Knowledge base search enhancement method and device, electronic equipment and storage medium

Abstract

The invention relates to a knowledge base search enhancement method and device, electronic equipment and a storage medium, and the method comprises the steps: carrying out preliminary retrieval enhancement generation matching from a knowledge base according to an input question, and obtaining at least one example question matched with the input question; acquiring description information of a first knowledge base to which the example question belongs; and inputting all the example questions and all the description information into a large model for analysis, and determining a knowledge base number corresponding to the input question, so as to query a search result corresponding to the input question in a target knowledge base corresponding to the knowledge base number. According to the method, preliminary matching can be carried out from a knowledge base by utilizing a retrieval enhancement generation technology according to an input problem to obtain a matched example problem, description information of a first knowledge base corresponding to the example problem is obtained, and then a knowledge base number corresponding to the input problem is decided by utilizing a large model to comprehensively analyze the example problem and the description information; therefore, the retrieval matching error rate of the knowledge base can be reduced, and the user experience and the problem solving efficiency are improved.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and in particular, to a knowledge base search enhancement method, device, electronic device, and storage medium. Background Art

[0002] In modern operation and maintenance platforms, the knowledge base system plays an extremely important role. It stores a large amount of operation and maintenance knowledge and can solve various types of operation and maintenance problems. However, due to the large repetition in the questioning methods of various problems, the traditional Retrieval Augmented Generation (RAG) technology still has certain deficiencies in terms of matching degree, resulting in a relatively high matching error rate, which affects the user experience and the efficiency of problem-solving. Summary of the Invention

[0003] This application provides a knowledge base search enhancement method, device, electronic device, and storage medium to solve the problem of how to reduce the knowledge base retrieval matching error rate.

[0004] In a first aspect, this application provides a knowledge base search enhancement method, and the method includes:

[0005] Perform a preliminary retrieval augmented generation match from the knowledge base according to the input question to obtain at least one example question that matches the input question;

[0006] Obtain the description information of the first knowledge base to which the example question belongs;

[0007] Input all the example questions and all the description information into a large model for analysis to determine the knowledge base number corresponding to the input question, so as to query the search result corresponding to the input question in the target knowledge base corresponding to the knowledge base number.

[0008] Optionally, before performing a preliminary retrieval augmented generation match from the knowledge base according to the input question to obtain at least one example question that matches the input question, the method further includes:

[0009] Perform the following maintenance processing for each knowledge base respectively:

[0010] Configure description information for the knowledge base, where the description information is used to characterize the problem type that the knowledge base solves;

[0011] Configure example questions and reply examples of the example questions for the knowledge base.

[0012] Optionally, performing a preliminary retrieval augmented generation match from the knowledge base according to the input question to obtain at least one example question that matches the input question includes:

[0013] Obtain the input problem;

[0014] Perform preliminary retrieval enhancement generation matching from all knowledge bases according to the input problem to obtain all example problems with a matching degree higher than a preset matching degree for the input problem.

[0015] Optionally, performing preliminary retrieval enhancement generation matching from all knowledge bases according to the input problem to obtain all example problems with a matching degree higher than a preset matching degree for the input problem includes:

[0016] Extract the vector features of the input problem;

[0017] Perform retrieval enhancement generation vector matching from all knowledge bases according to the vector features to obtain target vector features with a matching degree higher than a preset vector matching degree for the vector features.

[0018] Determine the example problems corresponding to the target vector features.

[0019] Optionally, inputting all the example problems and all the description information into a large model for analysis to determine the knowledge base number corresponding to the input problem, including:

[0020] Determine the first prompt word according to the example problem;

[0021] Determine the second prompt word according to the description information;

[0022] Input all the first prompt words and all the second prompt words into the large model for analysis to determine the knowledge base number corresponding to the input problem.

[0023] Optionally, query the search result corresponding to the input problem in the target knowledge base corresponding to the knowledge base number, including:

[0024] Determine the target knowledge base corresponding to the knowledge base number;

[0025] Query the search result corresponding to the input problem in the target knowledge base through the large model.

[0026] Optionally, the description information includes at least one of text information, label information, and category information.

