Business index analysis method, system and device, electronic equipment and program product

Abstract

The invention relates to a business index analysis method, system and device, electronic equipment and a program product, and relates to the technical field of computers. The method comprises the following steps: in response to a business index analysis request, performing intention analysis processing on the business index analysis request to obtain a business analysis intention; based on the business analysis intention, determining a target calling tool and a target query parameter from a pre-acquired tool description list; executing data extraction operation according to the target calling tool and the target query parameter to obtain initial business index data; and generating a business index analysis result based on the initial business index data. According to the method and the device, the intention recognition capability and tool calling for the business index analysis request are utilized to perform data extraction operation to obtain the business index analysis result, so that the request of the user can be responded in seconds, accurate analysis can be given, and the agile experience of'questioning and analyzing 'is realized.

Description

Technical Field

[0001] This disclosure relates to the field of computer technology, and more specifically, to a business indicator analysis method, a business indicator analysis system, a business indicator analysis device, an electronic device, and a computer program product. Background Technology

[0002] As the digital transformation of the telecommunications (CT) industry continues to advance, enterprises are increasingly demanding analysis of business metrics in their daily operations and management. Traditional telecommunications business metric analysis processes typically rely on personnel with specialized knowledge to manually construct queries or use reporting tools to extract data from databases before performing analysis.

[0003] It should be noted that the information disclosed in the background section above is only used to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0004] The purpose of this disclosure is to provide a business indicator analysis method, business indicator analysis system, business indicator analysis device, electronic device, computer-readable storage medium, and computer program product, thereby overcoming, to at least a certain extent, the problems of high dependence on professional personnel, high data usage threshold, and low data analysis efficiency in related technical solutions for telecommunications business indicator analysis.

[0005] Other features and advantages of this disclosure will become apparent from the following detailed description, or may be learned in part by practice of the invention.

[0006] According to a first aspect of this disclosure, a business indicator analysis method is provided, comprising: responding to a business indicator analysis request, performing intent parsing processing on the business indicator analysis request to obtain a business analysis intent; based on the business analysis intent, determining a target calling tool and target query parameters from a pre-acquired tool description list; performing a data extraction operation according to the target calling tool and the target query parameters to obtain initial business indicator data; and generating a business indicator analysis result based on the initial business indicator data.

[0007] In one exemplary embodiment of this disclosure, the method further includes: starting a pre-configured Model Context Protocol (MTP) server, which provides external services through a specified port; performing a connection configuration operation with the MTP server based on a MTP client to establish a communication connection with the MTP server; and obtaining the tool description list from the MTP server through the MTP client.

[0008] In one exemplary embodiment of this disclosure, obtaining the tool description list from the Model Context Protocol (MTP) server via the MTP client includes: sending a tool invocation request to the MTP server via the MTP client; and having the MTP server return the tool description list; the tool description list includes one or more of the following metadata information: database operations, metric extraction methods, and parameter restrictions.

[0009] In one exemplary embodiment of this disclosure, the step of performing intent parsing processing on the business indicator analysis request to obtain the business analysis intent includes: inputting the business indicator analysis request into a pre-trained indicator analysis model; and having the indicator analysis model perform intent parsing processing on the business indicator analysis request to obtain the business analysis intent, wherein the business analysis intent includes one or more of the following: data to be analyzed, the calling tool, and the query parameters; wherein the data to be analyzed includes one or more of the following: a query time range, a geographic range, a product identifier, and key business indicators.

[0010] In one exemplary embodiment of this disclosure, the step of performing data extraction operations based on the target invocation tool and the target query parameters to obtain initial business indicator data includes: constructing a structured query statement by the model context protocol server based on the target invocation tool and the target query parameters; and performing data extraction operations in a pre-built business database based on the structured query statement to obtain the initial business indicator data, wherein the initial business indicator data is structured business indicator data.

[0011] In one exemplary embodiment of this disclosure, generating business indicator analysis results based on the initial business indicator data includes: performing multidimensional analysis and semantic interpretation processing on the initial business indicator data based on a pre-built business knowledge base to obtain business indicator data; performing image-text conversion processing on the business indicator data to obtain the business indicator analysis results; the business knowledge base includes one or more of predefined business indicator descriptions, industry knowledge graphs, and business meaning databases.

[0012] In one exemplary embodiment of this disclosure, the method further includes: performing content extraction on the business indicator analysis results to obtain content in a specified format; performing block processing on the specified format content to obtain multiple content text blocks; performing vector transformation processing on the content text blocks to obtain content representation vectors; and storing the content representation vectors in a vector database, wherein the vector database is used to support semantic retrieval of the business indicator analysis results.

[0013] According to a second aspect of this disclosure, a business indicator analysis system is provided, comprising: a large model intent parsing module, used to perform intent parsing processing on business indicator analysis requests using a large language model to obtain business analysis intent, and to determine target calling tools and target query parameters from a tool description list based on the business analysis intent; a model context protocol client, used to generate a standardized calling request based on the target calling tool and the target query parameters, and to send the standardized calling request to a model context protocol server; the model context protocol server, used to perform data extraction operations based on the standardized calling request to obtain initial business indicator data; a large model data interpretation module, used to perform multidimensional analysis and semantic interpretation processing on the initial business indicator data based on the large language model and combined with business knowledge from a business knowledge base to obtain business indicator analysis results; a business database, used to store full business data, the full business data including one or more of user information data, business product information, business processing records, billing and accounting data, and user behavior data; and the business knowledge base, used to provide business knowledge of the target business to the large language model.

