user, and return an SSE streaming response to the business system based on the answer result, through the business system, according to the SSE streaming response, encapsulate the answer result into a consultation and question-answering result, and feedback the consultation and question-answering result to the user side; on the user side, perform a de-encapsulation process on the consultation and question-answering result to generate a consultation and question-answering result for the consultation question raised by the user, and display the consultation and question-answering result to the user. The present invention can realize the docking of the large language model with the business system, so that the user's questions can be efficiently answered based on the business system. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 is a flowchart of the method of the present invention;

[0053] Figure 2 is a schematic diagram of the method of the present invention;

[0054] Figure 3 is a structural diagram of the system of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0055] Now refer to the drawings to introduce the exemplary embodiments of the present invention. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. These embodiments are provided to disclose the present invention in detail and completely, and to fully convey the scope of the present invention to those skilled in the art. The terms in the exemplary embodiments shown in the drawings are not limitations on the present invention. In the drawings, the same units / components use the same reference numerals.

[0056] Unless otherwise specified, the terms (including scientific and technical terms) used herein have the ordinary meaning understood by those skilled in the art. In addition, it can be understood that the terms defined in the commonly used dictionary should be understood to have a meaning consistent with the context of their related fields, and should not be understood as idealized or overly formal meanings.

[0057] Embodiment 1:

[0058] The present invention also proposes a consultation and question-answering method based on a large language model, as Figure 1 shown, including:

[0059] Step 1: Package the large language model to generate a large model SDK development kit, integrate the large model SDK development kit into the project background, and dock the project background with the business system;

[0060] Step 2: On the user side, after the user asks a consulting question to the business system, through the business system, generate a Server-Sent Events (SSE) request based on the consulting question asked by the user, and forward the SSE request to the project background;

[0061] Step 3: Through the project background, answer the consulting question asked by the user according to the SSE request, and return an SSE streaming response to the business system based on the answer result. Through the business system, package the answer result into a consulting Q&A result according to the SSE streaming response, and feedback the consulting Q&A result to the user side;

[0062] Step 4: On the user side, perform unpacking processing on the consulting Q&A result to generate the consulting Q&A result of the consulting question asked by the user, and display the consulting Q&A result to the user.

[0063] Among them, the large model SDK development kit includes:

[0064] An SDK development kit with SSE call capabilities.

[0065] Among them, integrating the large model SDK development kit into the project background includes:

[0066] Use Java to develop a Spring Boot background project, and based on the developed Spring Boot background project, integrate the large model SDK development kit into the project background.

[0067] Among them, docking the project background with the business system includes:

[0068] In the project background, write an SSE asynchronous call interface, and dock with the business system through the SSE asynchronous call interface.

[0069] Among them, after the project background and the business system are docked, the large model SDK development kit in the project background is called through the business system. After the large model SDK development kit gives a consulting Q&A result to the consulting question asked by the user, based on the SSE asynchronous call interface, an SSE streaming response is generated according to the consulting Q&A result.

[0070] Among them, the method further includes:

[0071] Build the front-end project of the business system. Based on the front-end project, build a dialogue interaction box and a question-and-answer interface on the user side;

[0072] Among them, receive the user's consultation questions through the dialogue interaction box, and display the consultation question-and-answer results to the user through the question-and-answer interface.

[0073] Among them, according to the SSE streaming response, encapsulate the answer result into a consultation question-and-answer result, including:

[0074] According to the SSE streaming response, encapsulate the answer result into a consultation question-and-answer result in json message format and markdown format;

[0075] The consultation question-and-answer result in json message format is displayed as a response message;

[0076] The consultation question-and-answer result in markdown format is displayed as the text of the answer result.

[0077] Among them, use the typewriter effect and the streaming response effect to display the text of the answer result.

[0078] Among them, feedback the consultation question-and-answer results to the user in a batch asynchronous manner.

[0079] Among them, the schematic diagram of the present invention is as Figure 2 shown, combined with Figure 2 , further illustrate the present invention, as Figure 2 shown, the main steps of the present invention specifically include the following:

[0080] Encapsulate the capabilities of the large model into an SDK development toolkit that supports SSE calls. At the same time, encapsulate the return results in json format and markdown format. Encapsulate the self-developed large model into a large model SDK development toolkit. Hereinafter, the "large model SDK development toolkit" is simply referred to as "large model SDK".

[0081] Build the project background, use Java to develop a Spring Boot background project, and integrate the large model SDK into the background project.

[0082] In the background project engineering, write an interface that supports SSE asynchronous calls. For example, develop a question-and-answer interface. The interface should support SSE and asynchronous requests. The background business layer calls the large model SDK to make the interface have the streaming response effect of SSE.

[0083] Use Vue to build the front-end project, develop the corresponding business pages, develop the dialogue interaction box, call the question-and-answer interface provided by the background business system, and display the results of the SSE streaming response and the streaming response effect.

