Business data extraction system and method based on large model language

Abstract

The invention discloses a business data extraction system and method based on a large model language, belongs to the technical field of artificial intelligence, and aims to solve the technical problem of how to quickly and accurately extract business data to generate a report. Comprising a data middle table used for storing and managing business data and providing API service for the business data; the session management module is used for supporting a user to submit a data request in a natural language mode and performing session management on the data request submitted by the user; the data extraction module is used for analyzing the data request on the basis of a pre-configured large model language, acquiring business data matched with the user intention from the data medium table on the basis of an API service provided by the data medium table, generating a report file in a predetermined format on the basis of the business data, and returning a download link of the report file to the session management module; and the session management module is used for pushing the downloading link of the report file to the user through the chat robot so that the user can download the corresponding report file.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a business data extraction system and method based on a large model language. Background Art

[0002] In many business scenarios, extracting specific information from large amounts of data and generating reports is a time-consuming and error-prone task. Traditional data extraction relies on professional data analysts to configure data sets and reports in the BI system according to user needs. This method relies on manual operations, is inefficient, and is difficult to adapt to the frequent changes in business data extraction needs.

[0003] How to quickly and accurately extract business data to generate reports is a technical problem that needs to be solved. Summary of the invention

[0004] The technical task of the present invention is to address the above shortcomings and provide a business data extraction system and method based on a large model language to solve the technical problem of how to quickly and accurately extract business data to generate reports.

[0005] In a first aspect, the present invention provides a business data extraction system based on a large model language, including a session management module, a data extraction module and a data middle platform;

[0006] The data center is used to store and manage business data and provide API services for business data, supporting the data extraction module to read business data through the API services;

[0007] The session management module is configured with a chat robot, which is used to support users to submit data requests in natural language, and perform session management on the data requests submitted by users, associate the data requests submitted by the current user with context and historical information through session management, and forward the data requests after session management to the data extraction module in a structured manner;

[0008] The data extraction module is used to analyze data requests based on a preconfigured large model language to obtain user intent, and obtain business data matching the user intent from the data center based on the API service provided by the data center. It is used to generate a report file in a predetermined format based on the business data, and return the download link of the report file to the session management module. The session management module is used to push the download link of the report file to the user through a chat robot for the user to download the corresponding report file.

[0009] Preferably, a plurality of database tables are constructed in the data center, and business data related to each business scenario or business type is stored in a corresponding database table, and each database table is configured with an API for accessing the business data stored therein, and the API is used to read and return the business data in the corresponding database table based on the input;

[0010] Correspondingly, a mapping relationship table is constructed in the data extraction module, and the mapping relationship table is used to record the mapping relationship between the API of each database table and the relevant user intention. For the parsed user intention, the data extraction module is used to perform the following: obtain the API of the corresponding database table from the mapping relationship table, convert the user intention recognition into parameters of the input API to trigger the API service of the data middle station, and return the business data in the database table to the data extraction module through the corresponding API.

[0011] Preferably, for the business data returned by the API service, the data extraction module is used to generate an Excel report file from the business data, and send the download link of the Excel report file to the session management module, which is pushed to the user through the chat robot.

[0012] Preferably, the session management module is used to perform session management on data requests submitted by users based on a session tracking algorithm. When a user has a conversation with a chat robot, the session management module is used to create a unique session identifier for the user. During the conversation between the user and the chat robot, the session management module is used to associate each data request submitted by the user and the corresponding reply with the session identifier, so as to perform session management in a state machine-based manner.

[0013] In a second aspect, the present invention provides a method for extracting business data based on a large model language, which is used to extract business data through a business data extraction system based on a large model language as described in any one of the first aspects, and comprises the following steps:

[0014] Store and manage business data through the data center, provide API services for business data, and support data extraction modules to read business data through API services;

[0015] The user submits a data request to the session management module in a natural language, and the session management module performs session management on the data request submitted by the user. The data request submitted by the current user is associated with context and historical information through session management, and the data request after session management is forwarded to the data extraction module in a structured manner;

[0016] The data request is analyzed based on the large model language preconfigured in the data extraction module to obtain the user's intention, and the business data matching the user's intention is obtained from the data center based on the API service provided by the data center. Based on the business data, a report file in a predetermined format is generated through the data extraction module, and the download link of the report file is returned to the session management module. The session management module pushes the download link of the report file to the user through the chat robot for the user to download the corresponding report file.

