Information configuration method, information processing method, equipment, storage medium and program product

Abstract

The embodiment of the invention provides an information configuration method, an information processing method, equipment, a storage medium and a program product. In the embodiment of the invention, through decoupling a business index computational logic and an application system, computational logic description information expressed by a natural language is received, a corresponding relation between a business index and a data source is generated, and the business index is converted into a factor and configuration information thereof by utilizing an artificial intelligence (AI)-based language model; meanwhile, the computational logic is converted into a computational formula of a dependency factor, and finally independent and configurable computational logic configuration information is formed to be called by an application system. When the calculation logic needs to be updated, only the configuration information needs to be modified, the application system does not perceive, and synchronous upgrading or updating is not needed, so that business personnel can flexibly adjust the calculation relation between the business indexes according to actual requirements, and the flexibility and efficiency of business decision making are improved; and the dependence on developers and the adjustment cost of the application system are reduced.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and in particular, to an information configuration method, an information processing method, a device, a storage medium, and a program product. Background Art

[0002] With the continuous deepening of the digital transformation of enterprises, the calculation and analysis of relevant business indicators in application systems play an increasingly important role in decision-making support, operation optimization, and strategic planning. For example, in the e-commerce marketing scenario, the effectiveness of marketing activities can be evaluated based on business indicators such as user conversion rate, average order value, and repurchase rate, and subsequent marketing strategies can be adjusted according to the effectiveness of the marketing activities.

[0003] In the prior art, the calculation logic between business indicators is embedded in the application system in the form of hard coding. Specifically, the analysis module in the application system relies on the calculation logic between business indicators embedded in the form of hard coding to analyze the system operation situation and adjust subsequent operation strategies according to the analysis results. However, this hard coding method has obvious limitations: business personnel cannot directly adjust the calculation logic between business indicators and can only rely on developers to modify the code and redeploy the application system, which is inefficient and costly. Summary of the Invention

[0004] Multiple aspects of this application provide an information configuration method, an information processing method, a device, a storage medium, and a program product, so that business personnel can flexibly adjust the calculation relationship between business indicators according to actual needs, thereby improving the flexibility and efficiency of business decision-making, reducing the dependence on developers, and the cost of application system adjustment.

[0005] An embodiment of the present application provides an information configuration method, including: receiving at least one calculation logic description information expressed in natural language, each calculation logic description information including a business metric and a calculation logic between business metrics; performing the following operations for each calculation logic description information: generating an information description of the business metric according to the data source information of the business metric in the calculation logic description information, the information description representing the correspondence between the business metric and the data source information; according to the calculation logic description information, the information description of the business metric, and a first prompt word, calling a first language model based on artificial intelligence (AI), and under the prompt of the first prompt word, representing the business metrics in the calculation logic description information as factors, generating configuration information of the factors according to the information description of the business metric, and converting the calculation logic between the business metrics into a calculation formula depending on the factors, where a factor is an atomic representation of a business metric, and the configuration information of the factor is used to describe the manner information for reading a value for the factor; generating calculation logic configuration information corresponding to the calculation logic description information according to the calculation formula, the factors, and the configuration information of the factors, for an application system to call the calculation logic configuration information to perform analysis and processing based on the calculation logic between business metrics.

[0006] An embodiment of the present application further provides an information processing method, in which calculation logic configuration information corresponding to at least one calculation logic description information is pre-configured, and the method includes: receiving a first calculation logic description information expressed in natural language, the first calculation logic description information being any one of the at least one calculation logic description information; querying the calculation logic configuration information corresponding to each calculation logic description information according to the first calculation logic description information to obtain the first calculation logic configuration information corresponding to the first calculation logic description information; calling a parsing engine to parse the configuration information of the first factor in the first calculation logic configuration information, and obtaining the value of the first factor according to the manner information for reading a value for the first factor parsed from the configuration information of the first factor; calling a calculation engine to generate a calculation result corresponding to the first calculation logic description information according to the first calculation formula and the first factor in the first calculation logic configuration information, in combination with the value of the first factor.

[0007] An embodiment of the present application further provides an electronic device, including: a processor and a memory, the memory storing a computer program, and when the computer program is executed by the processor, causing the processor to be able to implement any step in the information configuration method and the information processing method.

[0008] An embodiment of the present application further provides a computer-readable storage medium storing a computer program, and when the computer program is executed by a processor, causing the processor to be able to implement any step in the information configuration method and the information processing method provided by the embodiment of the present application.

[0009] The embodiments of the present application also provide a computer program product, including computer programs / instructions, which, when executed by a processor, enable the processor to implement any step of the information configuration method and the information processing method provided by the embodiments of the present application.

[0010] In the embodiments of the present application, by decoupling the business metric calculation logic from the application system, receiving the calculation logic description information expressed in natural language, generating the correspondence between business metrics and data sources, and using a language model based on artificial intelligence (AI) to convert business metrics into factors and their configuration information, while converting the calculation logic into a calculation formula depending on factors, an independent and configurable calculation logic configuration information is finally formed for the application system to call. When it is necessary to update the calculation logic, only the configuration information needs to be modified, and the application system is unaware, without the need for synchronous upgrade or update. Thus, business personnel can flexibly adjust the calculation relationship between business metrics according to actual needs, improving the flexibility and efficiency of business decision-making, reducing the dependence on developers, and the cost of application system adjustment. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] The drawings described herein are used to provide a further understanding of the present application, and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application, and do not constitute an improper limitation to the present application. In the drawings:

[0012] Figure 1 is a schematic flowchart of an information configuration method provided by an exemplary embodiment of the present application;

[0013] Figure 2a is a schematic flowchart of a process for generating an information description of business metrics provided by an exemplary embodiment of the present application;

[0014] Figure 2b is a schematic flowchart of another process for generating an information description of business metrics provided by an exemplary embodiment of the present application;

[0015] Figure 2c is a schematic diagram of the information type of data sources corresponding to business metrics provided by an exemplary embodiment of the present application;

[0016] Figure 3a is a schematic flowchart of another information configuration method provided by an exemplary embodiment of the present application;

[0017] Figure 3b is a schematic diagram of the structure of a calculation formula provided by an exemplary embodiment of the present application;

[0018] Figure 4 is a schematic flowchart of yet another information configuration method provided by an exemplary embodiment of the present application;

[0019] Figure 5A flowchart of an information processing method provided by an exemplary embodiment of the present application;

[0020] Figure 6 A schematic structural diagram of an electronic device provided by an exemplary embodiment of the present application. Detailed implementation manners

[0021] To make the objectives, technical solutions, and advantages of the present application clearer, the technical solutions of the present application will be clearly and completely described below in conjunction with specific embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0022] It should be noted that in the case where the embodiments of the present application involve user information, the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the embodiments of the present application are all information and data that have been authorized by the user or fully authorized by all parties. Moreover, the collection, use, and processing of relevant data need to comply with relevant laws, regulations, and standards of relevant countries and regions, and corresponding operation entrances are provided for users to choose to authorize or refuse. In addition, various models involved in the present application (including but not limited to language models or large models) comply with relevant laws and standards.

[0023] In the prior art, the calculation logic between business metrics is usually embedded in the application system in the form of hard coding, resulting in business personnel being unable to directly adjust the calculation logic and having to rely on developers to modify the code and redeploy the system, which is not only inefficient but also costly. In addition, when the calculation logic needs to be updated, the entire application system may need to be synchronously upgraded or adjusted. This tightly coupled method severely limits the flexibility of business decision-making and the efficiency of the system. To solve the above technical problems, in the embodiments of the present application, by decoupling the calculation logic of business metrics from the application system, receiving calculation logic description information expressed in natural language, generating the correspondence between business metrics and data sources, and using a language model based on artificial intelligence (AI) to convert business metrics into factors and their configuration information, while converting the calculation logic into a calculation formula depending on the factors, an independent and configurable calculation logic configuration information is finally formed for the application system to call. When the calculation logic needs to be updated, only the configuration information needs to be modified, and the application system is unaware and does not need to be synchronously upgraded or updated. Thus, business personnel can flexibly adjust the calculation relationship between business metrics according to actual needs, improving the flexibility and efficiency of business decision-making, reducing the dependence on developers, and the cost of adjusting the application system.

[0024] The technical solutions provided by the embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0025] Figure 1 It is a schematic flowchart of an information configuration method provided by an exemplary embodiment of the present application. As Figure 1 shown, this method is applied to an information configuration device, and the information configuration method includes:

[0026] 11. Receive at least one computational logic description information expressed in natural language, and each computational logic description information includes business metrics and the computational logic between business metrics;

[0027] 12. Generate an information description of the business metric according to the data source information of the business metric in the computational logic description information;

[0028] 13. According to the computational logic description information, the information description of the business metric, and the first prompt word, call the first language model based on artificial intelligence (AI). Under the prompt of the first prompt word, represent the business metrics in the computational logic description information as factors, generate configuration information of the factors according to the information description of the business metric, and convert the computational logic between business metrics into a calculation formula depending on the factors;

[0029] 14. Generate computational logic configuration information corresponding to the computational logic description information according to the calculation formula, factors, and configuration information of the factors, so that the application system can call the computational logic configuration information to perform analysis and processing based on the computational logic between business metrics.

[0030] In the embodiments of the present application, the specific form or deployment method of the information configuration device is not limited. Any device that has the capabilities of data reception, processing, and output and can implement the information configuration method can be used as the information configuration device in the embodiments of the present application. For example, the information configuration device can be a server deployed in a data center, server cluster, local computer room, or cloud, etc., which can be a cloud server or a conventional server; it can also be a personal computer, mobile device, or other terminal devices with computing capabilities.

[0031] In the embodiments of the present application, the business refers to the specific activities or service areas carried out by an enterprise or organization during the operation of the application system. The type of business is not limited in the present application. For example, the business can be e-commerce transactions, recruitment services, real estate agency services, online education services, etc. These businesses involve multiple links and processes and require a series of corresponding business metrics to measure the performance and effect of the business.

[0032] In the embodiments of the present application, a business indicator is key data information used to quantify or describe the characteristics of business activities, which is a direct manifestation of the business operation results, is closely related to the business, and can reflect the specific status and achievements of the business. In the embodiments of the present application, the type of business indicator is not limited. For example, in the recruitment business, business indicators may include the number of resumes, the number of interview invitations, the number of hired people, etc.; in the e-commerce business, business indicators may include sales volume, order quantity, return rate, etc.