[0027] In a second aspect, the present application provides a knowledge base search enhancement device, and the device includes:

[0028] A preliminary matching module, configured to perform preliminary retrieval enhancement generation matching from a knowledge base according to an input problem to obtain at least one example problem that matches the input problem;

[0029] An acquisition module, configured to acquire the description information of the first knowledge base to which the example question belongs;

[0030] An enhanced matching module, configured to input all the example questions and all the description information into a large model for analysis, determine the knowledge base number corresponding to the input question, and query the search result corresponding to the input question in the target knowledge base corresponding to the knowledge base number.

[0031] In a third aspect, the present application provides an electronic device, including a processor, a communication interface, a memory, and a communication bus. Among them, the processor, the communication interface, and the memory complete mutual communication through the communication bus;

[0032] The memory is used to store a computer program;

[0033] The processor is configured to implement the knowledge base search enhancement method according to any one of the embodiments in the first aspect when executing the program stored on the memory.

[0034] In a fourth aspect, the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the knowledge base search enhancement method according to any one of the embodiments in the first aspect.

[0035] The above technical solutions provided by the embodiments of the present application have the following advantages compared with the prior art: The method provided by the embodiments of the present application performs preliminary retrieval enhancement generation matching from the knowledge base according to the input question to obtain at least one example question that matches the input question; acquires the description information of the first knowledge base to which the example question belongs; inputs all the example questions and all the description information into a large model for analysis, determines the knowledge base number corresponding to the input question, and queries the search result corresponding to the input question in the target knowledge base corresponding to the knowledge base number. This method can perform preliminary matching from the knowledge base using the retrieval enhancement generation technology according to the input question to obtain matching example questions, and acquire the description information of the first knowledge base corresponding to the example questions, and then use the large model to comprehensively analyze the example questions and the description information to decide the knowledge base number corresponding to the input question, thereby reducing the knowledge base retrieval matching error rate and improving the user experience and the efficiency of problem solving. Description of the Drawings

[0036] The drawings here are incorporated into the specification and constitute a part of this specification, showing embodiments consistent with the present application, and are used together with the specification to explain the principles of the present application.

[0037] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the accompanying drawings required for the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0038] One or more embodiments are exemplarily illustrated by the pictures in the corresponding accompanying drawings. These exemplary illustrations do not limit the embodiments. Elements with the same reference numerals in the drawings represent similar elements. Unless otherwise stated, the drawings in the figures do not constitute a proportional limitation.

[0039] Figure 1 It is a system architecture diagram of a method for enhancing knowledge base search provided by an embodiment of the present application;

[0040] Figure 2 It is a schematic flowchart of a method for enhancing knowledge base search provided by an embodiment of the present application;

[0041] Figure 3 It is a schematic flowchart of a method for enhancing knowledge base search provided by another embodiment of the present application;

[0042] Figure 4 It is a schematic structural diagram of a device for enhancing knowledge base search provided by an embodiment of the present application;

[0043] Figure 5 It is a schematic structural diagram of an electronic device provided by an embodiment of the present application. Detailed implementation manners

[0044] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are some, but not all, of the embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts fall within the scope of protection of the present application.

[0045] The following disclosure provides many different embodiments or examples for implementing different structures of the present application. To simplify the disclosure of the present application, the components and settings of specific examples are described below. Of course, they are only examples and are not intended to limit the present application. In addition, the present application may repeat reference numerals and / or letters in different examples. This repetition is for the purpose of simplification and clarity and does not itself indicate the relationship between the various embodiments and / or settings discussed.

[0046] To solve the technical problem of how to reduce the error rate of knowledge base retrieval matching in the prior art, the present application provides a knowledge base search enhancement method, device, electronic device and storage medium, which can reduce the error rate of knowledge base retrieval matching and improve the user experience and the efficiency of problem solving.

[0047] The first embodiment of the present application provides a knowledge base search enhancement method, which can be applied to a Figure 1 system architecture as shown. At least a client 101 and an operation and maintenance platform 102 are included in this system architecture, and a communication connection is established between the client 101 and the operation and maintenance platform 102.

[0048] This method can be applied to the operation and maintenance platform 102 in this system architecture. Among them, the client 101 can be an intelligent terminal such as a mobile phone, etc., or can also be devices such as a desktop computer, a tablet computer, a notebook, a supercomputer, etc. The operation and maintenance platform 102 internally includes a knowledge base system, and the knowledge base system can be established and store a large amount of operation and maintenance knowledge based on the RAG technology and the large model decision technology. The operation and maintenance platform 102 can be a local server, or can also be a cloud server, or can also be a server cluster.