[0014] According to a third aspect of this disclosure, a business indicator analysis apparatus is provided, comprising: a business intent parsing module, configured to, in response to a business indicator analysis request, perform intent parsing processing on the business indicator analysis request to obtain a business analysis intent; a tool parameter acquisition module, configured to, based on the business analysis intent, determine a target calling tool and target query parameters from a pre-acquired tool description list; a data extraction module, configured to, perform a data extraction operation based on the target calling tool and the target query parameters to obtain initial business indicator data; and an analysis result determination module, configured to, based on the initial business indicator data, generate a business indicator analysis result.

[0015] In one exemplary embodiment of this disclosure, the business indicator analysis device further includes a tool list acquisition module, configured to: start a pre-configured Model Context Protocol (MGP) server, which provides external services through a specified port; perform a connection configuration operation with the MGP server based on a MGP client to establish a communication connection with the MGP server; and acquire the tool description list from the MGP server through the MGP client.

[0016] In one exemplary embodiment of this disclosure, the tool list acquisition module includes a tool list acquisition unit, configured to: send a tool call request to the Model Context Protocol server through the Model Context Protocol client; and have the Model Context Protocol server return the tool description list; the tool description list includes one or more of the following metadata information: database operations, indicator extraction methods, and parameter restrictions.

[0017] In one exemplary embodiment of this disclosure, the business intent parsing module includes a business intent parsing unit, configured to: input the business indicator analysis request into a pre-trained indicator analysis model; and perform intent parsing processing on the business indicator analysis request by the indicator analysis model to obtain the business analysis intent, wherein the business analysis intent includes one or more of the following: data to be analyzed, the calling tool, and the query parameters; wherein the data to be analyzed includes one or more of the following: a query time range, a geographic range, a product identifier, and key business indicators.

[0018] In one exemplary embodiment of this disclosure, the data extraction module includes a data extraction unit, configured to: construct a structured query statement by a model context protocol server based on the target calling tool and the target query parameters; and perform a data extraction operation in a pre-built business database based on the structured query statement to obtain the initial business indicator data, wherein the initial business indicator data is structured business indicator data.

[0019] In one exemplary embodiment of this disclosure, the analysis result determination module includes an analysis result determination unit, configured to: perform multidimensional analysis and semantic interpretation processing on the initial business indicator data based on a pre-built business knowledge base to obtain business indicator data; perform image-text conversion processing on the business indicator data to obtain the business indicator analysis result; the business knowledge base includes one or more of predefined business indicator descriptions, industry knowledge graphs, and business meaning databases.

[0020] In one exemplary embodiment of this disclosure, the above-mentioned business indicator analysis device further includes a vector library construction module, configured to: perform content extraction operations on the business indicator analysis results to obtain content in a specified format; perform block processing on the specified format content to obtain multiple content text blocks; perform vector conversion processing on the content text blocks to obtain content representation vectors; and store the content representation vectors in a vector database, wherein the vector database is used to support semantic retrieval of the business indicator analysis results.

[0021] According to a fourth aspect of this disclosure, an electronic device is provided, comprising: a processor; and a memory storing computer-readable instructions, which, when executed by the processor, implement the business metric analysis method according to any one of the preceding claims.

[0022] According to a fifth aspect of this disclosure, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the business indicator analysis method according to any one of the preceding claims.

[0023] According to a sixth aspect of this disclosure, a computer program product is provided, including a computer program that, when executed by a processor, implements the business indicator analysis method described in any one of the preceding claims.

[0024] The technical solution provided in this disclosure may include the following beneficial effects:

[0025] The business indicator analysis method in the exemplary embodiments of this disclosure, on the one hand, provides a fast, accurate, and manual-intervention-free business indicator analysis solution by automatically extracting and intelligently interpreting business data based on the user's business data analysis request. On the other hand, by utilizing intent recognition capabilities and tool calls for business indicator analysis requests, data extraction operations are performed to obtain business indicator analysis results. This allows for a response to user requests within seconds and provides accurate analysis, achieving an agile experience of "analysis upon request."

[0026] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0027] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure. It is obvious that the drawings described below are merely some embodiments of this disclosure, and those skilled in the art can obtain other drawings based on these drawings without any inventive effort. In the drawings:

[0028] Figure 1 A flowchart illustrating a business metric analysis method according to an exemplary embodiment of the present disclosure is shown schematically;

[0029] Figure 2 An overall flowchart of a business metrics analysis method according to an exemplary embodiment of the present disclosure is illustrated schematically;

[0030] Figure 3 A schematic diagram illustrating the system architecture of a business metrics analysis system according to an exemplary embodiment of the present disclosure is provided.

[0031] Figure 4 The diagram illustrates the interaction flowchart between internal modules in a business metrics analysis system according to an exemplary embodiment of the present disclosure.

[0032] Figure 5 A block diagram of a business metrics analysis apparatus according to an exemplary embodiment of the present disclosure is shown schematically;

[0033] Figure 6 The illustration schematically shows a computer-readable storage medium according to an exemplary embodiment of the present disclosure;

[0034] Figure 7 A block diagram of an electronic device according to an exemplary embodiment of the present disclosure is shown schematically. Detailed Implementation

[0035] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the embodiments set forth herein; rather, they are provided so that this disclosure will be thorough and complete, and will fully convey the concept of the exemplary embodiments to those skilled in the art. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted.

[0036] Furthermore, the described features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. Numerous specific details are provided in the following description to give a thorough understanding of embodiments of this disclosure. However, those skilled in the art will recognize that the technical solutions of this disclosure can be practiced without one or more of the specific details described, or other methods, components, apparatuses, steps, etc., can be employed. In other instances, well-known structures, methods, apparatuses, implementations, materials, or operations are not shown or described in detail to avoid obscuring various aspects of this disclosure.