[0084] The front - end page sends the content message of the question to the business background through the Q&A interface of SSE. The background triggers the call of the large - model SDK. The large - model SDK returns the result to the background, and then the background responds to the front - end page. The front - end page parses the message in json format and displays the result in markdown format on the page for the user to view.

[0085] The key points are as follows:

[0086] The functions of the large - model are encapsulated into an SDK, separating the development of the large - model from the development of business applications, simplifying the development process of artificial intelligence GPT. The final result output by the research and development of the large - model is the "large - model SDK", which can be provided for business systems to use.

[0087] The business system integrates the large - model SDK and focuses on developing business and processes. Both sides can develop, update, and maintain independently.

[0088] "SSE" refers to "Server - Sent Events", which is a technology that enables the server to send real - time updates to the browser. SSE allows the server - side to actively push data to the client without the client having to frequently send requests to the server. This is very useful for application scenarios that require real - time data updates.

[0089] The large - model SDK and the background business system use SSE as the way of remote call. Compared with traditional http calls, the result content can be sent to the caller in batches asynchronously, with a faster response speed and a smooth real - time user experience.

[0090] The front - end and the background business use SSE as the way of remote call. Compared with traditional http calls, it is similar to the http call method, with less development difficulty. At the same time, a ciphertext request header can be added to the request header to achieve the functions of message confidentiality and authentication. One SSE request from the front - end can obtain multiple consecutive responses from the back - end, which matches the asynchronous response characteristics of the large - model.

[0091] The json message format is used as the end - to - end response message format, and at the same time, the large - model processing results are assembled using markdown syntax. The json format is suitable for network transmission and parsing. The markdown format is suitable for text display. The markdown format is more suitable for streaming transmission and partial transmission, and the front - end can also show good results when it gets part of the content. The front - end page is convenient for display and has a strong user experience.

[0092] Embodiment 2:

[0093] On the other hand, the present invention also proposes a consultation Q&A system 200 based on a large - language model, asFigure 3 As shown in the figure, it includes:

[0094] A docking unit 201, which is used to encapsulate a large language model to generate a large model SDK development toolkit, integrate the large model SDK development toolkit into the project background, and dock the project background with the business system;

[0095] A request unit 202, which is used on the user side. When a user asks a consultation question to the business system, through the business system, according to the consultation question raised by the user, generate a Server-Sent Events (SSE) request and forward the SSE request to the project background;

[0096] A Q&A unit 203, which is used through the project background. According to the SSE request, answer the consultation question raised by the user, and based on the answer result, return an SSE streaming response to the business system. Through the business system, according to the SSE streaming response, encapsulate the answer result into a consultation Q&A result and feedback the consultation Q&A result to the user side;

[0097] An output unit 204, which is used on the user side to perform de-encapsulation processing on the consultation Q&A result, generate the consultation Q&A result of the consultation question raised by the user, and display the consultation Q&A result to the user.

[0098] Among them, the large model SDK development toolkit includes:

[0099] An SDK development toolkit with SSE call capabilities.

[0100] Among them, integrating the large model SDK development toolkit into the project background includes:

[0101] Use Java to develop a Spring Boot background project. Based on the developed Spring Boot background project, integrate the large model SDK development toolkit into the project background.

[0102] Among them, docking the project background with the business system includes:

[0103] In the project background, write an SSE asynchronous call interface and dock with the business system through the SSE asynchronous call interface.

[0104] Among them, after the project background and the business system are docked, the large model SDK development toolkit in the project background is called through the business system. After the large model SDK development toolkit gives a consultation Q&A result to the consultation question raised by the user, based on the SSE asynchronous call interface, according to the consultation Q&A result, generate an SSE streaming response.

[0105] Among them, the system further includes: an interface unit 205, which is used to establish a front-end project of the business system. Based on the front-end project, in the request unit 202, a dialogue interaction box on the user side is built, and in the output unit 204, a question-and-answer interface on the user side is built;

[0106] Among them, the consultation questions of the user are received through the dialogue interaction box, and through the question-and-answer interface, the consultation question-and-answer results are displayed to the user.

[0107] Among them, according to the SSE streaming response, the solution result is encapsulated into a consultation question-and-answer result, including:

[0108] According to the SSE streaming response, the solution result is encapsulated into a consultation question-and-answer result in json message format and markdown format;

[0109] The consultation question-and-answer result in json message format is displayed as a response message;

[0110] The consultation question-and-answer result in markdown format is displayed as the text of the solution result.

[0111] Among them, a typewriter effect and a streaming response effect are used to display the text of the solution result.

[0112] Among them, the consultation question-and-answer results are fed back to the user in a batch asynchronous manner.

[0113] The present invention can realize the docking of the large language model and the business system, so that the user's questions can be efficiently answered based on the business system.

[0114] Embodiment 3:

[0115] Based on the same inventive concept, the present invention further provides a computer device, which includes a processor and a memory. The memory is used to store a computer program, and the computer program includes program instructions. The processor is used to execute the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or may also be 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. It is the computing core and control core of the terminal, and is suitable for implementing one or more instructions. Specifically, it is suitable for loading and executing one or more instructions in the computer storage medium to implement the corresponding method flow or corresponding function, so as to implement the steps of the method in the above embodiments.