[0049] Next, based on this system architecture, the knowledge base search enhancement method will be described in detail. As Figure 2 shown, the knowledge base search enhancement method includes:

[0050] Step 201, perform preliminary retrieval enhancement generation matching from the knowledge base according to the input question to obtain at least one example question that matches the input question.

[0051] The user can connect to the operation and maintenance platform through the client and submit a question to the operation and maintenance platform. The operation and maintenance platform can perform preliminary RAG matching from the knowledge base system according to the input question, and obtain example questions that match the input question through search. The number of obtained example questions is not limited, and can be one or multiple.

[0052] In one embodiment, before performing preliminary retrieval enhancement generation matching from the knowledge base according to the input question to obtain at least one example question that matches the input question, the method further includes: performing the following maintenance processing for each knowledge base respectively: configuring description information for the knowledge base, where the description information is used to characterize the problem type that the knowledge base solves; configuring example questions and reply examples of the example questions for the knowledge base.

[0053] In this embodiment, a knowledge base system can be constructed first. The knowledge base system includes multiple knowledge bases. For example, Knowledge Base A, Knowledge Base B, Knowledge Base C, Knowledge Base D, etc. Each knowledge base can be described in detail, and the description information of the knowledge base can be configured to clearly record the problems that the knowledge base can solve. Specifically, the description information can be in the form of text information, tag information, category information, etc., without limitation.

[0054] For example, Knowledge Base A can be described as "used to solve server connection problems", while Knowledge Base B can be described as "dealing with database performance optimization problems". At the same time, the user can also maintain relevant example questions and reply examples for each knowledge base, and the example questions and reply examples can be used for subsequent matching and searching processes.

[0055] For example, in Knowledge Base A, the following example questions may be included:

[0056] Example Question 1: "How to handle the situation where the server cannot be connected?"

[0057] Example Question 2: "Analysis of the reasons for server connection failure and solutions?"

[0058] While in Knowledge Base B, the following example questions may be included:

[0059] Example Question 1: "How to optimize database query performance?"

[0060] Example Question 2: "What are the reasons for slow database response and the solutions?"

[0061] In one embodiment, a preliminary retrieval enhancement generation matching is performed from the knowledge bases according to the input question to obtain at least one example question that matches the input question, including: obtaining the input question; performing a preliminary retrieval enhancement generation matching from all knowledge bases according to the input question to obtain all example questions whose matching degree with the input question is higher than a preset matching degree.

[0062] In this embodiment, a preliminary matching can be performed on the knowledge base system according to the user's input question using the RAG technology, and example questions with a high matching degree with the input question can be obtained through searching, such as all example questions whose matching degree with the input question is higher than a preset matching degree. For example, for the question "Why can't my server be connected?" input by the user, the system may preliminarily match the example question "How to handle the situation where the server cannot be connected?" in Knowledge Base A.

[0063] In one embodiment, a preliminary retrieval enhancement is performed from all knowledge bases according to the input question to generate a match, and all example questions with a matching degree higher than a preset matching degree with the input question are obtained, including: extracting the vector features of the input question; performing retrieval enhancement from all knowledge bases according to the vector features to generate a vector match, and obtaining target vector features with a matching degree higher than a preset vector matching degree with the vector features; and determining the example questions corresponding to the target vector features.

[0064] In this embodiment, when performing RAG matching, the vector features of the input question can be extracted first. Of course, the first vector features of the example questions are pre-extracted in the knowledge base system. Based on the vector features of the input question, a preliminary RAG matching based on vector features can be performed in the knowledge base system, and target vector features with a matching degree higher than the preset matching degree with the vector features of the input question can be obtained from the first vector features. Then, based on the target vector features, the corresponding example questions are determined, so that the example questions corresponding to the user's input question can be accurately matched, and the recognition accuracy can be improved.

[0065] Step 202: Obtain the description information of the first knowledge base to which the example questions belong.