[0033] In the embodiments of the present application, the calculation logic description information refers to the calculation logic relationship between business indicators described in the form of natural language. The calculation logic description information is expressed in natural language, enabling non-technical personnel to easily define the calculation logic between complex business indicators without having to delve into programming or technical details. For example, an example of calculation logic description information described in natural language is: The comprehensive performance is composed of dividing the GMV amount by the GMV target, multiplying by the ratio of 0.5, adding the resume data divided by the resume target, multiplying by the ratio of 0.3, and then adding the number of completed leads divided by the total number of leads, multiplying by the ratio of 0.2. This way of expressing calculation logic description information not only clearly expresses the business indicators and the calculation logic between them, but also reflects the contribution degree of different business indicators to the comprehensive performance through the way of weight allocation.

[0034] In the embodiments of the present application, the input method of the calculation logic description information is not limited. For example, the calculation logic description information can be manually input by business personnel into the application system, can also be read from a specified file (such as an Excel file stored locally), or can be downloaded from the server (such as a cloud server). In addition, the calculation logic description information can also be synchronously obtained through an API interface from other systems (such as a data warehouse).

[0035] In the embodiments of the present application, the data source type of the service metrics in the calculation logic description information is not limited. As long as the specific value of the service metrics can be obtained through the means supported by the application system, it can be regarded as the data source. The data source of the service metrics in the calculation logic description information refers to the way where the specific value or information of the service metrics is stored or obtained. The data source provides actual numerical support for the service metrics, enabling the calculation logic to perform operations and analyses based on real data. For example, the data source can be a database. The service metrics can be stored in a relational database or a non-relational database. For example, a third-party database such as MySQL, PostgreSQL, or MongoDB can be used, or a self-built database can also be used. The data source can also be a service interface. The service interface can be an RPC interface (Remote Procedure Call), that is, data is obtained from other services or systems through remote procedure call (RPC). The service interface can also be an HTTP (HyperText Transfer Protocol) request, and data is obtained from a third-party service (such as an e-commerce platform, an advertising system) by calling an external API interface (Application Programming Interface). In addition, the data source can also include forms such as local files, real-time data streams, and cache systems.

[0036] In the embodiments of the present application, the information description of the service metrics is generated according to the data source information of the service metrics in the calculation logic description information, expressing the corresponding relationship between the service metrics and the data source information. The data source information refers to the detailed description used to identify and locate the storage location of the specific value or information of the service metrics, providing the application system with the path and method to obtain the value of the service metrics. For example, in a database, the data source information can include the type of the database (such as MySQL, PostgreSQL, or MongoDB), the name of the database, the table name, and the field name. These information can enable the application system to accurately extract the required value of the service metrics from the database. If the data source is an RPC interface, the data source information can include the URL (Uniform Resource Locator) of the interface, the input parameter information, and the output parameter information. For an HTTP request, the data source information can include the URL of the API, the request method, the input parameter information, and the output parameter information.

[0037] In the embodiments of the present application, the first language model based on artificial intelligence AI refers to a large-scale language model that has been trained and can understand and process natural language inputs and generate results in a format that meets the expected output. Specifically, the first language model can, under the prompt of the first prompt word, represent the business metrics in the computational logic description information as factors, generate the configuration information of the factors according to the information description of the business metrics, and convert the computational logic between the business metrics into a calculation formula that depends on the factors, thereby realizing the automated conversion from natural language to computational logic.

[0038] In the embodiments of the present application, the first prompt word refers to a set of texts or structured information that provides guiding inputs for the first language model. Its role is to help the first language model understand the natural language description content provided by business personnel and generate the corresponding calculation formula and the configuration information of the factors. By being embedded in the computational logic description information, it is used to guide the first language model to perform parsing and generation actions, ensuring the accuracy and consistency of the parsing process.

[0039] In an alternative embodiment, the first prompt word can be a structured prompt word, which clarifies the role positioning of the first language model as a "computational logic parsing expert", requiring it to have strong natural language understanding capabilities to accurately identify key elements such as business metrics, computational logic, and data sources. At the same time, the structured prompt word also stipulates the step logic that the first language model needs to follow. For example, extracting business metrics and their computational logic relationships from the computational logic description information and converting the business metrics and their computational logic relationships into calculation formulas. In addition, the structured prompt word further standardizes the types and formats of the result information that the first language model needs to generate, including calculation formulas presented in the form of mathematical expressions, configuration information of factors presented in the form of structured data (such as JSON or XML), and other auxiliary information (such as question prompts or improvement suggestions). This design not only clarifies the task scope and output standards of the first language model but also improves the intelligent level and business adaptation ability of the application system, providing a more efficient and convenient support tool for business personnel.

[0040] In the embodiments of the present application, a factor refers to the smallest computational unit after the atomic abstraction of a business metric and is the most basic component in the computational logic, used to represent the specific information of the business metric. Each factor corresponds to a specific business metric, and through its configuration information, it can determine how to obtain values from the actual data source. For example, a factor can be an English letter, a symbol, a variable name, or other identifiers. The specific form depends on the expression requirements of the computational logic and the design specifications of the application system, and the embodiments of the present application do not limit this. For example, assuming that the business metrics "GMV amount" and "GMV target" are abstracted as factors, they can be represented as: GMV amount = A, GMV target = B.

[0041] In the embodiments of this application, the configuration information of a factor refers to the information on how to read the value of the factor. Its structure is that the configuration information of the factor is stored in the database and is used to guide the application system on how to read the specific data of the factor from the specified data source. For example, if the factor is "sales amount", its configuration information may include: the data source type is "MySQL database", the database name is "sales_db", the table name is "sales_data", the field name is "amount", and the query condition is "date BETWEEN '2024-01-01' AND '2024-01-31'".

[0042] In the embodiments of this application, the calculation formula of a dependent factor refers to the conversion of the calculation logic between business metrics into a mathematical expression based on factors. The calculation formula defines how to calculate the final result according to the actual values of the factors by combining factors, mathematical operators, and functions. For example, assuming the business logic is described as "performance consists of 50% of the GMV completion ratio, 30% of the resume completion ratio, and 20% of the lead completion ratio", the calculation formula of the dependent factor can be expressed as performance = 0.5×(A / B)+0.3×(C / D)+0.2×(E / F). Here, the letters A - F represent the factors corresponding to different business metrics. The business metric corresponding to A is the GMV amount, the business metric corresponding to B is the GMV target, the business metric corresponding to C is the resume data, the business metric corresponding to D is the resume target, the business metric corresponding to E is the number of completed leads, and the business metric corresponding to F is the total number of leads.

[0043] In the embodiments of this application, the calculation logic configuration information refers to a set of complete configuration data corresponding to the calculation logic description information generated based on the calculation formula, factors, and the configuration information of the factors. The generated calculation logic configuration information is stored in the database for the application system to call to execute specific calculation tasks, providing standardized input for the calculation engine of the application system to ensure that the calculation logic can be correctly parsed and executed.

[0044] Among them, the calculation logic configuration information includes the detailed configuration required to implement the calculation logic, such as the configuration information of factors, calculation formulas, and the corresponding relationships between factors and business indicators. For example, if the business logic is described as "Performance consists of 50% of the GMV completion ratio, 30% of the resume completion ratio, and 20% of the lead completion ratio", the calculation logic configuration information can be expressed as "Calculation formula: Performance = 0.5×(A / B)+0.3×(C / D)+0.2×(E / F), where A = GMV amount, B = GMV target, C = resume data, D = resume target, E = number of leads completed, F = total number of leads; Configuration information of factors: A: {"factor_name": "GMV amount", "data_source_type": "database", "database_name": "sales_db", "query_statement": "SELECT amount AS gmv_amount FROM gmv_table WHERE date BETWEEN '2024-01-01' AND '2024-01-31'"} (From the table gmv_table in the database sales_db, select the field amount and alias it as gmv_amount, and at the same time, the screening condition is that the date field date is between January 1, 2024, and January 31, 2024. Its meaning is to obtain the value of the GMV amount in January 2024.). The configuration information of factors B, C, D, E, and F is in the same format as that of factor A above, and will not be shown in detail here.

[0045] The independently configurable calculation logic configuration information allows the application system, when it needs to update the calculation logic, to simply modify the configuration information without the application system being aware, without the need for synchronous upgrades or updates. This enables business personnel to flexibly adjust the calculation relationships between business indicators according to actual needs, improving the flexibility and efficiency of business decision-making, reducing dependence on developers, and the cost of application system adjustment.

[0046] Figure 2a This is a schematic flowchart of a process for generating an information description of business indicators provided by an exemplary embodiment of the present application. As Figure 2a shown, when a user (such as a business person) needs to define a certain business indicator, they can trigger the interactive interface of the system by performing a first input operation to input a specific identifier, and then select the data source of the business indicator, and then add data source information. The data source information refers to the way where the specific value or information of the business indicator is stored or obtained, which provides actual numerical support for the business indicator, enabling the calculation logic to perform operations and analyses based on real data.

[0047] Figure 2bAnother process schematic diagram for generating information descriptions of business metrics provided by an exemplary embodiment of this application. As Figure 2b shown, a user (such as a businessperson) inputs calculation logic description information, determines at least one business metric included in the calculation logic description information, and generates an information description of the business metric according to the data source information of the business metric in the calculation logic description information, including: for any business metric, in response to a first input operation of a specific identifier, select the data source information type corresponding to the business metric from multiple data source information types. According to the data source information, the system can generate an information description of the business metric, clarify the correspondence between the business metric and the data source, and provide a basis for subsequent calculation logic configuration and execution.

[0048] In the embodiments of this application, the specific form of the specific identifier is not limited. The specific identifier can be "@" or "#". For example, for the business metric "GMV amount" in the calculation logic description information, when "@" is entered in the input box, the system will recognize the identifier and pop up an interactive window to prompt the user to select the specific data source information type of the business metric. Through the above method, the user can select the data source information type of the current business metric from multiple options, thus laying a foundation for subsequent detailed configuration.

[0049] Figure 2c A schematic diagram of the data source information type corresponding to the business metric provided by an exemplary embodiment of this application. As Figure 2c shown, the specific type of the data source information type corresponding to the business metric is not limited. The data source information type refers to the classification method for describing the storage or acquisition path of business metric values. For example, the data source information type can be a database type or other types such as a service interface type. Among them, the service interface type can be further divided into an RPC interface and an HTTP request, etc.

[0050] Optionally, if the data source information type corresponding to the business metric is a database type, determine the target database adapted to the business metric, and generate an information description of the business metric according to the target data table and target field in the target database adapted to the business metric. As Figure 2bAs shown, determine the target database adapted to the business metrics, including: responding to the second input operation for a specific identifier, presenting at least one candidate database; responding to the first selection operation, taking the selected candidate database from the at least one candidate database as the target database; generating an information description of the business metrics according to the target data table and target fields adapted to the business metrics in the target database, including: responding to the third input operation for a specific identifier, presenting the data tables in the target database and the fields in the data tables; responding to the second selection operation, selecting the target data table and target fields, generating query condition description information adapted to the target data table and target fields; responding to the fourth input operation for a specific identifier, presenting the data tables in the target database and the fields in the data tables; responding to the second selection operation, selecting the target data table and target fields, generating query result description information adapted to the target data table and target fields; generating an information description of the business metrics according to the query condition description information and the query result description information.