[0116] Embodiment 4:

[0117] Based on the same inventive concept, the present invention further provides a storage medium, specifically a computer-readable storage medium (Memory). The computer-readable storage medium is a memory device in the computer device and is used to store programs and data. It can be understood that the computer-readable storage medium here can include both the built-in storage medium in the computer device and, of course, the extended storage medium supported by the computer device. The computer-readable storage medium provides a storage space, and this storage space stores the operating system of the terminal. And, one or more instructions suitable for being loaded and executed by the processor are also stored in this storage space. These instructions can be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk memory. The one or more instructions stored in the computer-readable storage medium can be loaded and executed by the processor to implement the steps of the method in the above embodiments.

[0118] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memory, CD-ROM, optical memory, etc.) containing computer-usable program code. The solutions in the embodiments of the present invention can be implemented in various computer languages. For example, object-oriented programming languages such as Java and interpreted scripting languages such as JavaScript, etc.

[0119] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present invention. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, as well as the combination of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices generate means for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0120] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including instruction means that implement the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0121] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process. Thus, the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0122] Although the preferred embodiments of the present invention have been described, those skilled in the art can make additional changes and modifications once they know the basic creative concept. Therefore, the appended claims are intended to be construed to include the preferred embodiments as well as all changes and modifications falling within the scope of the present invention.

[0123] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention also intends to include these changes and modifications.

Claims

1. A consultation question-answering method based on a large language model, characterized in that: include: Encapsulate the large language model, generate a large model SDK development toolkit, integrate the large model SDK development toolkit into the project background, and connect the project background with the business system; On the user side, when the user asks a consultation question to the business system, the business system generates a server push event SSE request based on the consultation question raised by the user, and forwards the SSE request to the project background; Through the project background, according to the SSE request, the consulting question raised by the user is answered, and based on the answer result, an SSE streaming response is returned to the business system. Through the business system, according to the SSE streaming response, the answer result is encapsulated as a consulting question and answer result, and the consulting question and answer result is fed back to the user side; On the user side, the consultation question and answer result is decapsulated to generate a consultation question and answer result of the consultation question raised by the user, and the consultation question and answer result is displayed to the user.

2. The consultation question-and-answer method according to claim 1, characterized in that: The large model SDK development toolkit includes: SDK development toolkit with SSE calling capabilities.

3. The consultation question-and-answer method according to claim 1, characterized in that: Integrating the large model SDK development toolkit into the project background includes: Use Java to develop the Spring Boot background project, and based on the developed Spring Boot background project, integrate the large model SDK development toolkit into the project background.

4. The consultation question-and-answer method according to claim 1, characterized in that: The connecting the project backend with the business system includes: In the project background, write an SSE asynchronous call interface, and connect with the business system through the SSE asynchronous call interface.

5. The inquiry question and answer method according to claim 4, characterized in that: After the project background and the business system are connected, the big model SDK development toolkit in the project background is called through the business system. After the big model SDK development toolkit gives the consultation question and answer results to the consultation questions raised by the user, it generates an SSE streaming response based on the SSE asynchronous call interface and the consultation question and answer results.

6. The inquiry question and answer method according to claim 1, characterized in that: The method further comprises: Establish a front-end project for the business system, and build a user-side dialogue interaction box and question-and-answer interface based on the front-end project; The user's consultation questions are received through a dialogue interaction box, and the consultation question and answer results are displayed to the user through a question and answer interface.

7. The consultation question-and-answer method according to claim 1, characterized in that: The step of encapsulating the answer result into a consultation question and answer result according to the SSE streaming response includes: According to the SSE streaming response, the answer result is encapsulated into a consultation question and answer result in json message format and markdown format; The consultation question and answer result in the json message format is displayed as a response message; The consultation question and answer results in the markdown format are displayed as the text of the answer results.

8. A consultation question-answering system based on a large language model, characterized in that: include: A docking unit, used to encapsulate the large language model, generate a large model SDK development toolkit, integrate the large model SDK development toolkit into the project background, and dock the project background with the business system; A request unit is used to generate a server push event SSE request through the business system according to the consulting question raised by the user on the user side, and forward the SSE request to the project background; The question-and-answer unit is used to answer the consulting questions raised by the user through the project background according to the SSE request, and return an SSE streaming response to the business system based on the answer result. Through the business system, according to the SSE streaming response, the answer result is encapsulated as a consulting question-and-answer result, and the consulting question-and-answer result is fed back to the user side; The output unit is used to decapsulate the consultation question and answer result on the user side, generate the consultation question and answer result of the consultation question raised by the user, and display the consultation question and answer result to the user.

9. A computer device, characterized in that: include: one or more processors; a processor for executing one or more programs; When the one or more programs are executed by the one or more processors, the method according to any one of claims 1 to 7 is implemented.

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

Images