[0037] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities can be implemented in software, or in one or more software-hardened modules, or in different network and / or processor devices and / or microcontroller devices.

[0038] Traditional telecommunications service metric analysis processes typically rely on personnel with specialized knowledge to manually construct queries or use reporting tools to extract data from databases before analysis. This process suffers from high barriers to entry, low efficiency, and poor timeliness, hindering rapid response to business needs.

[0039] In recent years, with significant breakthroughs in large language model capabilities, intelligent agents possessing natural language understanding and contextual reasoning abilities have been increasingly applied to various data analysis and automated question-answering scenarios. Against this backdrop, how to introduce large model capabilities into telecommunications business systems and, in conjunction with model context protocols, achieve an automated closed loop from user intent parsing and data querying to business indicator analysis and result generation has become an important application direction for intelligent operations in the telecommunications industry.

[0040] Based on this, according to the embodiments of this disclosure, a business indicator analysis method, a business indicator analysis system, a business indicator analysis device, an electronic device, a computer-readable storage medium, and a computer program product are proposed.

[0041] In this article, it is important to understand the terms used, such as Large Language Model (LLM), which refers to a class of artificial intelligence models trained on massive amounts of text data. These models can understand and generate natural language and perform tasks such as chatting, writing, translating, programming, and question answering.

[0042] The Model Context Protocol (MCP) is a standardized interface bridge for large language models to call external tools and connect to other business systems. It gives models "hands-on ability," enabling them not only to answer questions but also to proactively call tools, query data, and complete tasks, making models truly "super assistants" for enterprises.

[0043] Furthermore, the number of any elements in the accompanying drawings is for illustrative purposes only and not for limitation, and any naming is for distinction only and has no limiting meaning.

[0044] The principles and spirit of this disclosure will be explained in detail below with reference to several representative embodiments.

[0045] First, in this example embodiment, a business indicator analysis method is provided. The business indicator analysis method of this disclosure can be implemented using a server or using a terminal device. The terminal described in this disclosure can include mobile terminals such as mobile phones, tablets, laptops, handheld computers, and personal digital assistants (PDAs), as well as fixed terminals such as desktop computers. Figure 1 The illustration shows a schematic diagram of a business metric analysis method flow according to some embodiments of the present disclosure. (Reference) Figure 1 The business metric analysis method may include the following steps:

[0046] Step S110: In response to the business indicator analysis request, perform intent parsing processing on the business indicator analysis request to obtain the business analysis intent;

[0047] Step S120: Based on the business analysis intent, determine the target calling tool and target query parameters from the pre-acquired tool description list;

[0048] Step S130: Perform data extraction operation according to the target calling tool and target query parameters to obtain initial business indicator data;

[0049] Step S140: Generate business indicator analysis results based on the initial business indicator data.

[0050] According to the business indicator analysis method in this example embodiment, on the one hand, it provides a fast, accurate, and manual-intervention-free business indicator analysis solution by automatically extracting and intelligently interpreting business data based on the user's business data analysis request. On the other hand, it utilizes intent recognition capabilities and tool calls for business indicator analysis requests to perform data extraction operations and obtain business indicator analysis results. It can respond to user requests in seconds and provide accurate analysis, achieving an agile experience of "analysis upon request".

[0051] The business indicator analysis method in this example embodiment will be further explained below.

[0052] This disclosed business indicator analysis method provides a telecommunications business indicator analysis scheme based on a large language model. This method can be implemented using a pre-built business indicator analysis system. Users can ask questions to the large model using natural language, thereby achieving automatic extraction and intelligent interpretation of telecommunications business data. (Reference) Figure 2 , Figure 2 The diagram schematically illustrates an overall flowchart of a business metrics analysis method according to an exemplary embodiment of this disclosure. Before processing a user's business metrics analysis request, system configuration and connection operations between various modules can be completed first. In step S200, a connection is established between the MCP client and the MCP server.

[0053] In one exemplary embodiment of this disclosure, a pre-configured Model Context Protocol (MCP) server is started, providing external services through a specified port; a connection configuration operation is performed on the MCP client to establish a communication connection with the MCP server; and a tool description list is obtained from the MCP server through the MCP client. This tool description list may be a list of tools registered by the MCP server, representing all the capabilities that the MCP server can provide.

[0054] The business indicator analysis system disclosed herein may include a large language model for processing business indicator analysis requests, an MCP client, and an MCP server. For the client, Cheery Studio client can be used. This client supports adding MCP server information and has a built-in MCP client module, allowing direct tool call requests to the MCP server. For the large language model, it is integrated with the system via an Application Programming Interface (API) call method, submitting user requests to the large language model for processing. The MCP server, capable of extracting telecommunications business indicator data, can be deployed in a local environment, and its connection information can be configured in the Cheery Studio client for system use. One possible configuration of the connection information is as follows:

[0055]

[0056] The configuration information mentioned above includes the name, type, description, whether it is active, and the URL address of the MCP server.

[0057] The specific implementation process is as follows: First, the MCP server is started and provides services externally via HTTP port 8080. After configuring the corresponding connection information in the Cheery Studio client, the user can establish a connection with the MCP server via HTTP. Once the connection is established, the client's built-in MCP module will proactively request a list of tool descriptions from the server, including tool names and function descriptions, thereby pre-obtaining all the capabilities provided by the server. After obtaining the list of tool descriptions, when the user initiates a query to the large language model in the Cheery Studio client, the client inputs the obtained tool information along with the user's request into the large language model to complete semantic understanding and tool selection.