[0051] Among them, when the user needs to define a certain business metric and the data source information type of the business metric is of the database type, the system will present at least one candidate database for the user to choose according to the pre-configured database connection information, and the user can select the target database adapted to the current business metric. The target database contains at least one data table and at least one field. The target data table refers to the specific data table that stores the data related to the business metric when the data source of the business metric is of the database type. It is a part of the logical organization in the database, composed of rows and columns. Each row represents a record, and each column represents an attribute or field. The selection of the target data table determines which table the system extracts the values required for the business metric from. The target field refers to the field in the target data table used to store the specific value of a specific business metric. It is a certain column in the target data table. The target field clarifies the specific position of the business metric in the data table, ensuring that the system can accurately read the required data. The query condition description information adapted to the target data table and target fields refers to the description information used to filter the records that meet specific conditions from the target data table. Repeatedly execute the response to the second selection operation, select the target data table and target fields, and obtain the final query condition description information. The query result description information adapted to the target data table and target fields refers to the description of the query result. Repeatedly execute the response to the second selection operation, select the target data table and target fields, and obtain the final query result description information. For example, the query condition can be "the date is between January 1, 2024 and January 31, 2024", which is used to limit the time range of the data; the query result description information can be "return the sum of the field amount", which is used to define which data to extract from the target data table as the value of the business metric.

[0052] For example, if the business metric is "GMV amount", the user performs a second input operation "@". From at least one candidate database provided by the system, in response to a first selection operation, the target database selected is sales_db which stores sales data. After the user selects the target database, the system responds to the user's third input operation "@" and further displays all the data tables and field information in the target database. The system responds to a second selection operation where the user selects the specific data table gmv_table and field amount related to the business metric as the target data table and target field. Based on the above-selected target data table and target field, the system generates query condition description information adapted to the target data table (gmv_table) and target field (amount): date = '2023-10-01'. The user continues to perform a fourth input operation "@". The system responds to the fourth input operation "@" and further displays all the data tables and field information in the target database. The system responds to the second selection operation to select the specific data table gmv_table and field amount related to the business metric. Based on the above target data table and target field, the system generates query result description information adapted to the target data table (gmv_table) and target field (amount): only the amount field needs to be returned, without other irrelevant fields. According to the query condition description information and query result description information, an information description of the business metric is generated. For the GMV amount, the information description is: "GMV amount, database sales_db, table gmv_table, field amount, query condition: date = '2023-10-01', query result description information: only the amount field needs to be returned, without other irrelevant fields."

[0053] Optionally, if the data source information type corresponding to the business metric is a service interface type, then a target service interface adapted to the business metric is determined, and based on the interface information of the target service interface, an information description of the business metric is generated. As Figure 2b shown, determining the target service interface adapted to the business metric includes: responding to a fifth input operation with a specific identifier to display at least one candidate service interface; responding to a third selection operation to use the selected candidate service interface among the at least one candidate service interfaces as the target service interface; generating an information description of the business metric based on the interface information of the target service interface, including: responding to a sixth input operation with a specific identifier to obtain the input parameter information of the target service interface entered by the user; responding to a seventh input operation with a specific identifier to obtain the output parameter information of the target service interface entered by the user; generating an information description of the business metric based on the input parameter information and the output parameter information.

[0054] Among them, if the data source information type corresponding to the business indicator is the service interface type, the system will display at least one candidate service interface for the user to select according to the pre-configured service interface connection information, and the user can select the target service interface that adapts to the current business indicator. A candidate service interface refers to a list of service interfaces available for selection displayed by the system according to business requirements, and the candidate service interfaces can come from internal systems or external services. A target service interface refers to a specific interface selected from the candidate service interfaces that can provide the value of the required business indicator. Input parameter information refers to the input parameters that need to be provided when calling the target service interface, and the input parameter information is used to specify the conditions or scope of the interface call. Repeat the sixth input operation in response to a specific identifier to obtain the final input parameter information. Output parameter information refers to the output parameters related to the business indicator in the data returned by the target service interface. Repeat the seventh input operation in response to a specific identifier to obtain the final output parameter information. For example, the input parameter information can be {"date_range": "2024-01-01 to 2024-01-31"}, which is used to specify the query time range; the output parameter information can be {"activity": "value"}, where activity is the specific value of the business indicator returned by the interface.

[0055] For example, if the business indicator is "resume data", the user performs the fifth input operation "@", and from at least one candidate service interface provided by the system, responds to the third selection operation to select the target service interface as / resume-service that provides resume-related data; after the user selects the target service interface, the system responds to the user's sixth input operation "@" to obtain the input parameter information of the target service interface entered by the user, that is, the user can enter the input parameter information of the target service interface / resume-service as {user_id: 123}, indicating that the resume data of user ID 123 needs to be queried, and repeat this step to obtain the final input parameter information; then, the system responds to the user's seventh input operation "@" to obtain the output parameter information of the target service interface entered by the user, that is, the user can specify the output parameter information of the target service interface / resume-service as the JSON path data.count, indicating that the value of the count field in the returned data needs to be extracted as the value of the business indicator, and repeat this step to obtain the final output parameter information; after the user completes the input of the input parameter information and the output parameter information, the system generates an information description of the business indicator according to the obtained input parameter information and output parameter information above. For example, for the business indicator "resume data", the information description is: "Resume data, service interface / resume-service, input parameter information is {user_id: 123}, output parameter information is JSON path data.count."

[0056] Figure 3aSchematic flowchart of another information configuration method provided for an exemplary embodiment of the present application. As Figure 3a shown, after the user inputs the calculation logic description information and obtains the information description of the business metrics, based on the calculation logic description information, the information description of the business metrics, and the first prompt word, the first language model based on artificial intelligence (AI) is called. Under the prompt of the first prompt word, the business metrics in the calculation logic description information are represented as factors, the configuration information of the factors is generated according to the information description of the business metrics, and the calculation logic between the business metrics is converted into a calculation formula depending on the factors, including: inputting the calculation logic description information, the information description of the business metrics, and the first prompt word into the first language model, and under the prompt of the first prompt word, parsing the calculation logic description information to obtain the business metrics, mathematical operators, and the logical relationship between the business metrics and the mathematical operators; representing the business metrics as factors respectively, and combining the factors with the data operators according to the logical relationship between the business metrics and the mathematical operators to obtain the calculation formula.

[0057] Among them, the calculation logic description information, the information description of the business metrics, and the first prompt word are input. The calculation logic description information provided by the user in natural language form (such as "Performance is obtained by dividing the GMV amount by the GMV target, multiplying by the ratio of 0.5, and adding the resume data divided by the resume target, multiplying by the ratio of 0.3") is received by the system and input into the first language model together with the information description of the business metrics (such as the GMV amount is from the field amount in the table gmv_table of the database sales_db) and the first prompt word (used to guide the language model to correctly parse the input content). Then, the calculation logic description information is parsed, and the first language model performs semantic analysis and structured processing on the calculation logic description information under the guidance of the first prompt word. Through this process, the system can extract the key elements in the calculation logic, including business metrics (such as GMV amount, GMV target, resume data, etc.), mathematical operators (such as addition, subtraction, multiplication, division), and their logical relationships (such as GMV amount divided by GMV target, resume data divided by resume target, etc.). The extraction results lay the foundation for subsequent factor representation and formula generation. Representing the business metrics as factors and generating the calculation formula, the system abstracts the extracted business metrics as factors respectively (such as A represents the GMV amount, B represents the GMV target), and combines the factors with the data operators according to the logical relationship between the business metrics and the mathematical operators, so as to obtain the calculation formula depending on the factors.

[0058] Figure 3b Schematic structural diagram of the calculation formula provided for an exemplary embodiment of the present application. As Figure 3bAs shown, the structure of the calculation formula can be: calculation result = factor A (+ / x / ÷ / =) factor B (< / &) factor C (function).... Among them, the calculation result is the result obtained by performing mathematical operations between at least one factor. For example, the above calculation logic "performance is calculated by dividing the GMV amount by the GMV target, multiplying by a ratio of 0.5, adding the resume data divided by the resume target, and multiplying by a ratio of 0.3" can be converted into the calculation formula: performance = 0.5×(A / B)+0.3×(C / D).

[0059] In an optional embodiment, after the calculation logic description information, the information description of the business indicator, and the first prompt word are input into the first language model, the first language model can gradually parse and process the input content at different levels through a hierarchical semantic parsing method, and finally generate a calculation formula. Specifically, the first language model can include an input layer, an encoding layer, an intermediate layer, and an output layer. The calculation logic description information (such as "performance is calculated by dividing the GMV amount by the GMV target, multiplying by a ratio of 0.5, adding the resume data divided by the resume target, and multiplying by a ratio of 0.3"), the information description of the business indicator (such as "the GMV amount is from the field amount in the table gmv_table of the database sales_db"), and the first prompt word (such as "parse the calculation logic") are input into the input layer. In the input layer, word segmentation, part-of-speech tagging, and vectorization operations are performed to convert the natural language text into an embedding vector that can be processed by the first natural language model. The result obtained is the embedding vector, which is used to represent the local semantic information of the input text and identify the key business indicators (such as "GMV amount", "GMV target", "resume data") and mathematical operators (such as "divide by", "multiply by"). The embedding vector obtained from the input layer is input into the encoding layer, and the embedding vector is encoded through a multi-layer neural network (such as the Transformer architecture) to extract deeper semantic information. The result obtained is the global semantic structure, and the first natural language model can understand the logical relationship between business indicators (such as "GMV amount divided by GMV target"). The semantic information extracted by the encoding layer is input into the intermediate layer, and logical relationship mapping and factor abstraction operations are performed. The result obtained is to abstract the business indicators into factors (such as abstracting "GMV amount" as factor A and "GMV target" as factor B), and identify the operation relationship between them (such as A / B). The factors and logical relationships generated by the intermediate layer are input into the output layer, and formula combination and optimization operations are performed. The result obtained is the final calculation formula (such as performance = 0.5×(A / B)+0.3×(C / D)).

