Operational research result semantization method and device, electronic equipment and medium

Abstract

The invention provides a large-model-driven operational research result semantization method and device, electronic equipment and a computer readable storage medium. The method comprises the steps of obtaining optimized numerical data corresponding to an operation planning optimization problem and business background information of the operation planning optimization problem; based on a pre-trained large model, combining the optimized numerical data with the service background information, and identifying service elements corresponding to the optimized numerical data; and generating business interpretation corresponding to the operation planning optimization problem at least according to the corresponding relationship between the optimization numerical data and the business elements. The business interpretation of the numerical data corresponding to the operation planning optimization problem is generated by using the language generation and understanding capability of the large model, and a decision maker can understand the business significance of the numerical result corresponding to the operation planning optimization problem without depending on the auditing mathematical background or professional knowledge, so that a more intelligent business decision is made.

Description

Technical Field

[0001] This application relates to the field of large model technologies. Specifically, it relates to a method and device for semanticizing the results of operations research driven by a large model, an electronic device, and a computer-readable storage medium. Background Art

[0002] With the rapid development of big data and artificial intelligence technologies, enterprises increasingly rely on data-driven decision support tools in the decision-making process. Among them, operations research models, due to their excellent performance in fields such as resource optimization, supply chain management, and logistics scheduling, have become an important means for enterprises to improve operational efficiency.

[0003] However, the complexity and abstractness of operations research models often make it difficult to directly interpret and apply their numerical results to actual business scenarios. There are two main reasons: one is that the numerical results output by the model cannot intuitively reflect the business logic, making it difficult for decision-makers to understand the underlying business meaning; the other is the lack of an effective feedback mechanism to guide the iterative optimization and parameter adjustment of the model. Summary of the Invention

[0004] In view of the above problems, this application aims to provide a method and device for semanticizing the results of operations research driven by a large model, an electronic device, and a computer-readable storage medium, to solve the problem that the numerical results of operations research optimization problems in the prior art are difficult to directly interpret and apply to actual business scenarios.

[0005] In a first aspect, this application provides a method for semanticizing the results of operations research driven by a large model, the method including:

[0006] Obtain the optimized numerical data corresponding to the operations research optimization problem, and the business background information of the operations research optimization problem;

[0007] Based on a pre-trained large model, combine the optimized numerical data with the business background information to identify the business elements corresponding to the optimized numerical data;

[0008] Generate a business interpretation corresponding to the operations research optimization problem at least according to the correspondence between the optimized numerical data and the business elements.

[0009] In a possible implementation manner, the step of based on a pre-trained large model, combining the optimized numerical data with the business background information to identify the business elements corresponding to the optimized numerical data includes:

[0010] Based on a pre-trained large model, according to the business background information, identify the key elements corresponding to the operations research optimization problem, and the mathematical relationship between the business elements and the key elements;

[0011] Calculate the change trend of the key factor according to the optimized numerical data and the mathematical relationship between the business factor and the key factor;

[0012] Generate a business explanation corresponding to the operational research optimization problem at least according to the corresponding relationship between the optimized numerical data and the business factor, including:

[0013] Generate a business explanation corresponding to the operational research optimization problem according to the corresponding relationship between the optimized numerical data and the business factor, and the change trend of the key factor.

[0014] In a possible implementation manner, the method further includes:

[0015] Obtain an operational research model corresponding to the operational research optimization problem;

[0016] Based on a pre-established operational research knowledge base, according to the operational research model, identify and obtain the operational research logic for obtaining the optimized numerical data;

[0017] Combine the operational research logic with the business background information to generate a business logic, and generate a business explanation corresponding to the operational research optimization problem according to the business logic.

[0018] In a possible implementation manner, the generating a business explanation corresponding to the operational research optimization problem according to the business logic includes:

[0019] Based on a pre-trained language generation model, convert the business logic into a natural language text;

[0020] Generate a business explanation corresponding to the operational research optimization problem according to the natural language text.

[0021] In a possible implementation manner, the converting the business logic into a natural language text based on a pre-trained language generation model includes:

[0022] Use the business background information of the operational research optimization problem to adaptively adjust the language generation model;

[0023] Based on the adaptively adjusted language generation model, convert the business logic into a natural language text.

[0024] In a possible implementation manner, the identifying and obtaining the operational research logic for obtaining the optimized numerical data based on a pre-established operational research knowledge base according to the operational research model includes:

[0025] Determine the business complexity corresponding to the operational research optimization problem based on the business background information, and determine the logic depth according to the business complexity;

[0026] Based on a pre - established knowledge base, according to the operations research model, identify and obtain the operations research logic that matches the logical depth.

[0027] In a possible implementation manner, the operations research logic for identifying and obtaining the optimized numerical data based on the pre - established operations research knowledge base and according to the operations research model includes:

[0028] Receive the logical suggestions sent by the user according to the interaction interface, and based on the operations research knowledge related to the logical suggestions in the pre - established operations research knowledge base, according to the operations research model, identify and obtain the operations research logic of the optimized numerical data.

[0029] In a second aspect, the present application provides an operations research result semanticization device driven by a large model, including:

[0030] A data acquisition module, configured to acquire the optimized numerical data corresponding to the operations research optimization problem and the business background information of the operations research optimization problem;

[0031] A large model module, configured to combine the optimized numerical data with the business background information based on a pre - trained large model, and identify the business elements corresponding to the optimized numerical data;

[0032] A language generation module, configured to generate a business interpretation corresponding to the operations research optimization problem at least according to the correspondence between the optimized numerical data and the business elements.

[0033] In a third aspect, an embodiment of the present application further provides an electronic device, including: a processor and a memory communicatively connected to the processor;

[0034] The memory stores computer - executable instructions;

[0035] The processor executes the computer - executable instructions stored in the memory to implement the method in any possible implementation manner of the first aspect above.