[0066] Since each example question corresponds to a fixed knowledge base, the first knowledge base to which the example question belongs can be determined through the corresponding relationship between the example question and the knowledge base, so that the description information of the first knowledge base can be accurately obtained. These description information can help the knowledge base system better understand the content of the knowledge base behind the matched example questions. For example, for the preliminarily matched example question "How to handle the situation where the server cannot be connected?", its corresponding description information is "Used to solve server connection problems".

[0067] Step 203: Input all the example questions and all the description information into the large model for analysis, and determine the knowledge base number corresponding to the input question, so as to query the search result corresponding to the input question in the target knowledge base corresponding to the knowledge base number.

[0068] Finally, the knowledge base system can input the preliminary matching results, that is, the example questions, and the description information of the knowledge base corresponding to the example questions into the large model for comprehensive analysis. The large model deeply processes all the input information, and based on the understanding and analysis, decides the knowledge base number that best matches the input question, so as to query the search result corresponding to the input question in the target knowledge base corresponding to the knowledge base number. The search result is the answer to the input question.

[0069] This method can perform a preliminary match from the knowledge base using retrieval-augmented generation technology based on the input question, obtain the matching sample questions, and acquire the description information of the first knowledge base corresponding to the sample questions. Then, it uses a large model to comprehensively analyze the sample questions and the description information to determine the knowledge base number corresponding to the input question, thereby reducing the error rate of knowledge base retrieval matching and improving the user experience and the efficiency of problem-solving.

[0070] In one embodiment, all sample questions and all description information are input into the large model for analysis to determine the knowledge base number corresponding to the input question, including: determining the first prompt word according to the sample question; determining the second prompt word according to the description information; inputting all the first prompt words and all the second prompt words into the large model for analysis to determine the knowledge base number corresponding to the input question.

[0071] In this embodiment, the large model can be a trained general language large model, which can at least process text data and understand natural language. The large model can be guided by the prompt to perform analysis and processing. For example, first determine the first prompt word according to the sample question, determine the second prompt word according to the description information, and then guide the large model to perform analysis and processing according to all the first prompt words and the second prompt words, so as to accurately determine the knowledge base number corresponding to the input question. For example, the large model finally determines that the knowledge base number corresponding to the user's input question "Why can't my server connect?" is A.

[0072] In one embodiment, query the search result corresponding to the input question in the target knowledge base corresponding to the knowledge base number, including: determining the target knowledge base corresponding to the knowledge base number; querying the search result corresponding to the input question in the target knowledge base through the large model.

[0073] In this embodiment, determine the corresponding target knowledge base according to the knowledge base number. For example, determine that the target knowledge base corresponding to the number A is Knowledge Base A, then the search result corresponding to the input question can be queried in Knowledge Base A through the large model, that is, the answer to the input question, and the search result is returned to the client.

[0074] In a specific embodiment, the knowledge base search enhancement method is as Figure 3 , including:

[0075] Step S0, knowledge base construction.

[0076] Construct a knowledge base and complete the maintenance of the description information and sample questions of the knowledge base. The user makes a detailed description of each knowledge base, clearly recording the types of problems that the knowledge base can solve. The description information should include but not be limited to text descriptions, tags, and category information.

[0077] For example, knowledge base A can be described as "used to solve server connection problems", while knowledge base B can be described as "dealing with database performance optimization problems". The user maintains relevant question and answer examples for each knowledge base. These Q&A examples will be used in the subsequent matching and search processes.

[0078] For example, in knowledge base A, it may contain the following question examples:

[0079] Example question 1: "How to handle the situation where the server cannot be connected?"

[0080] Example question 2: "Analysis of the reasons for server connection failure and solutions?"

[0081] While in knowledge base B, it may contain the following question examples:

[0082] Example question 1: "How to optimize database query performance?"

[0083] Example question 2: "What are the reasons for slow database response and the solutions?"

[0084] Step S1, user question. The input question of the user can be obtained through the operation and maintenance platform.

[0085] Step S2, search tool.

[0086] When the user enters a question in the operation and maintenance platform, a preliminary RAG match is performed through the search engine.

[0087] Step S3, preliminary RAG match.

[0088] The system first uses RAG technology for preliminary matching by searching for example questions with a high degree of match to the input question. For example, for the user's input question "Why can't my server be connected?", the system may preliminarily match the example question "How to handle the situation where the server cannot be connected?" in knowledge base A.