[0058] In one exemplary embodiment of this disclosure, obtaining a tool description list from a Model Context Protocol (MGP) server via a MGP client includes: sending a tool invocation request to the MGP server via the MGP client; and having the MGP server return a tool description list. The tool description list includes one or more of the following metadata information: database operations, metric extraction methods, and parameter limitations.

[0059] The business indicator analysis system initiates a tool call request to the MCP server through the MCP client. After receiving the tool call request, the MCP server will return a list of tool descriptions that can be called by the MCP server. This list of tool descriptions may include, but is not limited to, metadata information such as database operations, indicator extraction methods, and parameter constraints.

[0060] Database operations refer to a series of operations such as reading, writing, updating, and deleting data in a database. Indicator extraction methods can be methods for extracting business indicator data, including but not limited to full extraction, incremental extraction, structured data extraction, and unstructured data extraction. Parameter constraints are mathematical expressions in mathematical programming that restrict the range of values ​​for decision variables. Through the above steps, the MCP client can obtain a list of callable tools from the MCP server.

[0061] In one exemplary embodiment of this disclosure, step S110, which involves performing intent parsing processing on the business indicator analysis request to obtain the business analysis intent, includes: inputting the business indicator analysis request into a pre-trained indicator analysis model; and having the indicator analysis model perform intent parsing processing on the business indicator analysis request to obtain the business analysis intent, wherein the business analysis intent includes one or more of the following: data to be analyzed, tools to be invoked, and query parameters; wherein the data to be analyzed includes one or more of the following: time range to be queried, geographical range, product identifier, and key business indicators.

[0062] Specifically, the business metric analysis request can be a user's request command sent in natural language to execute a specified business metric analysis task. The metric analysis model can be an analysis model for business metrics trained on a large language model. The business analysis intent can be the true intent obtained after the large language model processes the user's input business metric analysis request. The query time range can be the time range corresponding to the business metric to be analyzed. The geographic range can be the regional range corresponding to the business metric to be analyzed. The product identifier can be the unique identifier corresponding to the business product to be analyzed. Key business metrics can be the type of the business metric to be analyzed and other relevant information.

[0063] Continue to refer to Figure 2 In step S210, the system receives a natural language instruction from the user, which can be in text or voice form (converted to text via speech recognition). For example, the user inputs "Help me analyze the business scale and revenue of XX product in XX province for the current year"; where business scale typically includes business data such as the number of new users, the number of churned users, and the number of new users, this request is a complete business indicator query request.

[0064] In step S220, a list of callable tool descriptions is obtained as context information and sent to the large language model along with the user's original request. After obtaining the above tool description list, when the MCP client receives the user's business indicator analysis request, it can use the tool description list as context information and send it along with the user's original request (i.e., the business indicator analysis request) to the large language model for parsing.

[0065] In step S230, the large language model performs semantic understanding on the natural language request input by the user, parses the user's business intent, and determines the type of tool to be invoked based on the tool description list. After receiving the above business indicator analysis request, the request is sent to a pre-trained indicator analysis model, which is trained based on the large language model.

[0066] The indicator analysis model performs intent parsing on the business indicator analysis request to determine the corresponding business analysis intent. This intent includes multiple parameters such as region, time, product name, and indicator type. It identifies the business product to be analyzed, specific business indicators (such as new user acquisition, dismantling volume, revenue, etc.), time range, and geographical range (such as a specific province or city). By employing a large language model to perform intent recognition processing on the business indicator analysis request, the user's true intent can be determined, thus enabling subsequent data extraction operations.

[0067] For step S120, based on the business analysis intent, the target calling tool and target query parameters are determined from the pre-acquired tool description list.

[0068] In one exemplary embodiment of this disclosure, the target invocation tool can be an invocation tool corresponding to the actual business analysis intent. The target query parameters can be query parameters corresponding to the actual business analysis intent.

[0069] The large language model performs semantic understanding on the natural language requests input by users, parsing out the user's business intent, including the time range, region, product identifier, key business indicators, etc., and combines the tool description list to determine the type of tool to be called (such as business indicator query tool, revenue data extraction tool, etc.), as well as the specific parameters required (such as time range, region code, product code, indicator type, etc.).

[0070] The specific implementation process is as follows: The MCP client first calls the MCP server interface to obtain a list of registered tools, including the tool name and tool description. The MCP server is based on the MCP open-source Python Software Development Kit (SDK). Each tool on the server is registered through the @tool decorator. The @tool decorator defines the tool's name and description, and its name and function description are recorded in the name and description fields, respectively, thus providing clear calling semantics for the large language model.

[0071] After obtaining the tool list, the MCP client inputs the tool name and description along with the user's natural language request into the large language model. Based on this, the large language model completes tool selection and parameter parsing, determining the name of the target tool to be invoked and its input parameters (including specific values ​​for province, time range, product code, and indicator type). The following is an implementation method for a telecommunications business data extraction tool provided by the MCP server:

[0072] from mcp.server import FastMcp

[0073] import tool from mcp.tool

[0074] from mcp.prompt import prompt

[0075] app=FastMCP("chinatelecom mcp server")

[0076] @tool(name=”get_product_data”,description=”Search for all values ​​of a specific province, time range, product, and metric; the input parameters for province, time range, product code, and metric type are province, period, datetime, and product_id, respectively”)

[0077] def get_product_data(province:str,period:datetime,product_id:int,category:str)->List[dict]:

[0078] #Implementation code section

[0079] return product_data

[0080] The tool "get_product_data" mentioned in the code above is used to retrieve business indicator data for telecommunications service products. Once the large language model understands the meaning of the tool, it can accurately invoke the corresponding tool to obtain the required data.

