Urban management intelligent question and answer assistant method based on large language model

Abstract

The invention provides a city management intelligent question and answer assistant method based on a large language model, which comprises the following steps: establishing a large language model, and adopting a pre-trained large language model as a core; constructing a professional knowledge vector library which comprises knowledge contents in the field of city management, including laws and regulations and policy documents; for business function calling, the intelligent city management business function is embedded into a large language model through a self-defined function, user input is analyzed through the large language model, and the business function is called according to the user; multi-modal interaction is realized, the input form of the interaction comprises characters and voices, and the system can be combined with a large language model technology and a smart city management service port to realize intelligent dialogue with a user, quickly answer various problems related to city management and guide the user to use a smart city management service function. And the convenience and timeliness of the original service system are improved through the intelligent and integrated intelligent assistant.

Description

Technical Field

[0001] The present invention relates to the technical field of large language model application, and in particular to an urban management intelligent question-answering assistant method based on a large language model. Background Art

[0002] With the acceleration of urbanization, urban management faces many challenges, such as environmental sanitation, urban appearance, and maintenance of public facilities. In recent years, although some cities have begun to try to use the Internet, big data, artificial intelligence and other technical means to build smart city management business systems and improve urban management efficiency, they lack an integrated and intelligent question-and-answer assistant to quickly respond to various questions and needs of citizens and urban management staff.

[0003] Since its introduction, the application of big language models in various fields has shown great potential and value, especially in natural language processing such as text understanding and dialogue systems, which can provide users with timely and accurate answers and feedback. However, due to the limitation of training data, big language models lack professional knowledge in specific fields, resulting in unsatisfactory answers in specific fields. Therefore, the existing technology based on big language models is not intelligent enough in urban management, and cannot provide auxiliary guidance for smart city management related businesses. Summary of the invention

[0004] In view of the shortcomings of the prior art, the purpose of the present invention is to provide an intelligent question-answering assistant method for urban management based on a large language model to solve the problems raised in the above-mentioned background technology. The present invention can combine the large language model technology and the smart city management service port to realize intelligent dialogue with users, quickly answer various questions related to urban management, and guide users to use smart city management service functions.

[0005] In order to achieve the above object, the present invention is implemented through the following technical solution: a method for an urban management intelligent question-answering assistant based on a large language model, comprising the following steps:

[0006] Step 1: Build a large language model, using a pre-trained large language model as the core;

[0007] Step 2: Create a professional knowledge vector library, which includes knowledge content in the field of urban management, including laws, regulations, and policy documents;

[0008] Step 3: For business function calls, embed the smart city management business functions into the big language model through custom functions, use the big language model to analyze user input, and call business functions based on users;

[0009] Step 4: Implement multimodal interaction, where the input forms include text and voice.

[0010] Furthermore, users interact with the large language model microservice through interactive pages using text input or voice input, and perform urban management instruction question and answer processes and business function jumps on the server side.

[0011] Furthermore, the business function jump displays the business function page to the user on the Web front end through the jump address.

[0012] Furthermore, the large language model microservice will receive the urban management knowledge answer content generated by the urban management knowledge question and answer and feed it back to the user's interactive interface.

[0013] Furthermore, the knowledge content in the field of urban management also includes the establishment of urban management organizations, the establishment of urban management organizations, the urban management command manual, the urban management assessment method, the personnel management system, the urban appearance and environmental sanitation management regulations, the urban management law enforcement methods, the urban management regulations of different cities, and the urban management law enforcement methods of different cities.

[0014] Furthermore, the urban management field knowledge content is uploaded to a vector database for storage after text segmentation and text vectorization.

[0015] Furthermore, the voice input process in step 4 is as follows: the user inputs a voice file, and a natural language is generated through a voice recognition module.

[0016] Furthermore, it also includes an urban management knowledge question and answer module. In the urban management knowledge question and answer module, after the user inputs a question, an answer is generated based on ChatLLM and returned to the user end.

[0017] Furthermore, the questions input by the user are retrieved from the private knowledge base through the text vectorization model. After the corresponding private knowledge is retrieved, the answer is also generated based on ChatLLM and returned to the user end.

[0018] Beneficial effects of the present invention:

[0019] 1. The urban management intelligent question-and-answer assistant method based on a large language model combines large language model technology, RAG technology and function call technology with the smart city management business system, and improves the convenience and timeliness of the original business system through an intelligent and integrated intelligent assistant.

[0020] 2. The urban management intelligent question-answering assistant method based on a large language model builds a professional knowledge base in the urban management field, improving the accuracy of the intelligent assistant when answering professional questions in the urban management field.

[0021] 3. The urban management intelligent question-and-answer assistant method based on the large language model can combine the large language model technology and the smart city management business port to achieve intelligent dialogue with users, quickly answer various questions related to urban management, and guide users to use smart city management business functions. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 A flowchart of an urban management intelligent question-answering assistant method based on a large language model according to the present invention;

[0023] Figure 2 Constructing a flow chart for the urban management professional knowledge vector database of the present invention;

[0024] Figure 3 It is a processing logic diagram of the urban management knowledge question and answer module of the present invention;

[0025] Figure 4 It is a logic diagram for processing the business function jump module of the present invention;

[0026] Figure 5 This is a logic diagram of voice input processing of the present invention. DETAILED DESCRIPTION

[0027] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the present invention is further explained below in conjunction with specific implementation methods.