[0060] Optionally, after the calculation logic description information, the information description of business metrics, and the first prompt are input into the first language model, the first language model can also gradually convert the calculation logic described in natural language into a calculation formula through a multi-stage feature extraction and step-by-step parsing method. Specifically, the first natural language model can include an input layer, a feature extraction layer, a logic parsing layer, and a formula generation layer. Among them, the input layer is the same as the input layer method description in the above hierarchical semantic parsing method and will not be elaborated here. The embedding vector obtained by the input layer is input into the feature extraction layer, and a multi-layer neural network (such as Bi-LSTM or Transformer Encoder) is used to perform feature extraction operations on the embedding vector to extract the feature information of key business metrics and operators. The result obtained is a feature vector. The model can identify business metrics (such as "GMV amount", "GMV target") and their corresponding data source information (such as database tables and fields), and extract the features of mathematical operators (such as "divide by", "multiply by"). The feature vector obtained by the feature extraction layer is input into the logic parsing layer for logical relationship parsing operations. The result obtained is the logical relationship between business metrics (such as "GMV amount divided by GMV target"), and the business metrics are abstracted into factors (such as abstracting "GMV amount" as factor A and "GMV target" as factor B), and at the same time, their operation relationship (such as A / B) is identified. The factors and logical relationships generated by the logic parsing layer are input into the formula generation layer for formula combination and optimization operations. The result obtained is the final calculation formula (such as performance = 0.5×(A / B)+0.3×(C / D)).

[0061] As Figure 3a shown, according to the calculation logic description information, the information description of business metrics, and the first prompt, the first language model based on artificial intelligence AI is called. Under the prompt of the first prompt, the business metrics in the calculation logic description information are represented as factors, and according to the information description of business metrics, the configuration information of factors is generated. Among them, according to the different data source information of business metrics, the way of generating the configuration information of factors will be different. The following is a description by case:

[0062] Case 1: When the data source information of the business metric is the target database, the target database includes the target data table and target fields adapted to the business metric. The information description of the business metric includes the query condition description information and query result description information adapted to the target data table and target fields. Then, according to the query condition description information and query result description information, the database statement required to query the target data table and target fields from the target database is generated, and combined with the identifier of the target database, it is used as the configuration information of the factor. The identifier of the target database refers to the information used to uniquely identify and locate the target database, ensuring that the system can accurately connect to the correct database to obtain the required data.

[0063] For example, the calculation logic description information is "Performance is calculated as GMV amount divided by the GMV target, multiplied by a ratio of 0.5, plus resume data divided by the resume target, multiplied by a ratio of 0.3". Among them, the information description of the business indicator as the GMV amount is "The GMV amount is sourced from the field 'amount' in the table 'gmv_table' of the database'sales_db', and the query condition is 'date BETWEEN '2024-01-01' AND '2024-01-31'". The first prompt is "Parse the calculation logic and generate factor configurations". Based on this information, the system identifies the target database (sales_db), target data table (gmv_table), and target field (amount) for the business indicator of GMV amount, and generates a database query statement for the business indicator of GMV amount according to the query condition: SELECT amount AS gmv_amount

[0064] FROM gmv_table

[0065] WHERE date BETWEEN '2024-01-01' AND '2024-01-31';

[0066] Combined with the identifier of the target database, generate the configuration information of the factor for the business indicator of GMV amount: {"factor_name": "GMV amount", "data_source_type": "database", "database_name": "sales_db", "query_statement": "SELECT amount AS gmv_amount FROM gmv_table WHERE date BETWEEN '2024-01-01' AND '2024-01-31'"}. The configuration information of the factors for the GMV target, resume data, and resume target is in the same format as the above configuration information of the factor for GMV amount, which will not be elaborated here.

[0067] Case 2: When the data source information of the business metric is the target service interface, the information description of the business metric includes the input parameter information and output parameter information of the target service interface. Then, the access path information required to access the target service interface is generated based on the input parameter information and output parameter information and used as the configuration information of the factor. Among them, when the data source of the business metric is the target service interface, the system also generates the configuration information of the factor according to the information description of the business metric. For example, the calculation logic description information is "Performance is equal to user activity multiplied by weight 0.5 plus user retention rate multiplied by weight 0.5", the information description of the business metric is "User activity is sourced from the RPC interface getUserActivity, the input parameter is {"date_range": "2024-01-01 to 2024-01-31"}, and the output parameter is {"activity": "value"}", and the first prompt word is "Parse the calculation logic and generate factor configuration". Based on this information, the system identifies the target service interface (getUserActivity) and its input and output parameter information, and generates the access path information: http: / / api.example.com / getUserActivity?date_range=2024-01-01%20to%202024-01-31. The generated configuration information of the factor is as follows: {"factor_name": "User activity", "data_source_type": "service_interface", "interface_name": "getUserActivity", "access_path": "http: / / api.example.com / getUserActivity?date_range=2024-01-01%20to%202024-01-31", "input_params": {"date_range": "2024-01-01 to 2024-01-31"}, "output_param": "activity"} (This JSON describes a factor named "User activity" whose data is sourced from a service interface. By calling the http: / / api.example.com / getUserActivity interface and passing the parameter date_range=2024-01-01 to 2024-01-31, the user activity data for January 2024 can be obtained. In the result returned by the interface, the field activity contains the specific user activity value, and the system will extract this field as the actual value of the factor for subsequent calculation or analysis).

[0068] In an optional embodiment, the first language model includes an input layer, an encoding layer, a parsing layer, and an output layer. The process of generating the configuration information of the factor in combination with this model structure is as follows: Taking the data source type as a database as an example, the calculation logic description information (such as "Performance is calculated by dividing the GMV amount by the GMV target and multiplying by a ratio of 0.5, plus dividing the resume data by the resume target and multiplying by a ratio of 0.3"), the information description of business metrics (such as "The GMV amount is sourced from the field amount in the table gmv_table of the database sales_db, and the query condition is date BETWEEN '2024-01-01' AND '2024-01-31'"), and the first prompt (such as "Parse the calculation logic and generate factor configuration") are input into the input layer of the first language model. Tokenization, part-of-speech tagging, and vectorization operations are performed to obtain embedding vectors, and the input text is preprocessed to extract local semantic information, identifying key business metrics (such as "GMV amount", "GMV target") and mathematical operators (such as "divide by", "multiply by"); then the embedding vectors obtained from the input layer are input into the encoding layer, and the embedding vectors are encoded through a multi-layer neural network (such as the Transformer architecture) to extract deeper semantic information. The encoding layer understands the logical relationship between business metrics (such as "GMV amount divided by GMV target") and identifies the data source type of the business metrics (such as the target database or target service interface); then the semantic information extracted by the encoding layer is input into the parsing layer for parsing operations of business metrics. If the data source of the business metric is the target database, the target database (such as sales_db), the target data table (such as gmv_table), and the target field (such as amount) are identified. At the same time, according to the information description of the business metric, the query condition description information (such as date BETWEEN '2024-01-01' AND '2024-01-31') and the query result description information (such as "Return the sum of the field amount") are extracted, and the parsing layer converts the information description of the business metric into specific query conditions and result descriptions; finally, the business metric information obtained from the parsing layer is input into the output layer to generate the database statement required to query the target data table and target field from the target database. For example, generate the SQL statement:

[0069] SELECT SUM(amount)AS gmv_amount

[0070] FROM gmv_table

[0071] WHERE date BETWEEN '2024-01-01' AND '2024-01-31'; Combine the identifier of the target database (such as the database name sales_db or connection information) to generate the configuration information of the factor: {"factor_name": "GMV amount", "data_source_type": "database", "database_name": "sales_db", "query_statement": "SELECT SUM(amount) AS gmv_amount FROM gmv_table WHERE date BETWEEN '2024-01-01' AND '2024-01-31'"}.

[0072] Optionally, if the data source type is a service interface type, the processing steps are the same as those when the data source is a target database, except that different input contents result in different output results. For example, input the calculation logic description information (such as "Performance is calculated by multiplying user activity by a weight of 0.5 and adding user retention rate multiplied by a weight of 0.5"), the information description of business metrics (such as "User activity is sourced from the RPC interface getUserActivity, with the input parameter {"date_range": "2024-01-01 to 2024-01-31"} and the output parameter {"activity": "value"}) and the first prompt word (such as "Parse the calculation logic and generate factor configuration") into the input layer of the first language model; input the above input content into the input layer, perform word segmentation, part-of-speech tagging, and vectorization operations to obtain embedding vectors. The input layer preprocesses the input text, extracts local semantic information, and identifies key business metrics (such as "user activity", "user retention rate") and mathematical operators (such as "multiply", "add"); input the embedding vectors obtained from the input layer into the encoding layer, and perform encoding operations on the embedding vectors through a multi-layer neural network (such as the Transformer architecture). The encoding layer is to understand the logical relationship between business metrics (such as "multiplying user activity by a weight of 0.5 and adding user retention rate multiplied by a weight of 0.5") and identify the data source type of the business metrics (such as the target service interface); input the semantic information extracted by the encoding layer into the parsing layer to perform parsing operations on business metrics. If the data source of the business metric is a target service interface, identify the target service interface (such as getUserActivity) and its input parameter information (such as {"date_range": "2024-01-01 to 2024-01-31"}) and output parameter information (such as {"activity": "value"}). At the same time, according to the information description of the business metric, extract the input parameter information and output parameter information. The parsing layer converts the information description of the business metric into specific interface call parameters; input the business metric information obtained from the parsing layer into the output layer to generate the access path information required to access the target service interface.For example, generate an access path: http: / / api.example.com / getUserActivity?date_range=2024-01-01%20to%202024-01-31. Combine it with the identifier of the target service interface (such as the interface name or URL) to generate the configuration information of the factor: {"factor_name": "User activity", "data_source_type": "service_interface", "interface_name": "getUserActivity", "access_path": "http: / / api.example.com / getUserActivity?date_range=2024-01-01%20to%202024-01-31", "input_params": {"date_range": "2024-01-01 to 2024-01-31"}, "output_param": "activity"}.

[0073] Through the content described in the above embodiments, the embodiments of the present application can generate the configuration information of the factor according to the information description of the business metrics, such as Figure 3a As shown, the calculation formula, factor, and the configuration information of the factor generated in the above embodiments are output, and the calculation logic configuration information is generated and stored in the database for the application system to call the calculation logic configuration information for analysis and processing based on the calculation logic between the business metrics. When it is necessary to update the calculation logic, only the configuration information needs to be modified, and the application system is unaware, without the need for synchronous upgrade or update. Thus, business personnel can flexibly adjust the calculation relationship between business metrics according to actual needs, improving the flexibility and efficiency of business decision-making, reducing the dependence on developers, and the cost of application system adjustment.