[0081] In one exemplary embodiment of this disclosure, step S130, which involves performing a data extraction operation based on the target invocation tool and the target query parameters to obtain initial business indicator data, includes: the model context protocol server constructing a structured query statement based on the target invocation tool and the target query parameters; and performing a data extraction operation in a pre-built business database based on the structured query statement to obtain initial business indicator data, wherein the initial business indicator data is structured business indicator data.

[0082] The structured query statement can be a query statement with a fixed format and model; for example, it can be a Structured Query Language (SQL) statement. The initial business indicator data can be business indicator data extracted directly from the business database based on the structured query statement.

[0083] Continue to refer to Figure 2 In step S240, based on the parsed tools and parameters, the MCP client sends the standardized call request to the MCP server to perform data extraction. After completing tool selection and parameter parsing, and determining the target call tool and target query parameters, the large language model can use the above parameters to construct a structured query statement to obtain a standardized call request. The MCP client sends the standardized call request to the MCP server. The MCP server, based on the target tool name and input parameters, connects to the pre-built business database, constructs an SQL query statement, and executes the data extraction task. The extracted results are structured business indicator data, including multiple indicator fields requested by the user. Through the data extraction operation, the required business indicator data can be accurately obtained, serving as the data foundation for the final business indicator analysis results.

[0084] In one exemplary embodiment of this disclosure, step S140, generating business indicator analysis results based on initial business indicator data, includes: performing multidimensional analysis and semantic interpretation processing on the initial business indicator data based on a pre-built business knowledge base to obtain business indicator data; performing image-text conversion processing on the business indicator data to obtain business indicator analysis results; the business knowledge base includes one or more of predefined business indicator descriptions, industry knowledge graphs, and business meaning databases.

[0085] Multidimensional analysis processing can be the process of analyzing initial business indicator data from multiple dimensions. Semantic interpretation processing can be the process of interpreting the semantics of initial business indicator data by combining target business knowledge from a business knowledge base. Business indicator data can be the indicator data obtained after multidimensional analysis and semantic interpretation of the initial business indicator data. Business indicator descriptions can be explanations of each indicator item contained in the business indicator data. Industry knowledge graphs can be knowledge graphs corresponding to a specific industry. Business meaning bases can be databases constructed from the specific meanings of multiple businesses.

[0086] Continue to refer to Figure 2 In step S250, the MCP server returns the query results to the large language model data interpretation module, which, in conjunction with the business knowledge base, performs multidimensional analysis and semantic interpretation of the initial business indicator data. The interpretation module combines predefined indicator descriptions, industry knowledge graphs, and business meaning databases from the business knowledge base to perform multidimensional analysis and semantic interpretation of the business indicator data, thus obtaining the business indicator data.

[0087] In step S260, the large language model data interpretation module returns the analysis results to the user in a combined text and graphics format. Based on the user's original intent, the interpretation module generates a corresponding natural language analysis report from the business indicator data, and can optionally generate charts (bar charts, line charts, etc.) to enhance readability and expressiveness. In the specific implementation process, various document data related to the target business are first imported into the system. These document types cover common formats such as PDF, Word, Excel, and PowerPoint. Simultaneously, the large language model data interpretation module returns the analysis results to the user in a combined text and graphics format, including trend analysis of indicator data, business scale assessment, revenue contribution, etc. Through the above steps, business indicator analysis results in a specified text and graphics format can be generated for easy viewing by the user.

[0088] In one exemplary embodiment of this disclosure, content extraction is performed on the business indicator analysis results to obtain content in a specified format; the specified format content is segmented to obtain multiple content text blocks; the content text blocks are vectorized to obtain content representation vectors; and the content representation vectors are stored in a vector database, which is used to support semantic retrieval of the business indicator analysis results.

[0089] The specified format content can be business indicator analysis results presented in a specific format, such as a graphic format. The content text block can be a text block obtained by segmenting the specified format content. The content representation vector can be text content represented in a vector format. The vector database can be a database that stores content representation vectors.

[0090] After obtaining the business metric analysis results, the PyMuPDF tool is used to parse the document, extract the text content, and divide the text into appropriately sized blocks using intelligent segmentation technology to ensure contextual integrity and semantic coherence. Next, an embedding model is used to convert each text block into a high-dimensional vector representation, and these vectors are stored in a vector database to support efficient semantic retrieval. When a user initiates a query, the system finds the most relevant text blocks from the vector database based on the search requirements, combines these search results with the extracted business data, and then passes them to a large language model for multi-dimensional analysis and deep semantic interpretation. Vectorizing the business metric analysis results enables subsequent semantic retrieval of business data by users.

[0091] In summary, the business indicator analysis method disclosed herein responds to a business indicator analysis request by performing intent parsing to obtain the business analysis intent; based on the business analysis intent, it determines the target invocation tool and target query parameters from a pre-acquired tool description list; it performs data extraction operations according to the target invocation tool and target query parameters to obtain initial business indicator data; and it generates business indicator analysis results based on the initial business indicator data. On one hand, it provides a fast, accurate, and manual-intervention-free business indicator analysis solution by automatically extracting and intelligently interpreting business data according to the user's business data analysis request. On the other hand, by utilizing the intent recognition capability and tool invocation for business indicator analysis requests, it performs data extraction operations to obtain business indicator analysis results, enabling a response time of seconds to user requests and providing accurate analysis, achieving an agile "question-as-analysis" experience. Furthermore, this disclosure leverages the standardized interface of the Model Context Protocol (MCP) to give the entire system good scalability, allowing dynamic access to new MCP servers to provide new data processing tools and providing a general data processing framework.