[0089] The system further obtains the knowledge base description information corresponding to the preliminarily matched example question. These description information can help the system more accurately understand the content of the knowledge base behind the matching result. For example, the description information for the preliminarily matched information "How to handle the situation where the server cannot be connected?" is "used to solve server connection problems".

[0090] Step S4, obtain the knowledge base number through large model decision-making.

[0091] Finally, the system inputs the preliminary matching results and their corresponding knowledge base description information into the large model for comprehensive analysis. The large model will deeply process all the information and, based on its understanding and analysis, decide the knowledge base number that best matches the user's question. For example, the large model finally determines that the user's question "Why can't my server connect?" should belong to Knowledge Base A.

[0092] Step S5, search results.

[0093] If the target knowledge base corresponding to number A is determined to be Knowledge Base A, then the large model can query the search results corresponding to the input question in Knowledge Base A, that is, the answer to the input question, and return the search results to the client.

[0094] In this embodiment, the knowledge base maintenance description information and Q&A examples can be pre - processed. By performing preliminary RAG matching to obtain high - score example questions and combining the corresponding knowledge base description information, the large model is finally used for comprehensive analysis and decision - making. Through the multi - level matching mechanism, the accuracy and robustness of knowledge base search are significantly improved, and the matching error rate is reduced. Through experiments, the solution of this application can improve the accuracy of various RAG queries and provide more accurate query services for users. For example, in the "Operation and Maintenance Assistant" project, according to the provided metric query ability, the accuracy can be determined to reach more than 95%.

[0095] Based on the same technical concept, the second embodiment of this application provides a knowledge base search enhancement device, as Figure 4 , the device includes:

[0096] A preliminary matching module 401, configured to perform preliminary retrieval enhancement generation matching from the knowledge base according to the input question to obtain at least one example question that matches the input question;

[0097] An acquisition module 402, configured to acquire the description information of the first knowledge base to which the example question belongs;

[0098] An enhanced matching module 403, configured to input all the example questions and all the description information into the large model for analysis, determine the knowledge base number corresponding to the input question, and query the search results corresponding to the input question in the target knowledge base corresponding to the knowledge base number.

[0099] This device can perform preliminary matching from the knowledge base according to the input question using the retrieval enhancement generation technology to obtain matching example questions, acquire the description information of the first knowledge base corresponding to the example questions, and then use the large model to comprehensively analyze the example questions and description information to decide the knowledge base number corresponding to the input question, thereby reducing the knowledge base retrieval matching error rate and improving the user experience and the efficiency of problem - solving.

[0100] Such asFigure 5 As shown in the figure, an embodiment of the present application provides an electronic device, including a processor 111, a communication interface 112, a memory 113, and a communication bus 114. Among them, the processor 111, the communication interface 112, and the memory 113 complete mutual communication through the communication bus 114.

[0101] The memory 113 is used to store computer programs.

[0102] In an embodiment of the present application, when the processor 111 executes the program stored on the memory 113, it implements the knowledge base search enhancement method provided by any one of the foregoing method embodiments, including:

[0103] Perform preliminary retrieval enhancement generation matching from the knowledge base according to the input question to obtain at least one example question that matches the input question;

[0104] Obtain the description information of the first knowledge base to which the example question belongs;

[0105] Input all the example questions and all the description information into a large model for analysis to determine the knowledge base number corresponding to the input question, so as to query the search result corresponding to the input question in the target knowledge base corresponding to the knowledge base number.

[0106] The communication bus mentioned above for the terminal can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into an address bus, a data bus, a control bus, etc. For the sake of simplicity, only a thick line is used to represent it in the figure, but it does not mean that there is only one bus or one type of bus.

[0107] The communication interface is used for communication between the above terminal and other devices.

[0108] The memory may include a Random Access Memory (RAM), or may also include a non-volatile memory, such as at least one disk memory. Optionally, the memory may also be at least one storage device located far from the aforementioned processor.

[0109] The above-mentioned processor may be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it may also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0110] The embodiment of the present application also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the knowledge base search enhancement method provided by any one of the foregoing method embodiments.