[0074] In an alternative embodiment, for any calculation logic, there can be multiple versions of the calculation logic configuration information, which may be generated due to changes in business requirements, updates of data sources, or optimization of the calculation logic. For example, in different time periods or business scenarios, the calculation logic may need to adjust the weights of factors, modify the calculation formula, or change the data sources of the factors. To adapt to such dynamic changes, the application system maintains a version management mechanism for each calculation logic to ensure that different versions of the calculation logic can be effectively recorded and distinguished. By supporting multi-version management, the application system can not only meet the diverse needs of business development but also quickly apply the previous version when needed, thereby improving the flexibility and stability of the application system. For example, when problems occur in the latest version of the calculation logic configuration information during actual application, it is possible to quickly switch to a previous stable version to ensure business continuity. This design provides strong support for the management of calculation logic in complex business environments.

[0075] In an alternative embodiment, after receiving the calculation logic description information expressed in natural language, the system generates the corresponding calculation logic configuration information based on this calculation logic description information and stores these calculation logic configuration information in a database. This process ensures the standardization and reusability of the calculation logic. After storage, the application system can perform analysis and calculations based on these defined business metrics through the corresponding calculation formulas to support specific business requirements. The following takes the use of the first calculation logic as an example for illustration.

[0076] In an alternative embodiment, receive the first calculation logic description information expressed in natural language, where the first calculation logic description information is any one of at least one calculation logic description information; according to the first calculation logic description information, query the calculation logic configuration information corresponding to each calculation logic description information to obtain the first calculation logic configuration information corresponding to the first calculation logic description information; call the parsing engine to parse the configuration information of the first factor in the first calculation logic configuration information, and obtain the value of the first factor according to the way information for reading the value of the first factor parsed from the configuration information of the first factor; call the calculation engine to generate the calculation result corresponding to the first calculation logic description information based on the first calculation formula and the first factor in the first calculation logic configuration information, in combination with the value of the first factor.

[0077] Among them, in the embodiments of the present application, the specific type of the parsing engine is not limited. The parsing engine is a module that specifically parses the configuration information of factors, and converts the configuration information of factors (such as database query statements or service interface call templates) into specific information on how to read values. For example, the parsing engine can be an SQL-based parser for parsing database query statements; it can also be an API parser for generating actual HTTP requests; it can also be a script parser for executing custom data extraction logic. For example, if the configuration information of a factor contains an SQL statement, the parsing engine will parse it into a database query operation that can be directly executed; if the configuration information of a factor contains an API call template, the parsing engine will generate an actual HTTP request. Call the calculation engine to generate the calculation result corresponding to the first calculation logic description information according to the first calculation formula and the first factor in the first calculation logic configuration information, in combination with the value of the first factor. The calculation engine is a module specifically used to execute calculation logic, and completes specific numerical calculations according to the calculation formula in the calculation logic configuration information, in combination with the actual values of the calculation factors, and outputs the final calculation result. For example, the calculation engine can be a script-based calculation engine for performing mathematical operations; it can also be a distributed computing framework for processing large-scale data; it can also be a real-time calculation engine for quickly responding to calculation requirements, and the embodiments of the present application do not limit this.

[0078] For example, the first calculation logic description information input by the user is: "Performance is calculated as GMV amount divided by GMV target, multiplied by a ratio of 0.5, plus resume data divided by resume target, multiplied by a ratio of 0.3." The system queries the database and finds that the first calculation logic configuration information corresponding to this first calculation logic and description information is "Calculation formula: Performance = 0.5×(A / B) + 0.3×(C / D); Corresponding relationship between business indicators and factors: A = GMV amount, B = GMV target, C = resume data, D = resume target; Configuration information of factors: A: {"factor_name": "GMV amount", "data_source_type": "database", "database_name": "sales_db", "query_statement": "SELECT SUM(amount) AS gmv_amount FROM gmv_table WHERE date BETWEEN '2024-01-01' AND '2024-01-31'"}, B: {"factor_name": "GMV target", "data_source_type": "database", "database_name": "sales_db", "query_statement": "SELECT SUM(target) AS gmv_target FROM gmv_table WHERE date BETWEEN '2024-01-01' AND '2024-01-31'"}, C: {"factor_name": "Resume data", "data_source_type": "database", "database_name": "hr_db", "query_statement": "SELECT COUNT(*) AS resume_count FROM resume_table WHERE date BETWEEN '2024-01-01' AND '2024-01-31'"}, {"factor_name": "Resume target", "data_source_type": "database", "database_name": "hr_db", "query_statement": "SELECT COUNT(*) AS resume_target FROM resume_table WHERE date BETWEEN '2024-01-01' AND '2024-01-31'"}.The database statements for the configuration information of the factors from A to D have the same meaning as the database statements in the foregoing embodiments, and are all used to extract the numerical values of the corresponding business metrics from the specified data sources. Since their functions and roles have been described in detail in the foregoing embodiments, they will not be elaborated here. The system calls the parsing engine to parse the configuration information of the factors and obtain the actual numerical values: for factor A (GMV amount), factor B (GMV target), factor C (resume data), and factor D (resume target), the parsing engine executes SQL queries and respectively obtains the query results as gmv_amount = 1000000, gmv_target = 1200000, resume_count = 500, and resume_target = 600. The system calls the calculation engine and calculates according to the calculation formula performance = 0.5×(GMV amount / GMV target) + 0.3×(resume data / resume target). Substituting the actual numerical values for calculation: performance = 0.5×(1000000 / 1200000) + 0.3×(500 / 600), and the calculation result is: performance = 0.5×0.8333 + 0.3×0.8333 ≈ 0.4167 + 0.25 ≈ 0.6667. Finally, the system outputs the calculation result: performance ≈ 0.6667. Through the above steps, the system realizes the automatic conversion from the calculation logic described in natural language to the specific calculation result, ensuring the flexibility and accuracy of the calculation logic.

[0079] In an alternative embodiment, receive the second calculation logic description information expressed in natural language and the target value corresponding to the second calculation logic description information, where the second calculation logic description information is any one of at least one calculation logic description information; according to the second calculation logic description information, query the calculation logic configuration information corresponding to each calculation logic description information to obtain the second calculation logic configuration information corresponding to the second calculation logic description information; call the parsing engine to parse the configuration information of the second factor in the second calculation logic configuration information, and obtain the numerical value of the second factor according to the information on the way to read the numerical value for the second factor parsed from the configuration information of the second factor; according to the second calculation formula and the second factor in the second calculation logic configuration information, combine the numerical value of the second factor, the target value, and the second prompt word, and call the second language model based on AI to analyze the influence relationship information between each second factor and the target value under the prompt of the second prompt word; optimize the calculation formula in the second calculation logic configuration information according to the influence relationship information.

[0080] Among them, the target value refers to the calculated result value expected by the user, which is used to measure the effectiveness and accuracy of the current calculation logic. The second language model based on artificial intelligence AI is an AI model specifically used to analyze and optimize the calculation logic. The second language model and the first language model can be the same model or different models, and this is not limited in the embodiments of the present application. The function of the second language model is to analyze the influence relationship information between each factor and the target value under the guidance of the second prompt word, and provide suggestions for optimizing the calculation formula. The second prompt word is natural language prompt information used to guide the second language model based on AI to correctly analyze the influence relationship between the factor and the target value. For example, "Please analyze the influence degree of each factor on the target value and give optimization suggestions." Among them, the influence relationship information refers to the influence degree of the factor on the target value, the sensitivity analysis result, and other related relationship data. For example, the influence degree of the GMV amount on the target performance is 0.6. The optimized calculation formula refers to adjusting the existing calculation formula according to the analyzed influence relationship information to improve the consistency between the calculation result and the target value. For example, optimizing the original formula "Performance = 0.5×(A / B) + 0.3×(C / D)" to "Performance = 0.6×(A / B) + 0.2×(C / D)".

[0081] Figure 4 It is a schematic flowchart of another information configuration method provided by an exemplary embodiment of the present application. As Figure 4 shown, the system obtains the calculation logic description information and the target value input by the user (such as a businessperson), obtains the second calculation logic configuration information from the database according to the calculation logic description information, and calls the parsing engine. The parsing engine parses the second factor according to the second calculation logic configuration information to obtain the value of the second factor. According to the second calculation formula and the second factor in the second calculation logic configuration information, combined with the value of the second factor, the target value, and the second prompt word, the second language model based on AI is called. Under the prompt of the second prompt word, the influence relationship information between each second factor and the target value is analyzed, including: inputting the second calculation formula, the second factor, the value of the second factor, the target value, and the second prompt word into the second language model. Under the prompt of the second prompt word, determine the calculation items included in the second calculation formula and the weight of each calculation item. Each calculation item describes the calculation relationship between some second factors. Among them, the second factor refers to all factors included in the second calculation formula. The calculation items included in the second calculation formula refer to each independent calculation part in the calculation formula, and each part describes the operation relationship between a group of factors.

[0082] For example, 0.5×(A / B) is a calculation item, which describes the division operation between factor A and factor B. The weight of each calculation item refers to the importance or contribution ratio of each calculation item in the total calculation formula. For example, 0.5 indicates that the weight of calculation item 1 in the total formula is 50%. The calculation relationship between the second factors in the description part of each calculation item refers to the specific operation relationship between the factors in each calculation item. For example, A / B describes the division relationship between factor A and factor B. Suppose the second calculation formula is performance = 0.5×(A / B) + 0.3×(C / D), where: calculation item 1: 0.5×(A / B), with a weight of 0.5, describes the division relationship between factor A and factor B; calculation item 2: 0.3×(C / D), with a weight of 0.3, describes the division relationship between factor C and factor D.

[0083] In an optional embodiment, when the weight of one calculation item is set to 1 and the weights of other calculation items are set to 0, the result of the calculation item with the weight set to 1 is calculated as the influence weight on the target value. Among them, 1 represents that the weight of the current calculation item is set to 1, indicating that the contribution of this calculation item in the total formula is considered separately; 0 represents that the weights of other calculation items are set to 0, indicating that these calculation items are ignored in the current analysis. The target influence weight is a quantitative index that measures the contribution degree or importance of each factor to the target value, reflects the influence size of a certain factor on the final target value in the calculation formula, and is expressed in numerical form (such as percentage or standardized score). By calculating the target influence weight, it can be clarified which factors have a greater influence on the target value, so as to provide a basis for optimizing the calculation logic or adjusting the business strategy, and is presented in the form of a ranking of the factors affecting the target value. For example, suppose the second calculation formula is performance = 0.5×(A / B) + 0.3×(C / D), the factor values are A = 1000000, B = 1200000, C = 500, D = 600, set the weight of calculation item 1 to 1, and the weights of other calculation items to 0: calculation item 1: 1×(A / B) = 1×(1000000 / 1200000) ≈ 0.8333, the influence weight is 0.8333, indicating the independent influence of calculation item 1 on the target value. Set the weight of calculation item 2 to 1, and the weights of other calculation items to 0: calculation item 2: 1×(C / D) = 1×(500 / 600) ≈ 0.8333, the influence weight is 0.8333, indicating the independent influence of calculation item 2 on the target value.