[0092] It should be noted that although the steps of the method in this invention are described in a specific order in the accompanying drawings, this does not require or imply that the steps must be performed in that specific order, or that all the steps shown must be performed to achieve the desired result. Additional or alternative steps may be omitted, multiple steps may be combined into one step, and / or one step may be broken down into multiple steps.

[0093] Next, in this example embodiment, a business indicator analysis system is also provided, for reference. Figure 3 , Figure 3A schematic diagram of a business metrics analysis system according to an exemplary embodiment of the present disclosure is shown. The business metrics analysis system 300 may include: a large model intent parsing module 310, a model context protocol client 320, a model context protocol server 330, a large model data interpretation module 340, a business database 350, and a business knowledge base 360.

[0094] Specifically, the large model intent parsing module 310 is used to process the business indicator analysis request through the large language model to obtain the business analysis intent. Based on the business analysis intent, it determines the target tool to be invoked and the target query parameters from the tool description list. It is primarily responsible for semantic understanding and intent recognition of the analysis request input by the user through natural language, identifying the telecommunications products and specific business indicators to be analyzed. Simultaneously, it utilizes the Model Context Protocol (MCP) to obtain the structured text description provided by the MCP server, combines it with the large language model, and generates a standardized intent pattern for the MCP client to invoke.

[0095] The Model Context Protocol client (MCP client) 320 is used to generate standardized invocation requests based on the target invocation tool and target query parameters, and send the standardized invocation requests to the Model Context Protocol server. As a bridging component between the large language model and the backend business database, the MCP client constructs standardized query requests based on the structured parameters output by the large language model's intent parsing module and sends them to the MCP server.

[0096] The Model Context Protocol server (MCP server) 330 is used to perform data extraction operations based on standardized call requests to obtain initial business indicator data. The MCP server is deployed on the business database side, responsible for receiving requests from MCP clients, executing corresponding data access tasks, and returning the results to the large language model data interpretation module.

[0097] The large-scale model data interpretation module 340 is used to perform multi-dimensional analysis and semantic interpretation of initial business indicator data based on a large language model and business knowledge from a business knowledge base, thereby obtaining business indicator analysis results. Based on the large language model and combined with the telecommunications business knowledge base, it interprets the data and automatically generates structured analytical text and visualizations (such as charts and trend graphs) to help users understand the meaning and underlying reasons of business indicators.

[0098] Business database 350 is used to store all business data, which includes one or more of the following: user information data, business product information, business processing records, billing and accounting data, and user behavior data; and supports MCP server-side query and extraction operations for indicator data.

[0099] The Business Knowledge Base 360 ​​is used to provide business knowledge for the target business in the large language model. For example, it provides knowledge support in the telecommunications field for the large language model, including explanations of commonly used business terms in the telecommunications industry, indicator definitions, customer definitions, introductions to telecommunications products, billing logic for telecommunications products, business process knowledge, and other content.

[0100] This disclosure provides a business indicator analysis system based on a large language model, constructing an end-to-end closed-loop process from user requirement input to analysis result output. (Reference) Figure 4 , Figure 4 The diagram illustrates the interaction flowchart between internal modules in a business metrics analysis system according to an exemplary embodiment of the present disclosure. Users submit business metrics analysis requests via natural language, without needing to understand the complex terminology or database structure of specific business domains. The large language model parsing module 310 performs semantic parsing and intent recognition, automatically extracting key parameters.

[0101] Then, through the Model Context Protocol (MCP), the MCP client 320 and the MCP server 330 collaboratively obtain tool capability descriptions to help the large model determine which data extraction tool to use. The data extraction tool provided by the MCP server 330 connects to the business database, automatically generates and executes query statements, and obtains initial business indicator data. The obtained structured data is then passed to the large model data interpretation module 340, where it is combined with the business knowledge base for semantic interpretation and trend analysis. Finally, the analysis results are returned to the user, quickly completing an accurate business indicator analysis process without manual intervention.

[0102] Furthermore, in this example embodiment, a business metric analysis device is also provided. (See reference) Figure 5 The business indicator analysis device 500 may include: a business intent parsing module 510, a tool parameter acquisition module 520, a data extraction module 530, and an analysis result determination module 540.

[0103] Specifically, the business intent parsing module 510 is used to respond to the business indicator analysis request, perform intent parsing processing on the business indicator analysis request, and obtain the business analysis intent; the tool parameter acquisition module 520 is used to determine the target calling tool and target query parameters from the pre-acquired tool description list based on the business analysis intent; the data extraction module 530 is used to perform data extraction operations according to the target calling tool and target query parameters to obtain initial business indicator data; and the analysis result determination module 540 is used to generate business indicator analysis results based on the initial business indicator data.

[0104] In one exemplary embodiment of this disclosure, the business metric analysis device 500 further includes a tool list acquisition module, configured to: start a pre-configured Model Context Protocol (MGP) server, which provides external services through a specified port; perform connection configuration operations with the MGP server based on the MGP client to establish a communication connection with the MGP server; and acquire a tool description list from the MGP server through the MGP client.

[0105] In one exemplary embodiment of this disclosure, the tool list acquisition module includes a tool list acquisition unit, configured to: send a tool call request to the Model Context Protocol (MGP) server via a MGP client; and have the MGP server return a tool description list; the tool description list includes one or more of the following metadata information: database operations, metric extraction methods, and parameter constraints.