[0111] The device embodiments described above are merely illustrative. 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 may be distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0112] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a general-purpose hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the essence of the above technical solution, or the part that contributes to the related technology, can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0113] It should be understood that the terms used herein are for the purpose of describing particular example embodiments only and are not intended to be limiting. Unless the context clearly dictates otherwise, the singular forms "a", "an", and "the" as used herein may also include the plural forms. The terms "comprising", "including", "containing", and "having" are inclusive and thus specify the presence of the stated features, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, steps, operations, elements, components, and / or combinations thereof. The method steps, processes, and operations described herein are not to be construed as necessarily requiring them to be performed in the particular order described or illustrated, unless the order of performance is explicitly stated. It should also be understood that additional or alternative steps may be used.

[0114] It should be understood that the specific embodiments described herein are merely for explaining the present application and are not used to limit the present application. In the description, suffixes such as "module", "component", or "unit" used to denote elements are only for the convenience of explaining the present application and have no specific meaning in themselves. Therefore, "module", "component", or "unit" may be used interchangeably.

[0115] The above are only specific embodiments of the present application, enabling those skilled in the art to understand or implement the present application. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to these embodiments shown herein, but rather to the broadest scope consistent with the principles and novel features claimed herein.

Claims

1. A knowledge base search enhancement method, characterized in that: The method comprises: Performing preliminary search enhancement and matching generation from a knowledge base according to the input question to obtain at least one example question matching the input question; Obtaining description information of the first knowledge base to which the example question belongs; All of the example questions and all of the description information are input into the big model for analysis, and the knowledge base number corresponding to the input question is determined, so as to query the search results corresponding to the input question in the target knowledge base corresponding to the knowledge base number.

2. The method according to claim 1, characterized in that Before performing preliminary search enhancement and matching generation from a knowledge base according to the input question to obtain at least one example question matching the input question, the method further includes: Perform the following maintenance processing for each knowledge base: Configuring description information for the knowledge base, wherein the description information is used to characterize the type of problem solved by the knowledge base; Configure the knowledge base with sample questions and sample responses to the sample questions.

3. The method according to claim 1, characterized in that Performing preliminary search enhancement and matching generation from a knowledge base according to the input question to obtain at least one example question matching the input question includes: Obtaining the input question; A preliminary search is performed from all knowledge bases according to the input question to enhance the generation of matches, and all example questions whose matching degree with the input question is higher than a preset matching degree are obtained.

4. The method according to claim 3, characterized in that According to the input question, a preliminary search is performed from all knowledge bases to enhance the generation of matches, and all example questions with a matching degree higher than a preset matching degree of the input question are obtained, including: Extracting vector features of the input question; Retrieve and enhance the generation of vector matching from all knowledge bases according to the vector feature, and obtain a target vector feature whose matching degree with the vector feature is higher than a preset vector matching degree; Determine the example problem corresponding to the target vector feature.

5. The method according to claim 1, characterized in that Input all the example questions and all the description information into the big model for analysis, and determine the knowledge base number corresponding to the input question, including: Determine a first prompt word according to the example question; Determine a second prompt word according to the description information; All of the first prompt words and all of the second prompt words are input into the large model for analysis to determine the knowledge base number corresponding to the input question.

6. The method according to claim 1, characterized in that Querying the target knowledge base corresponding to the knowledge base number for search results corresponding to the input question includes: Determine the target knowledge base corresponding to the knowledge base number; The search results corresponding to the input question are queried in the target knowledge base through the large model.

7. The method according to claim 2, characterized in that The description information includes at least one of text information, tag information and category information.

8. A knowledge base search enhancement device, characterized in that: The device comprises: A preliminary matching module, used to perform preliminary retrieval enhancement and matching generation from a knowledge base according to an input question, and obtain at least one example question matching the input question; An acquisition module, used for acquiring description information of the first knowledge base to which the example question belongs; The enhanced matching module is used to input all the example questions and all the description information into the large model for analysis, determine the knowledge base number corresponding to the input question, and query the search results corresponding to the input question in the target knowledge base corresponding to the knowledge base number.

9. An electronic device, characterized in that: It includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other through the communication bus; Memory, used to store computer programs; The processor is used to implement the knowledge base search enhancement method described in any one of claims 1-7 when executing the program stored in the memory.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the knowledge base search enhancement method according to any one of claims 1 to 7 is implemented.

Images