[0017] Preferably, a plurality of database tables are constructed in the data center, and business data related to each business scenario or business type is stored in a corresponding database table, and each database table is configured with an API for accessing the business data stored therein, and the API is used to read and return the business data in the corresponding database table based on the input;

[0018] Correspondingly, a mapping relationship table is constructed in the data extraction module, which is used to record the mapping relationship between the API of each database table and the relevant user intention. For the parsed user intention, the data extraction module executes the following: obtain the API of the corresponding database table from the mapping relationship table, convert the user intention recognition into parameters of the input API to trigger the API service of the data middle station, and return the business data in the database table to the data extraction module through the corresponding API.

[0019] Preferably, for the business data returned by the API service, an Excel report file is generated from the business data through the data extraction module, and the download link of the Excel report file is sent to the session management module, which pushes it to the user through the chat robot.

[0020] Preferably, the session management module performs session management on data requests submitted by users based on a session tracking algorithm. When a user has a conversation with a chat robot, a unique session identifier is created for the user through the session management module. During the conversation between the user and the chat robot, each data request submitted by the user and the corresponding reply are associated with the session identifier through the session management module, and session management is performed in a state machine-based manner.

[0021] The business data extraction system and method based on the big model language of the present invention have the following advantages:

[0022] 1. Improved user experience: The chatbot is used to handle user business data extraction needs, solving the tedious problem of logging into various business systems to download data and then merging and processing it when extracting user data;

[0023] 2. Improved data extraction efficiency: The data center stores various business data through database tables and provides APIs for each database table. It uses large language model (LLM) technology to identify user intent and match the data center API, thereby improving the degree of automation in the data extraction process. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0025] The present invention is further described below in conjunction with the accompanying drawings.

[0026] Figure 1 This is a flowchart of a business data extraction method based on a large model language in Example 2. DETAILED DESCRIPTION

[0027] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments so that those skilled in the art can better understand the present invention and implement it. However, the embodiments are not intended to limit the present invention. In the absence of conflict, the embodiments of the present invention and the technical features in the embodiments may be combined with each other.

[0028] The embodiments of the present invention provide a business data extraction system and method based on a large model language, which are used to solve the technical problem of how to quickly and accurately extract business data to generate reports.

[0029] Embodiment 1:

[0030] The present invention discloses a business data extraction system based on a large model language, comprising a session management module, a data extraction module and a data middle platform.

[0031] The data center is used to store and manage business data and provide API services for business data. It supports the data extraction module to read business data through API services.

[0032] As a specific implementation of the data middle platform, multiple database tables are constructed in the data middle platform. The business data related to each business scenario or business type is stored in a corresponding database table. Each database table is configured with an API for accessing the business data stored therein. The API is used to read and return the business data in the corresponding database table based on the input.

[0033] The session management module is configured with a chatbot to support users in submitting data requests in natural language and to perform session management on the data requests submitted by users. Through session management, the data request submitted by the current user is associated with the context and historical information, and the data request after session management is forwarded to the data extraction module in a structured manner.

[0034] As a specific implementation of the session management module, the session management module is used to perform session management on data requests submitted by users based on a session tracking algorithm. When a user has a conversation with a chat robot, the session management module is used to create a unique session identifier for the user. During the conversation between the user and the chat robot, the session management module is used to associate each data request submitted by the user and the corresponding reply with the session identifier, so as to perform session management in a state machine-based manner.

[0035] In this embodiment, the session management module receives the user's natural language request based on the chatbot, parses the request, and triggers the data extraction service; and maintains the user's session state based on the Chatbot to ensure that the context and historical information of the request are processed correctly.

[0036] The data extraction module is used to analyze data requests based on the preconfigured big model language to obtain user intent, and obtain business data that matches the user intent from the data center based on the API service provided by the data center. It is used to generate a report file in a predetermined format based on the business data, and return the download link of the report file to the session management module. The session management module is used to push the download link of the report file to the user through the chat robot for the user to download the corresponding report file.