[0106] In one exemplary embodiment of this disclosure, the business intent parsing module 510 includes a business intent parsing unit, configured to: input a business indicator analysis request into a pre-trained indicator analysis model; and perform intent parsing processing on the business indicator analysis request by the indicator analysis model to obtain a business analysis intent, wherein the business analysis intent includes one or more of the following: data to be analyzed, tools to be invoked, and query parameters; wherein the data to be analyzed includes one or more of the following: a time range to be queried, a geographical range, a product identifier, and key business indicators.

[0107] In one exemplary embodiment of this disclosure, the data extraction module 530 includes a data extraction unit, configured to: construct a structured query statement by the model context protocol server based on the target calling tool and the target query parameters; and perform a data extraction operation in a pre-built business database based on the structured query statement to obtain initial business indicator data, wherein the initial business indicator data is structured business indicator data.

[0108] In one exemplary embodiment of this disclosure, the analysis result determination module 540 includes an analysis result determination unit, configured to: perform multidimensional analysis and semantic interpretation processing on initial business indicator data based on a pre-built business knowledge base to obtain business indicator data; perform graphic-text conversion processing on the business indicator data to obtain business indicator analysis results; the business knowledge base includes one or more of predefined business indicator descriptions, industry knowledge graphs, and business meaning databases.

[0109] In one exemplary embodiment of this disclosure, the business indicator analysis device 500 further includes a vector library construction module, used for: performing content extraction operations on the business indicator analysis results to obtain content in a specified format; performing block processing on the specified format content to obtain multiple content text blocks; performing vector conversion processing on the content text blocks to obtain content representation vectors; and storing the content representation vectors in a vector database, the vector database being used to support semantic retrieval of the business indicator analysis results.

[0110] The specific details of the virtual modules of each business indicator analysis device mentioned above have been described in detail in the corresponding business indicator analysis methods. For any undisclosed details, please refer to the implementation methods in the method section, and therefore will not be repeated here.

[0111] It should be noted that although several modules or units of the business indicator analysis device have been mentioned in the detailed description above, this division is not mandatory. In fact, according to embodiments of this disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.

[0112] An exemplary embodiment of this disclosure also provides a computer program product. The computer program product includes a computer program that, when executed by a processor, implements the aforementioned business metric analysis method.

[0113] In one implementation, the computer program product may be a tangible product containing a computer program, such as a computer-readable storage medium storing the computer program. (See reference...) Figure 6 , Figure 6 The schematic diagram illustrates a computer-readable storage medium 600 according to an exemplary embodiment of the present disclosure. The computer-readable storage medium 600 can be a storage medium based on electrical, magnetic, optical, electromagnetic, infrared, or other signals, including but not limited to: random access memory (RAM), read-only memory (ROM), magnetic tape, floppy disk, flash memory, hard disk drive (HDD), solid-state drive (SSD), etc. Exemplarily, a computer program product can be implemented as a non-volatile storage medium storing a computer program, such as read-only memory, NAND flash memory, etc.

[0114] In one implementation, the computer program product can be an intangible product containing a computer program. For example, the computer program product can be implemented as a virtual digital product, such as an executable file, installation package, or other digital file storing the computer program.

[0115] Computer program code can be written in one or more programming languages. Examples of programming languages ​​include C, Java, and C++. Program code can execute entirely on the user's computing device, partially on the user's computing device, or as a standalone software package. It can also execute partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, such as a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via an internet connection provided by a mobile network operator).

[0116] Computer programs can be carried or transmitted via signals such as electricity, magnetism, light, electromagnetic fields, and infrared radiation. Electronic devices can convert the signals carrying computer programs into digital signals, thereby running the computer programs. When a computer program runs on an electronic device, its code is used to cause the electronic device to execute (more specifically, to execute by the processor of the electronic device) the method steps of various exemplary embodiments of this disclosure, such as the aforementioned business indicator analysis method.

[0117] Exemplary embodiments of this disclosure also provide an electronic device, which may include a processor and a memory. The memory stores executable instructions of the processor, such as a computer program. The processor executes the executable instructions to perform the method steps of various exemplary embodiments of this disclosure. Furthermore, the electronic device may also include a display for displaying a graphical user interface.

[0118] The following is for reference. Figure 7 The electronic device is illustrated by way of a general-purpose computing device. It should be understood that... Figure 7 The electronic device 700 shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments disclosed herein.

[0119] like Figure 7 As shown, the electronic device 700 may include: a processor 710, a memory 720, a bus 730, an I / O (input / output) interface 740, a network adapter 750, and a display 760.

[0120] The memory 720 may include volatile memory, such as RAM 721 and cache unit 722, and may also include non-volatile memory, such as ROM 723. The memory 720 may also include one or more program modules 724, including but not limited to: an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include an implementation of a network environment. For example, program module 724 may include the modules described above.

[0121] The processor 710 may include one or more processing units, such as an AP (Application Processor), a modem processor, a GPU (Graphics Processing Unit), an ISP (Image Signal Processor), a controller, an encoder, a decoder, a DSP (Digital Signal Processor), a baseband processor, and / or an NPU (Neural-Network Processing Unit).

[0122] The processor 710 can be used to execute executable instructions stored in the memory 720, such as the aforementioned business indicator analysis method.

[0123] Bus 730 is used to connect different components of electronic device 700 and may include a data bus, an address bus and a control bus.

[0124] Electronic device 700 can communicate with one or more external devices 800 (such as keyboard, mouse, external controller, etc.) through I / O interface 740.