[0084] In an alternative embodiment, with the goal of achieving the target value using the second calculation formula, an exploratory solution is performed for each second factor to obtain the critical value corresponding to each second factor. Here, the exploratory solution is to dynamically adjust the value of the factor through an optimization algorithm (such as gradient descent, genetic algorithm, etc.) so that the result of the calculation formula is as close as possible to the target value. The critical value refers to the optimal value of the factor that makes the result of the calculation formula reach or approach the target value. For example, assume the second calculation formula is performance = 0.5×(A / B) + 0.3×(C / D), and the target value is 0.7: Perform an exploratory solution for factor A: Assume the critical value of factor A found through the optimization algorithm is 1,100,000. At this time, 0.5×(1,100,000 / 1,200,000)≈0.4583. Perform an exploratory solution for factor C: Assume the critical value of factor C found through the optimization algorithm is 550. At this time, 0.3×(550 / 600)≈0.275.

[0085] In an alternative embodiment, based on the influence weight and the critical value, determine the factor combination that has the greatest impact on the target value. The factor combination includes at least one second factor; the influence weight of each second factor on the target value, the critical value of each second factor, and the factor combination are used as influence relationship information. Here, the factor combination that has the greatest impact on the target value refers to the group of factors that contribute the most to the target value among all possible factor combinations, that is, the group of factors that can achieve the target value faster by adjusting which business indicators. The influence relationship information is the information that describes the influence weight, critical value, and their combination of each factor on the target value, and is used to guide business optimization. For example, assume the influence weight of factor A is 0.4167 and the critical value is 1,100,000; the influence weight of factor C is 0.25 and the critical value is 550: Determine the factor combination [A, C] as the combination that has the greatest impact on the target value. The influence relationship information is: factor A "weight: 0.4167, critical value: 1,100,000", factor C "weight: 0.25, critical value: 550".

[0086] Optionally, the second language model may include an input layer, an encoding layer, an analysis layer, and an output layer. The second calculation formula, the second factor, the value of the second factor, the target value, and the second prompt word are input into the input layer of the second language model, where word segmentation, part-of-speech tagging, and vectorization operations are performed, and embedding vectors are output. The input layer preprocesses this content to generate embedding vectors for representing the local meaning information of the input text. These embedding vectors provide a basis for subsequent semantic parsing and analysis. Then, the embedding vectors generated by the input layer are input into the encoding layer. The encoding layer is used to understand the logical relationship between the various factors in the calculation formula, and encodes the embedding vectors through a multi-layer neural network (Transformer architecture) to extract deeper semantic information and output each calculation term and its weight. The calculation terms and their weights extracted by the encoding layer are input into the analysis layer for calculating the influence weight and the critical value, and the influence weight and the critical value of each factor are output. Specifically, the influence weight is first calculated by setting the weight of each calculation term to 1 and the weights of other calculation terms to 0, and calculating the result of the calculation term with a weight of 1 as the influence weight on the brigade target value. Secondly, with the goal of the second calculation formula reaching the target value, an exploratory solution is performed for each factor to obtain the critical value of each factor. The influence weight and the critical value obtained by the analysis layer are input into the output layer for determining the factor combination, and influence relationship information is output. The output layer determines the factor combination with the greatest influence on the target value according to the influence weight and the critical value, and generates the final influence relationship information. Through the above process, the second language model can quantify the influence degree of each business indicator on the final calculation result, thereby providing a scientific basis for business decision-making. This analysis method can help business personnel clarify which indicators have the greatest influence on the target value, and then optimize business strategies to improve business efficiency and effectiveness.

[0087] In an optional embodiment, the exploratory solution of the critical value is a process of dynamically adjusting the values of the factors through a series of optimization algorithms to make the result of the calculation formula as close as possible to the target value. Specifically, the system can use methods such as the gradient descent method, the genetic algorithm, or the simulated annealing algorithm to achieve this goal. Taking the gradient descent method as an example, the system first sets an initial value for each factor, and then calculates the gradient of the objective function with respect to each factor, that is, the partial derivative. According to the gradient direction, the system gradually adjusts the factor values to minimize the difference between the objective function and the target value. This process gradually approaches the target value through iterative optimization until the preset number of iterations is reached or the difference between the objective function and the target value is less than a certain threshold. For example, assume that the second calculation formula is "Performance = 0.5×(A / B)+0.3×(C / D)", the target value is 0.7, and the initial factor values are A = 1000000, B = 1200000, C = 500, D = 600. The system calculates the gradient and updates the factor values, gradually optimizing factors A and C until the performance is close to the target value of 0.7.

[0088] Optionally, the system can also adopt a genetic algorithm, which is an optimization algorithm based on the principles of natural selection and genetics. The genetic algorithm finds the optimal solution by simulating the biological evolution process. The system first generates a set of random factor values as the initial population, and then calculates the fitness of each individual (factor combination), that is, the degree of proximity between the objective function and the target value. According to the fitness, the system selects excellent individuals for reproduction and generates new individuals through crossover and mutation operations. This process gradually approaches the target value through iterative evolution until the optimal solution is found or the preset number of iterations is reached. For example, the system generates 10 sets of random factor values as the initial population, calculates the fitness of each individual, and selects the individual with the highest fitness for reproduction. Through crossover and mutation operations, the system generates new individuals and repeats the above steps until the optimal solution is found.

[0089] For example, when the user inputs the calculation logic description information as "Performance is calculated by dividing the GMV amount by the GMV target, multiplying the result by a proportion of 0.5, adding the resume data divided by the resume target, multiplying the result by a proportion of 0.3, and then adding the number of leads completed divided by the total number of leads, multiplying the result by a proportion of 0.2" and sets the target value to 0.7, the system first preprocesses this information at the input layer, including word segmentation, part-of-speech tagging, and vectorization operations, to generate embedding vectors that can represent the local meaning information of the input text. The embedding vectors are passed to the encoding layer, which uses a multi-layer neural network to encode the embedding vectors, thereby extracting deeper semantic information, understanding the logical relationships between the various factors in the calculation formula, and also outputting each calculation term and its weight: calculation term 1 is 0.5×(GMV amount / GMV target), with a weight of 0.5; calculation term 2 is 0.3×(resume data / resume target), with a weight of 0.3; calculation term 3 is 0.2×(number of leads completed / total number of leads), with a weight of 0.2. The calculation terms and their weights extracted by the encoding layer are sent to the analysis layer, which calculates the influence weight of each factor. When the weight of calculation term 1 is 1, the result is 1×(GMV amount / GMV target). Assuming the initial values are GMV amount = 5000000 and GMV target = 6000000, the influence weight is 0.8333. Similarly, the influence weights of the resume data and the number of leads completed are 0.75 and 0.6667 respectively. After calculating the influence weights, the analysis layer further aims to reach the target value of 0.7 with the second calculation formula, and explores and solves each factor to find the critical value of each factor. This process can use a variety of optimization algorithms, such as the gradient descent method, genetic algorithm, or simulated annealing algorithm. Taking the gradient descent method as an example, the system first sets the initial value for each factor, then calculates the gradient (i.e., partial derivative) of the objective function with respect to each factor, and gradually adjusts the factor value according to the gradient direction to minimize the difference between the objective function and the target value. Through iterative optimization, the system finally finds the critical values that make the performance close to the target value of 0.7. Assuming that after optimization, the critical values are GMV amount = 5500000, resume data = 330, and number of leads completed = 220.

[0090] Finally, the influence weights and critical values of the factors obtained by the analysis layer are passed to the output layer. The influence weights and critical values of the factors in the output layer determine the factor combination that has the greatest impact on the target value and generate the final influence relationship information. For example, the output layer determines that ["GMV amount", "resume data", "number of leads completed"] is the factor combination that has the greatest impact on the target value and generates the following influence relationship information: GMV amount: weight is 0.8333, critical value is 5500000; resume data: weight is 0.75, critical value is 330; number of leads completed: weight is 0.6667, critical value is 220; factor combination: the factor combination that has the greatest impact on the target value includes GMV amount, resume data, and number of leads completed. Through the above process of generating the influence relationship, the impact degree of each business indicator on the final calculation result can be quantitatively analyzed, providing a scientific basis for business decision-making, helping business personnel clarify the impact of key indicators, optimize business strategies, improve business efficiency and effectiveness, and thus achieve precise decision-making and efficient operation.

[0091] In an optional embodiment, when optimizing the calculation formula in the second calculation logic configuration information according to the influence relationship information, the system can directly adjust the parameter configuration of the formula based on the influence relationship information (including the influence weight, critical value, and factor combination of the factor), for example, reallocating the weights of each calculation item or modifying the operation logic to better meet the business requirements. At the same time, the system outputs the influence relationship information to the business personnel, and the business personnel can analyze the rationality of the current formula based on this information and put forward optimization suggestions or instructions. After receiving the optimization instructions from the business personnel, the system further adjusts the formula and updates the configuration information. This method not only improves the accuracy and business flexibility of the calculation formula, but also reduces the trial-and-error cost through quantitative analysis, supports scientific decision-making, and ultimately helps to achieve business goals faster, improving the intelligent level and overall value of the system.

[0092] Figure 5 A flowchart of an information processing method provided by an exemplary embodiment of the present application. As Figure 5As shown, an information processing method is also provided. At least one piece of calculation logic configuration information corresponding to each piece of calculation logic description information is pre-configured. The method includes: receiving a first piece of calculation logic description information expressed in natural language, where the first piece of calculation logic description information is any one of at least one piece of calculation logic description information; querying the calculation logic configuration information corresponding to each piece of calculation logic description information according to the first piece of calculation logic description information to obtain the first piece of calculation logic configuration information corresponding to the first piece of calculation logic description information; calling a parsing engine to parse the configuration information of the first factor in the first piece of calculation logic configuration information, and obtaining the value of the first factor according to the way of reading the value for the first factor parsed from the configuration information of the first factor; calling a calculation engine to generate a calculation result corresponding to the first piece of calculation logic description information according to the first calculation formula and the first factor in the first piece of calculation logic configuration information, in combination with the value of the first factor. Here, the first factor refers to the factor corresponding to all business indicators in the first piece of calculation logic description information. Among them, the detailed implementation manners and beneficial effects of each step in the method of this embodiment have been described in detail in the foregoing embodiments, and will not be elaborated here.