[0037] For the business data returned by the API service, the data extraction module is used to generate an Excel report file from the business data, and send the download link of the Excel report file to the session management module, which is pushed to the user through the chat robot.

[0038] As a specific implementation of the data extraction module, a mapping relationship table is constructed in the data extraction module, which is used to record the mapping relationship between the API of each database table and the relevant user intention. For the parsed user intention, the data extraction module is used to perform the following: obtain the API of the corresponding database table from the mapping relationship table, convert the user intention recognition into parameters of the input API to trigger the API service of the data middle station, and return the business data in the database table to the data extraction module through the corresponding API.

[0039] In this embodiment, the data extraction module uses the large language model LLM to identify user intentions, determine the specific requirements of data extraction, and convert user intentions into parameters acceptable to the API in the data center. Based on the parameters, the API service of the data center can be triggered to read business data from the database table corresponding to the API and return it to the data extraction module. At the same time, the data extraction module automatically formats the extracted data and generates an Excel report, and returns the download link of the report to the session management module, which is returned to the user through Chatbot for download.

[0040] The system disclosed in this embodiment utilizes the intent recognition capability of the LLM model and combines the chat context sent by the user through the Chatbot to obtain the user's data extraction needs, and realizes rapid data extraction and report generation in an automated manner.

[0041] Embodiment 2:

[0042] The present invention provides a method for extracting business data based on a large model language, which extracts business data through the system disclosed in Example 1. The method comprises the following steps:

[0043] Step S100: Store and manage business data through the data middle platform, provide API services for business data, and support data extraction module to read business data through API services.

[0044] Multiple database tables are built in the data center. Business data related to each business scenario or business type is stored in a corresponding database table. Each database table is configured with an API for accessing the business data stored therein. The API is used to read and return the business data in the corresponding database table based on the input.

[0045] Step S200: The user submits a data request to the session management module in natural language. The session management module performs session management on the data request submitted by the user. The data request submitted by the current user is associated with context and historical information through session management, and the data request after session management is forwarded to the data extraction module in a structured manner.

[0046] As a specific implementation of step S200, session management is performed on data requests submitted by users based on a session tracking algorithm and through a session management module. When a user has a conversation with a chat robot, a unique session identifier is created for the user through the session management module. During the conversation between the user and the chat robot, each data request submitted by the user and the corresponding reply are associated with the session identifier through the session management module, and session management is performed in a state machine-based manner.

[0047] In this embodiment, the session management module receives the user's natural language request based on the chatbot, parses the request, and triggers the data extraction service; and maintains the user's session state based on the Chatbot to ensure that the context and historical information of the request are processed correctly.

[0048] Step S300: Analyze the data request based on the large model language preconfigured in the data extraction module to obtain the user's intention, and obtain the business data matching the user's intention from the data center based on the API service provided by the data center. Generate a report file in a predetermined format based on the business data through the data extraction module, and return the download link of the report file to the session management module. The session management module pushes the download link of the report file to the user through the chat robot for the user to download the corresponding report file.

[0049] For the business data returned by the API service, the business data is generated into an Excel report file through the data extraction module, and the download link of the Excel report file is sent to the session management module and pushed to the user through the chat robot.

[0050] As a specific implementation of data extraction, a mapping relationship table is constructed in the data extraction module, which is used to record the mapping relationship between the API of each database table and the relevant user intention. For the parsed user intention, the data extraction module is used to execute the following: obtain the API of the corresponding database table from the mapping relationship table, convert the user intention recognition into parameters of the input API to trigger the API service of the data middle station, and return the business data in the database table to the data extraction module through the corresponding API.

[0051] In this embodiment, the large language model LLM is used to identify user intentions and determine specific requirements for data extraction, and the user intentions are converted into parameters acceptable to the API in the data center. Based on the parameters, the API service of the data center can be triggered, and the business data is read from the database table corresponding to the API and returned to the data extraction module. At the same time, the extracted data is automatically formatted and an Excel report is generated through the data extraction module, and the download link of the report is returned to the session management module, and returned to the user through the Chatbot for download.