[0125] Electronic device 700 can communicate with one or more networks via network adapter 750. For example, network adapter 750 can provide mobile communication solutions such as 3G / 4G / 5G, or wireless communication solutions such as wireless LAN, Bluetooth, and near-field communication. Network adapter 750 can communicate with other modules of electronic device 700 via bus 730.

[0126] Electronic device 700 can display a graphical user interface via display 760, such as an interface that displays the results of business indicator analysis.

[0127] although Figure 7 As not shown in the diagram, other hardware and / or software modules may also be configured in the electronic device 700, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0128] As can be seen from the above, the technical solutions disclosed herein can be implemented as methods, apparatus, systems, computer program products, storage media, electronic devices, etc. Those skilled in the art will understand that various aspects of this disclosure can be specifically implemented in the following forms: a completely hardware implementation, a completely software implementation (including firmware, microcode, etc.), or a combination of hardware and software implementations, which may be referred to as "circuit," "module," or "system," respectively.

[0129] It should be understood that this disclosure is not limited to the specific methods, steps, or structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. Those skilled in the art will readily conceive of other embodiments based on the specific implementations provided in this disclosure. Therefore, the specific implementations provided in this disclosure are merely exemplary, and the scope and spirit of this disclosure are indicated by the claims, and should cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary technical means in the art not disclosed in this disclosure.

Claims

1. A service indicator analysis method characterized by, The method comprises the following steps: In response to a business index analysis request, performing intention analysis processing on the business index analysis request to obtain a business analysis intention; Based on the business analysis intention, determining a target calling tool and a target query parameter from a pre-acquired tool description list; Performing data extraction operations according to the target calling tool and the target query parameter to obtain initial business index data; Generating a business index analysis result based on the initial business index data.

2. The method of claim 1, wherein, The method further comprises the following steps: Starting a pre-configured model context protocol server, which provides external services through a specified port; Based on a model context protocol client, performing connection configuration operations with the model context protocol server to establish a communication connection between the model context protocol client and the model context protocol server; Obtaining the tool description list from the model context protocol server through the model context protocol client.

3. The method of claim 2, wherein, The method of obtaining the tool description list from the model context protocol server through the model context protocol client comprises the following steps: Sending a tool calling request to the model context protocol server through the model context protocol client; Returning the tool description list by the model context protocol server; the tool description list includes one or more of the following metadata information: database operations, index extraction methods, and parameter restriction conditions.

4. The method of claim 1, wherein, The method of performing intention analysis processing on the business index analysis request to obtain a business analysis intention comprises the following steps: Inputting the business index analysis request into a pre-trained index analysis model; Performing intention analysis processing on the business index analysis request by the index analysis model to obtain the business analysis intention, which includes one or more of the following: data to be analyzed, calling tools, and query parameters; The data to be analyzed includes one or more of the following: query time range, geographical range, product identifier, and key business index.

5. The method of claim 1, wherein, The method of performing data extraction operations according to the target calling tool and the target query parameter to obtain initial business index data comprises the following steps: Constructing a structured query statement according to the target calling tool and the target query parameter by the model context protocol server; Based on the structured query statement, performing data extraction operations in a pre-constructed business database to obtain the initial business index data, which is structured business index data.

6. The method of claim 1, wherein, The method of generating a business index analysis result based on the initial business index data comprises the following steps: Based on a pre-constructed business knowledge base, performing multi-dimensional analysis and semantic interpretation processing on the initial business index data to obtain business index data; Performing image-text conversion processing on the business index data to obtain the business index analysis result; The business knowledge base includes one or more of the following: pre-defined business index descriptions, industry knowledge graphs, and business meaning libraries.

7. The method according to any one of claims 1 to 6, characterized in that, The method further comprises the following steps: Performing content extraction operations on the business index analysis result to obtain specified format content; Performing block processing on the specified format content to obtain a plurality of content text blocks; The content text block is subjected to vector conversion processing to obtain a content representation vector; The content representation vector is stored in a vector database, which is used to support semantic retrieval of the business indicator analysis result.

8. A service indicator analysis system characterized by, Comprise: A large model intent analysis module is configured to analyze a business indicator analysis request by a large language model to obtain a business analysis intent, and determine a target calling tool and a target query parameter from a tool description list based on the business analysis intent; A model context protocol client is configured to generate a standardized calling request according to the target calling tool and the target query parameter, and send the standardized calling request to a model context protocol server; The model context protocol server is configured to perform a data extraction operation according to the standardized calling request to obtain initial business indicator data; A large model data interpretation module is configured to perform multi-dimensional analysis and semantic interpretation processing on the initial business indicator data based on the large language model and in combination with business knowledge of a business knowledge base to obtain a business indicator analysis result; A business database is configured to store full-amount business data, which includes one or more of user information data, business product information, business handling records, billing account data, and user behavior data; The business knowledge base is configured to provide business knowledge of a target business for the large language model.

9. A service indicator analysis apparatus characterized by comprising: Comprise: A business intent analysis module is configured to analyze a business indicator analysis request to obtain a business analysis intent in response to the business indicator analysis request; A tool parameter acquisition module is configured to determine a target calling tool and a target query parameter from a pre-acquired tool description list based on the business analysis intent; A data extraction module is configured to perform a data extraction operation according to the target calling tool and the target query parameter to obtain initial business indicator data; An analysis result determination module is configured to generate a business indicator analysis result based on the initial business indicator data.

10. An electronic device, comprising: Comprise: A processor; A memory for storing instructions executable by the processor; The processor is configured to execute the instructions to implement the business indicator analysis method of any one of claims 1 to 7.

11. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the business indicator analysis method of any one of claims 1 to 7.

Images