[0093] Optionally, in the information processing method, the first piece of calculation logic configuration information corresponding to the first piece of calculation logic description information may be a valid version in the version of the calculation logic configuration information used by the current application system. In this case, the system will preferentially select the valid version of the calculation logic configuration information for subsequent processing to ensure that the calculation result is based on valid business rules and business indicator definitions. Query the calculation logic configuration information corresponding to each piece of calculation logic description information according to the calculation logic configuration information used by the current application system to obtain the first piece of calculation logic configuration information corresponding to the first piece of calculation logic description information. Among them, the detailed implementation manner and beneficial effect of the information processing method are the same as those described in the above embodiments, and will not be elaborated here.

[0094] In addition, in some of the processes described in the above embodiments and the accompanying drawings, a plurality of operations appear in a specific order. However, it should be clearly understood that these operations may not be executed in the order in which they appear in this document or may be executed in parallel. The operation numbers such as 11, 12, etc. are only used to distinguish different operations, and the numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in order or in parallel. It should be noted that the descriptions such as "first" and "second" in this document are used to distinguish different messages, devices, modules, etc., do not represent a sequence, and do not limit that "first" and "second" are of different types.

[0095] Figure 6 A schematic structural diagram of an electronic device provided for an exemplary embodiment of the present application. As Figure 6As shown, the device includes: a memory 64 and a processor 65.

[0096] The memory 64 is used to store computer programs and can be configured to store various other data to support operations on the electronic device. Examples of such data include instructions for any application or method for operating on the electronic device, business metrics, factors, calculation logic description information, configuration information of factors, calculation formulas, etc.

[0097] The memory 64 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disks or optical discs.

[0098] The processor 65, coupled to the memory 64, is configured to execute the computer programs in the memory 64 for: an information configuration method, including: receiving at least one calculation logic description information expressed in natural language, each calculation logic description information including a business metric and the calculation logic between business metrics; performing the following operations for each calculation logic description information: generating an information description of the business metric according to the data source information of the business metric in the calculation logic description information, the information description representing the correspondence between the business metric and the data source information; calling a first language model based on artificial intelligence (AI) according to the calculation logic description information, the information description of the business metric, and a first prompt word, and under the prompt of the first prompt word, representing the business metrics in the calculation logic description information as factors, generating configuration information of the factors according to the information description of the business metric, and converting the calculation logic between business metrics into a calculation formula dependent on the factors, where the factor is an atomic representation of the business metric, and the configuration information of the factor is used to describe the information on the way to read values for the factor; generating calculation logic configuration information corresponding to the calculation logic description information according to the calculation formula, the factors, and the configuration information of the factors, for an application system to call the calculation logic configuration information to perform analysis and processing based on the calculation logic between business metrics.

[0099] In an alternative embodiment, the processor 65 generates an information description of the service metrics based on the data source information of the service metrics in the computing logic description information, including: for any service metric, in response to a first input operation of a specific identifier, selecting the data source information type corresponding to the service metric from multiple data source information types; if the data source information type corresponding to the service metric is of the database type, determining a target database adapted to the service metric, and generating an information description of the service metric according to the target data table and target fields in the target database that are adapted to the service metric; if the data source information type corresponding to the service metric is of the service interface type, determining a target service interface adapted to the service metric, and generating an information description of the service metric according to the interface information of the target service interface.

[0100] In an alternative embodiment, the processor 65 determines a target database adapted to the service metric, including: in response to a second input operation of a specific identifier, displaying at least one candidate database; in response to a first selection operation, using the selected candidate database among the at least one candidate databases as the target database; generating an information description of the service metric according to the target data table and target fields in the target database that are adapted to the service metric, including: in response to a third input operation of a specific identifier, displaying the data tables in the target database and the fields in the data tables; in response to a second selection operation, selecting the target data table and target fields, and generating query condition description information adapted to the target data table and target fields; in response to a fourth input operation of a specific identifier, displaying the data tables in the target database and the fields in the data tables; in response to a second selection operation, selecting the target data table and target fields, and generating query result description information adapted to the target data table and target fields; generating an information description of the service metric according to the query condition description information and the query result description information.

[0101] In an alternative embodiment, the processor 65 determines a target service interface adapted to the service metric, including: in response to a fifth input operation of a specific identifier, displaying at least one candidate service interface; in response to a third selection operation, using the selected candidate service interface among the at least one candidate service interfaces as the target service interface; generating an information description of the service metric according to the interface information of the target service interface, including: in response to a sixth input operation of a specific identifier, obtaining the input parameter information of the target service interface input by the user; in response to a seventh input operation of a specific identifier, obtaining the output parameter information of the target service interface input by the user; generating an information description of the service metric according to the input parameter information and the output parameter information.

[0102] In an alternative embodiment, the processor 65 invokes a first language model based on artificial intelligence (AI) according to the calculation logic description information, the information description of the service metrics, and the first prompt word. Under the prompt of the first prompt word, the service metrics in the calculation logic description information are represented as factors, the configuration information of the factors is generated according to the information description of the service metrics, and the calculation logic between the service metrics is converted into a calculation formula depending on the factors, including: inputting the calculation logic description information, the information description of the service metrics, and the first prompt word into the first language model, and under the prompt of the first prompt word, parsing the calculation logic description information to obtain service metrics, mathematical operators, and the logical relationship between the service metrics and the mathematical operators; representing the service metrics as factors respectively, and combining the factors with data operators according to the logical relationship between the service metrics and the mathematical operators to obtain a calculation formula; and generating the configuration information of the factors according to the information description of the service metrics.

[0103] In an alternative embodiment, the processor 65 generates the configuration information of the factors according to the information description of the service metrics, including: when the data source information of the service metric is the target database, the target database includes a target data table and a target field adapted to the service metric, and the information description of the service metric includes query condition description information and query result description information adapted to the target data table and the target field, then generating a database statement required to query the target data table and the target field from the target database according to the query condition description information and the query result description information, and combining the identifier of the target database as the configuration information of the factor; when the data source information of the service metric is the target service interface, the information description of the service metric includes the input parameter information and the output parameter information of the target service interface, then generating access path information required to access the target service interface according to the input parameter information and the output parameter information as the configuration information of the factor.

[0104] In an alternative embodiment, the processor 65 receives first calculation logic description information expressed in natural language, and the first calculation logic description information is any one of at least one calculation logic description information; queries the calculation logic configuration information corresponding to each calculation logic description information according to the first calculation logic description information to obtain the first calculation logic configuration information corresponding to the first calculation logic description information; invokes a parsing engine to parse the configuration information of the first factor in the first calculation logic configuration information, and obtains the value of the first factor according to the way information of reading the value for the first factor parsed from the configuration information of the first factor; and invokes a calculation engine to generate a calculation result corresponding to the first calculation logic description information according to the first calculation formula and the first factor in the first calculation logic configuration information, in combination with the value of the first factor.

[0105] In an alternative embodiment, the processor 65 receives second computational logic description information expressed in natural language and a target value corresponding to the second computational logic description information, where the second computational logic description information is any one of at least one computational logic description information; according to the second computational logic description information, query the computational logic configuration information corresponding to each computational logic description information to obtain the second computational logic configuration information corresponding to the second computational logic description information; call the parsing engine to parse the configuration information of the second factor in the second computational logic configuration information, and obtain the value of the second factor according to the way of reading the value for the second factor parsed from the configuration information of the second factor; according to the second calculation formula and the second factor in the second computational logic configuration information, combine the value of the second factor, the target value, and the second prompt word, and call the second language model based on AI to analyze the influence relationship information between each second factor and the target value under the prompt of the second prompt word; optimize the calculation formula in the second computational logic configuration information according to the influence relationship information.

[0106] In an alternative embodiment, the processor 65, according to the second calculation formula and the second factor in the second computational logic configuration information, combines the value of the second factor, the target value, and the second prompt word, and calls the second language model based on AI to analyze the influence relationship information between each second factor and the target value under the prompt of the second prompt word, including: inputting the second calculation formula, the second factor, the value of the second factor, the target value, and the second prompt word into the second language model, and under the prompt of the second prompt word, determining the calculation items included in the second calculation formula and the weight of each calculation item, where each calculation item describes the calculation relationship between some of the second factors; when setting the weight of one calculation item to 1 and setting the weights of other calculation items to 0, calculate the result of the calculation item with the weight set to 1 as the influence weight on the target value; aiming at the second calculation formula reaching the target value, perform exploratory solution for each second factor to obtain the critical value corresponding to each second factor; according to the influence weight and the critical value, determine the factor combination with the greatest influence on the target value, where the factor combination includes at least one second factor; take the influence weight of each second factor on the target value, the critical value of each second factor, and the factor combination as the influence relationship information.

[0107] A processor 65, coupled to a memory 64, is configured to execute a computer program in the memory 64 for: An information processing method, which pre-configures calculation logic configuration information corresponding to at least one piece of calculation logic description information. The method includes: receiving first calculation logic description information expressed in natural language, where the first calculation logic description information is any one of the at least one piece of calculation logic description information; querying the calculation logic configuration information corresponding to each piece of calculation logic description information according to the first calculation logic description information to obtain first calculation logic configuration information corresponding to the first calculation logic description information; invoking a parsing engine to parse the configuration information of a first factor in the first calculation logic configuration information, and obtaining a value of the first factor according to the way of reading a value for the first factor parsed from the configuration information of the first factor; invoking a calculation engine to generate a calculation result corresponding to the first calculation logic description information according to a first calculation formula and the first factor in the first calculation logic configuration information, in combination with the value of the first factor.

[0108] Further, as Figure 6 shown, the electronic device further includes: other components such as a communication component 66, a display 67, a power supply component 68, an audio component 69, etc. Figure 6 Only some components are schematically shown, and it does not mean that the electronic device only includes Figure 6 the components shown. Additionally, Figure 6 the components within the dashed box in

[0109] are optional components, rather than mandatory components, and can be determined according to the product form of the working node specifically.

[0110] The embodiment of the present application further provides a computer-readable storage medium storing a computer program, and when the computer program is executed, it can implement the steps in the above method embodiment.

[0111] The above-mentioned memory can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0112] The above-mentioned communication component is configured to facilitate communication between the device where the communication component is located and other devices in a wired or wireless manner. The device where the communication component is located can access a wireless network based on a communication standard, such as WiFi, 2G, 3G, 4G / LTE, 5G and other mobile communication networks, or a combination thereof. In an exemplary embodiment, the communication component receives a broadcast signal or broadcast-related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component further includes a near field communication (NFC) module to facilitate short-range communication. For example, the NFC module can be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technologies.

[0113] The above-mentioned display includes a screen, and the screen can include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen can be implemented as a touch screen to receive input signals from a user. The touch panel includes one or more touch sensors to sense touches, swipes and gestures on the touch panel. The touch sensors can sense not only the boundaries of touch or swipe actions, but also detect the duration and pressure associated with the touch or swipe operations.