[0028] See also Figures 1 to 5 The present invention provides the following technical solutions: a method for an intelligent urban management question-answering assistant based on a large language model, which can combine the large language model technology and the smart city management service port to realize intelligent dialogue with users, quickly answer various questions related to urban management, and guide users to use smart city management service functions. The specific solution content of the above method provided in this embodiment is as follows:

[0029] 1. Large language model: Using a pre-trained large language model as the core, through RAG and other technical means, it adapts to the question-answering needs in the field of urban management, and achieves deep understanding and accurate answers. This method combines large language model technology, RAG technology and function call technology with the smart city management business system, and improves the convenience and timeliness of the original business system through intelligent and integrated smart assistants;

[0030] 2. Professional knowledge vector library: Build a vector library of knowledge in the field of urban management, including laws, regulations, policy documents, etc., to provide a solid knowledge foundation for the question-answering system and improve the accuracy of the intelligent assistant when answering professional questions in the field of urban management.

[0031] 3. Business function call: embed smart city management business functions into the big language model through custom functions, use the big language model to analyze user input, and call business functions based on users;

[0032] 4. Multimodal interaction: supports multiple forms of input such as text and voice to improve the convenience and experience of user interaction.

[0033] This embodiment also provides the following specific implementation process to explain the above method, which is as follows:

[0034] S1, refer to Figure 2 , build a professional knowledge vector database in the field of urban management

[0035] This step specifically includes:

[0036] 1) Collect documents on relevant regulations, laws and regulations, industry knowledge, etc. in the field of urban management;

[0037] 2) Load the document using TikaDocumentReader;

[0038] 3) Inherit the document segmentation class TextSplitter in the Spring AI framework, modify the underlying logic to segment the document content by chapters, and obtain several text blocks;

[0039] 4) In the Spring AI framework vectorStore, use the mofanke / dmeta-embedding-zh model provided in ollama to vectorize the text block and import it into the PGVector vector database for storage to form a professional knowledge vector database in the urban management field.

[0040] S2, reference Figure 3 , build a knowledge question and answer service for urban management

[0041] This step specifically includes:

[0042] 1) Use the similaritySearch function in the Spring AI framework vectorStore to implement information retrieval in the urban management professional knowledge vector database

[0043] 2) In the Spring AI framework, OllamaChatModel inputs user questions, database search results and prompts into the Qwen2.5 model, generates urban management knowledge answers, and returns them to the user.

[0044] S3, refer to Figure 4 Build business function jump service

[0045] 1) Use Function in JAVA to write custom functions

[0046] 2) Use ChatClient in the Spring AI framework to enable this function in the Qwen2.5 model

[0047] S4, using FunASR speech recognition model reference Figure 5Build a voice input service to support user voice input.

[0048] The basic principles and main features of the present invention and the advantages of the present invention are shown and described above. For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the spirit or basic characteristics of the present invention.

[0049] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

Claims

1. A method for an urban management intelligent question-answering assistant based on a large language model, characterized in that: The following steps are involved: Step 1: Build a large language model, using a pre-trained large language model as the core; Step 2: Create a professional knowledge vector library, which includes knowledge content in the field of urban management, including laws, regulations, and policy documents; Step 3: For business function calls, embed the smart city management business functions into the big language model through custom functions, use the big language model to analyze user input, and call business functions based on users; Step 4: Implement multimodal interaction, where the input forms include text and voice.

2. According to claim 1, a method for urban management intelligent question-answering assistant based on a large language model is characterized in that: Users interact with the large language model microservice through interactive pages, using text input or voice input, and perform urban management instructions, question and answer processes, and business function jumps on the server.

3. According to claim 2, a method for urban management intelligent question-answering assistant based on a large language model is characterized in that: The business function jump displays the business function page to the user on the Web front end through the jump address.

4. According to claim 2, a method for urban management intelligent question-answering assistant based on a large language model is characterized in that: The large language model microservice will receive the urban management knowledge answers generated by the urban management knowledge question and answer and feed them back to the user's interactive interface.

5. The urban management intelligent question-answering assistant method based on a large language model according to claim 1 is characterized in that: The knowledge content in the field of urban management also includes the establishment of urban management organizations, the establishment of urban management agencies, the urban management command manual, the urban management assessment method, the personnel management system, the urban appearance and environmental sanitation management regulations, the urban management law enforcement methods, the urban management regulations of different cities, and the urban management law enforcement methods of different cities.

6. The urban management intelligent question-answering assistant method based on a large language model according to claim 5 is characterized in that: The urban management field knowledge content is uploaded to a vector database for storage after text segmentation and text vectorization.

7. The urban management intelligent question-answering assistant method based on a large language model according to claim 1 is characterized in that: The voice input process in step 4 is as follows: the user inputs a voice file, and a voice recognition module generates natural language.

8. The urban management intelligent question-answering assistant method based on a large language model according to claim 1 is characterized in that: It also includes an urban management knowledge question and answer module. In the urban management knowledge question and answer module, after the user inputs a question, an answer is generated based on ChatLLM and returned to the user end.

9. The urban management intelligent question-answering assistant method based on a large language model according to claim 8 is characterized in that: The questions input by the user are retrieved from the private knowledge base through the text vectorization model. After the corresponding private knowledge is retrieved, the answer is also generated based on ChatLLM and returned to the user.

Images