[0052] The business data extraction system and method based on the big model language provided by the present invention are introduced in detail above. Specific examples are used in this article to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for general technical personnel in this field, according to the idea of ​​the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A business data extraction system based on a large model language, characterized in that: Includes session management module, data extraction module and data middle platform; The data center is used to store and manage business data and provide API services for business data, supporting the data extraction module to read business data through the API services; The session management module is configured with a chat robot, which is used to support users to submit data requests in natural language, and perform session management on the data requests submitted by users, associate the data requests submitted by the current user with context and historical information through session management, and forward the data requests after session management to the data extraction module in a structured manner; The data extraction module is used to analyze data requests based on a preconfigured large model language to obtain user intent, and obtain business data matching the user intent from the data center based on the API service provided by the data center. It is used to generate a report file in a predetermined format based on the business data, and return the download link of the report file to the session management module. The session management module is used to push the download link of the report file to the user through a chat robot for the user to download the corresponding report file.

2. The business data extraction system based on the large model language according to claim 1 is characterized in that: The data center has multiple database tables. The business data related to each business scenario or business type is stored in a corresponding database table. Each database table is configured with an API for accessing the business data stored therein. The API is used to read and return the business data in the corresponding database table based on the input; Correspondingly, a mapping relationship table is constructed in the data extraction module, and the mapping relationship table is used to record the mapping relationship between the API of each database table and the relevant user intention. For the parsed user intention, the data extraction module is used to perform the following: obtain the API of the corresponding database table from the mapping relationship table, convert the user intention recognition into parameters of the input API to trigger the API service of the data middle station, and return the business data in the database table to the data extraction module through the corresponding API.

3. The business data extraction system based on large model language according to claim 1 is characterized in that: For the business data returned by the API service, the data extraction module is used to generate an Excel report file from the business data, and send the download link of the Excel report file to the session management module, which is pushed to the user through the chat robot.

4. The business data extraction system based on large model language according to any one of claims 1 to 3, characterized in that: The session management module is used to perform session management on data requests submitted by users based on a session tracking algorithm. When a user has a conversation with a chat robot, the session management module is used to create a unique session identifier for the user. During the conversation between the user and the chat robot, the session management module is used to associate each data request submitted by the user and the corresponding reply with the session identifier, so as to perform session management in a state machine-based manner.

5. A business data extraction method based on a large model language, characterized in that: The method is used to extract business data through a business data extraction system based on a large model language as described in any one of claims 1 to 4, comprising the following steps: Store and manage business data through the data center, provide API services for business data, and support data extraction modules to read business data through API services; The user submits a data request to the session management module in a natural language, and the session management module performs session management on the data request submitted by the user. The data request submitted by the current user is associated with context and historical information through session management, and the data request after session management is forwarded to the data extraction module in a structured manner; The data request is analyzed based on the large model language preconfigured in the data extraction module to obtain the user's intention, and the business data matching the user's intention is obtained from the data center based on the API service provided by the data center. Based on the business data, a report file in a predetermined format is generated through the data extraction module, and the download link of the report file is returned to the session management module. The session management module pushes the download link of the report file to the user through the chat robot for the user to download the corresponding report file.

6. The business data extraction method based on large model language according to claim 5 is characterized in that: The data center has multiple database tables. The business data related to each business scenario or business type is stored in a corresponding database table. Each database table is configured with an API for accessing the business data stored therein. The API is used to read and return the business data in the corresponding database table based on the input; Correspondingly, a mapping relationship table is constructed in the data extraction module, which is used to record the mapping relationship between the API of each database table and the relevant user intention. For the parsed user intention, the data extraction module executes the following: obtain the API of the corresponding database table from the mapping relationship table, convert the user intention recognition into parameters of the input API to trigger the API service of the data middle station, and return the business data in the database table to the data extraction module through the corresponding API.

7. The business data extraction method based on large model language according to claim 5 is characterized in that: For the business data returned by the API service, the business data is generated into an Excel report file through the data extraction module, and the download link of the Excel report file is sent to the session management module, which pushes it to the user through the chat robot.

8. The method for extracting business data based on a large model language according to any one of claims 5 to 7, characterized in that: The session management module performs session management on data requests submitted by users based on a session tracking algorithm. When a user has a conversation with a chat robot, a unique session identifier is created for the user through the session management module. During the conversation between the user and the chat robot, the session management module associates each data request submitted by the user and the corresponding reply with the session identifier, and performs session management in a state machine-based manner.

Images