[0114] The above power supply component provides power for various components of the device where the power supply component is located. The power supply component may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power for the device where the power supply component is located.

[0115] The above audio component can be configured to output and / or input audio signals. For example, the audio component includes a microphone (MIC). When the device where the audio component is located is in an operating mode, such as a call mode, a recording mode, and a voice recognition mode, the microphone is configured to receive external audio signals. The received audio signals can be further stored in a memory or transmitted via a communication component. In some embodiments, the audio component further includes a speaker for outputting audio signals.

[0116] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-readable storage media (including but not limited to disk memory, compact disc read-only memory (CD-ROM), optical memory, etc.) that contain computer-usable program code.

[0117] The present application is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices generate means for implementing the functions specified in Figure 1 one or more flows and / or Figure 1 blocks or multiple blocks.

[0118] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including instruction means that implement the functions specified in Figure 1 one or more flows and / or Figure 1 blocks or multiple blocks.

[0119] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide for implementing the process Figure 1 one process or multiple processes and / or blocks Figure 1 steps of the functions specified in one block or multiple blocks.

[0120] In a typical configuration, a computing device includes one or more processors (Central Processing Unit, CPU), an input / output interface, a network interface, and a memory.

[0121] The memory may include non-permanent memory in the computer-readable medium, in the form of random access memory (Random Access Memory, RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash memory (flash RAM). Memory is an example of a computer-readable medium.

[0122] Computer-readable media includes permanent and non-permanent, removable and non-removable media that can store information by any method or technology. The information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change random access memory (Phase-change Random Access Memory, PRAM), static random access memory (SRAM), dynamic random access memory (Dynamic Random Access Memory, DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (Digital Video Disc, DVD) or other optical storage, magnetic cassette tapes, magnetic disk storage or other magnetic storage devices, or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include transitory computer-readable media, such as modulated data signals and carrier waves.

[0123] It should also be noted that the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, such that a process, method, commodity or device comprising a series of elements not only includes those elements but also includes other elements not expressly listed, or further includes elements inherent to such process, method, commodity or device. Without further limitation, an element defined by the statement "comprising one..." does not exclude the presence of additional identical elements in the process, method, commodity or device comprising the element.

[0124] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, various changes and modifications can be made to the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included within the scope of the claims of the present application.

Claims

1. An information configuration method, characterized in that: include: receiving at least one computing logic description information expressed in a natural language, each computing logic description information including a business indicator and a computing logic between the business indicators; Perform the following operations for each calculation logic description information: Generate an information description of the business indicator according to the data source information of the business indicator in the calculation logic description information, wherein the information description represents a corresponding relationship between the business indicator and the data source information; According to the calculation logic description information, the information description of the business indicator and the first prompt word, a first language model based on artificial intelligence AI is called, and under the prompt of the first prompt word, the business indicator in the calculation logic description information is represented as a factor, and the configuration information of the factor is generated according to the information description of the business indicator, and the calculation logic between the business indicators is converted into a calculation formula that depends on the factor, wherein the factor is an atomic representation of the business indicator, and the configuration information of the factor is used to describe the method information of reading the value of the factor; According to the calculation formula, the factors and the configuration information of the factors, calculation logic configuration information corresponding to the calculation logic description information is generated, so that the application system calls the calculation logic configuration information to perform analysis and processing based on the calculation logic between the business indicators.

2. The method according to claim 1, characterized in that Generating information description of the business indicator according to data source information of the business indicator in the calculation logic description information includes: For any business indicator, in response to a first input operation of a specific identifier, selecting a data source information type corresponding to the business indicator from a plurality of data source information types; If the data source information type corresponding to the business indicator is a database type, determining a target database adapted to the business indicator, and generating an information description of the business indicator according to a target data table and a target field adapted to the business indicator in the target database; If the data source information type corresponding to the business indicator is a service interface type, a target service interface adapted to the business indicator is determined, and an information description of the business indicator is generated according to the interface information of the target service interface.

3. The method according to claim 2, characterized in that Determining a target database that matches the business indicator includes: In response to a second input operation of a specific identifier, displaying at least one candidate database; in response to a first selection operation, using a candidate database selected from the at least one candidate database as the target database; Generate information description of the business indicator according to the target data table and target field in the target database that are adapted to the business indicator, including: In response to a third input operation of a specific identifier, displaying a data table and fields in the data table in the target database; in response to a second selection operation, selecting the target data table and target field, and generating query condition description information adapted to the target data table and target field; In response to a fourth input operation of a specific identifier, displaying a data table and fields in the data table in the target database; in response to a second selection operation, selecting the target data table and target field, and generating query result description information adapted to the target data table and target field; Generate an information description of the business indicator according to the query condition description information and the query result description information.

4. The method according to claim 2, characterized in that: Determining a target service interface that is compatible with the business indicator includes: In response to a fifth input operation of the specific identifier, display at least one candidate service interface; in response to a third selection operation, use a selected candidate service interface from the at least one candidate service interface as the target service interface; Generate information description of the business indicator according to the interface information of the target service interface, including: In response to a sixth input operation of the specific identifier, obtaining input parameter information of the target service interface input by the user; In response to a seventh input operation of the specific identifier, obtaining output parameter information of the target service interface input by the user; Generate information description of the business indicator according to the input parameter information and the output parameter information.

5. The method according to claim 1, characterized in that: According to the calculation logic description information, the information description of the business indicator and the first prompt word, a first language model based on artificial intelligence AI is called, and under the prompt of the first prompt word, the business indicator in the calculation logic description information is expressed as a factor, and the configuration information of the factor is generated according to the information description of the business indicator, and the calculation logic between the business indicators is converted into a calculation formula that depends on the factor, including: Inputting the calculation logic description information, the information description of the business indicator and the first prompt word into the first language model, and parsing the calculation logic description information under the prompt of the first prompt word to obtain the business indicator, the mathematical operator, and the logical relationship between the business indicator and the mathematical operator; Representing the business indicators as factors respectively, and combining the factors with the mathematical operators according to the logical relationship between the business indicators and the mathematical operators to obtain the calculation formula; According to the information description of the business indicator, the configuration information of the factor is generated.

6. The method according to claim 5, characterized in that According to the information description of the business indicator, the configuration information of the factor is generated, including: In the case where the data source information of the business indicator is a target database, the target database includes a target data table and a target field adapted to the business indicator, and the information description of the business indicator includes query condition description information and query result description information adapted to the target data table and the target field, then a database statement required to query the target data table and the target field from the target database is generated according to the query condition description information and the query result description information, and combined with an identifier of the target database, as the configuration information of the factor; When the data source information of the business indicator is the target service interface, the information description of the business indicator includes the input parameter information and output parameter information of the target service interface, and the access path information required to access the target service interface is generated according to the input parameter information and the output parameter information as the configuration information of the factor.

7. The method according to any one of claims 1 to 6, characterized in that: Also includes: receiving first computing logic description information expressed in a natural language, where the first computing logic description information is any one of the at least one computing logic description information; According to the first computing logic description information, query the computing logic configuration information corresponding to each computing logic description information to obtain the first computing logic configuration information corresponding to the first computing logic description information; Invoke a parsing engine to parse the configuration information of the first factor in the first computing logic configuration information, and obtain a value of the first factor according to the manner of reading a value for the first factor parsed from the configuration information of the first factor; The calculation engine is called to generate a calculation result corresponding to the first calculation logic description information according to the first calculation formula and the first factor in the first calculation logic configuration information and the value of the first factor.

8. The method according to any one of claims 1 to 6, characterized in that: Also includes: receiving second computing logic description information expressed in a natural language and a target value corresponding to the second computing logic description information, wherein the second computing logic description information is any one of the at least one computing logic description information; According to the second computing logic description information, query the computing logic configuration information corresponding to each computing logic description information to obtain the second computing logic configuration information corresponding to the second computing logic description information; Invoke a parsing engine to parse the configuration information of the second factor in the second computing logic configuration information, and obtain a value of the second factor according to the manner of reading the value of the second factor parsed from the configuration information of the second factor; According to the second calculation formula in the second calculation logic configuration information and the second factor, in combination with the value of the second factor, the target value and the second prompt word, calling the AI-based second language model, and under the prompt of the second prompt word, analyzing the influence relationship information between each second factor and the target value; The calculation formula in the second calculation logic configuration information is optimized according to the influence relationship information.

9. The method according to claim 8, characterized in that According to the second calculation formula and the second factor in the second calculation logic configuration information, in combination with the value of the second factor, the target value, and the second prompt word, the second language model based on AI is called, and under the prompt of the second prompt word, the influence relationship information between each second factor and the target value is analyzed, including: Inputting the second calculation formula, the second factor, the value of the second factor, the target value and the second prompt word into the second language model, and under the prompt of the second prompt word, determining the calculation items included in the second calculation formula and the weight of each calculation item, wherein each calculation item describes the calculation relationship between parts of the second factors; When the weight of one calculation item is reset to 1 and the weights of other calculation items are reset to 0, the result of calculating the calculation item whose weight is set to 1 is used as the influence weight on the target value; With the second calculation formula reaching the target value as the goal, each second factor is solved in an exploratory manner to obtain a critical value corresponding to each second factor; Determine, according to the influence weight and the critical value, a factor combination that has the greatest influence on the target value, wherein the factor combination includes at least one second factor; The influence weight of each second factor on the target value, the critical value of each second factor and the factor combination are used as the influence relationship information.

10. An information processing method, characterized in that: Pre-configuring computing logic configuration information corresponding to at least one computing logic description information, the method comprising: receiving first computing logic description information expressed in a natural language, where the first computing logic description information is any one of the at least one computing logic description information; According to the first computing logic description information, query the computing logic configuration information corresponding to each computing logic description information to obtain the first computing logic configuration information corresponding to the first computing logic description information; Invoke a parsing engine to parse the configuration information of the first factor in the first computing logic configuration information, and obtain a value of the first factor according to the manner of reading a value for the first factor parsed from the configuration information of the first factor; The calculation engine is called to generate a calculation result corresponding to the first calculation logic description information according to the first calculation formula and the first factor in the first calculation logic configuration information and the value of the first factor.

11. An electronic device, characterized in that: include: A processor and a memory, wherein the memory is used to store a computer program, and when the computer program is executed by the processor, the processor is enabled to implement the steps in the method according to any one of claims 1 to 10.

12. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the processor is enabled to implement the steps in the method according to any one of claims 1 to 10.

13. A computer program product, characterized in that The method comprises a computer program / instruction, which, when executed by a processor, enables the processor to implement the steps of the method according to any one of claims 1 to 10.

Images