[0036] In a fourth aspect, an embodiment of the present application further provides a computer - readable storage medium, in which computer - executable instructions are stored, and when the computer - executable instructions are executed by a processor, they are used to implement the method in any possible implementation manner of the first aspect above.

[0037] The operations research result semanticization method, device, electronic device, and computer - readable storage medium provided by the present application utilize the language generation and understanding capabilities of the large model to generate a business interpretation of the numerical data corresponding to the operations research optimization problem, helping decision - makers understand the business significance of the numerical results corresponding to the operations research optimization problem without relying on a profound mathematical background or professional knowledge, so as to make more informed business decisions. Brief Description of the Drawings

[0038] The drawings herein are incorporated into and form a part of this specification, showing embodiments consistent with the present application, and are used together with the specification to explain the principles of the present application.

[0039] Figure 1 It is a schematic flowchart of a method for semanticizing the results of operations research driven by a large model provided by an embodiment of the present application;

[0040] Figure 2 It is a schematic flowchart of some steps of another method for semanticizing the results of operations research driven by a large model provided by an embodiment of the present application;

[0041] Figure 3 It is a schematic flowchart of some steps of another method for semanticizing the results of operations research driven by a large model provided by an embodiment of the present application;

[0042] Figure 4 It is a schematic flowchart of some steps of another method for semanticizing the results of operations research driven by a large model provided by an embodiment of the present application;

[0043] Figure 5 It is a schematic flowchart of some steps of another method for semanticizing the results of operations research driven by a large model provided by an embodiment of the present application;

[0044] Figure 6 It is a schematic structural diagram of a device for semanticizing the results of operations research driven by a large model provided by an embodiment of the present application;

[0045] Figure 7 It is a hardware structure diagram of an electronic device provided by an embodiment of the present application.

[0046] Through the above drawings, specific embodiments of the present application have been shown, and there will be more detailed descriptions hereinafter. These drawings and the textual descriptions are not intended to limit the scope of the concept of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments.

[0047] Description of Reference Numerals:

[0048] 601 - Data Acquisition Module; 602 - Large Model Module; 603 - Language Generation Module

[0049] 701 - Processor; 702 - Memory; 703 - Communication Interface; 704 - Communication Bus. Detailed Description of the Embodiments

[0050] Exemplary embodiments will be described in detail herein, and examples thereof are shown in the accompanying drawings. When the following description refers to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. On the contrary, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0051] In the embodiments of the present application, terms such as "first" and "second" are used to distinguish identical or similar items with basically the same functions and roles. Those skilled in the art can understand that terms such as "first" and "second" do not limit the quantity and execution order, and "first", "second", etc. do not necessarily mean different.

[0052] It should be noted that in the embodiments of the present application, words such as "exemplary" or "for example" are used to represent examples, illustrations or explanations. Any embodiment or design solution described as "exemplary" or "for example" in the present application should not be construed as being more preferred or having more advantages than other embodiments or design solutions. Rather, the use of words such as "exemplary" or "for example" is intended to present relevant concepts in a specific manner. In the embodiments of the present application, "at least one" means one or more, and "a plurality" means two or more.

[0053] It should be noted that in the embodiments of the present application, "when..." can be at the instant when a certain situation occurs, or within a period of time after a certain situation occurs. The embodiments of the present application do not make specific limitations on this.

[0054] In some related technologies, NLP (Natural Language Processing) technology can be used to interpret and present the data analysis results of the numerical values output by the operations research optimization model.

[0055] Specifically, by using NLP technology to automatically generate a data analysis report, the complex data analysis results are converted into a description in natural language, which is convenient for users to understand. For example, the Quill platform of Narrative Science company can automatically convert structured data into natural language text to help users quickly interpret data results.

[0056] In some related technologies, NLP technology can also be combined with the visualization function, and visualization technology is used to visualize the data analysis results. However, this kind of visualization focuses on graphical display rather than in-depth business logic interpretation.

[0057] That is to say, related technologies are limited to the surface description of data results and lack the ability to understand and explain deep business logic, which limits the application effect of related technologies in complex business environments.

[0058] To solve the above technical problems, embodiments of the present application provide a method and apparatus for semanticizing the results of operations research driven by a large model, an electronic device, and a computer-readable storage medium. By using the language generation and understanding capabilities of the large model, business interpretations of the numerical data corresponding to the operations research optimization problem are generated, helping decision-makers understand the business significance of the numerical results corresponding to the operations research optimization problem without relying on an auditing mathematical background or professional knowledge, so as to make more informed business decisions.

[0059] Figure 1 As shown in the flowchart of a method for semanticizing the results of operations research driven by a large model provided by an embodiment of the present application, Figure 1 The specific implementation process of this method may include the following steps:

[0060] S110: Obtain the optimization numerical data corresponding to the operations research optimization problem, as well as the business background information of the operations research optimization problem.

[0061] Operations research optimization can consist of branches such as "linear programming" and "integer programming". Operations research optimization problems can be problems such as "linear programming" and "integer programming".

[0062] Obtaining the optimization numerical data corresponding to the operations research optimization problem can be to establish an operations research model according to the operations research optimization problem, input the operations research model corresponding to the operations research optimization problem into an operations research model solver, and obtain the numerical results output by the operations research model solver, that is, the optimization numerical data corresponding to the operations research optimization problem.

[0063] Specifically, the numerical results output by the operations research model solver can be obtained through a data interface. Through the data interface, different types of operations research models and operations research model solvers can be flexibly integrated.

[0064] Obtaining the business background information of the operations research optimization problem can be to obtain the business scenario corresponding to the operations research optimization problem, and use the business elements related to the operations research optimization problem and the relationships between the business elements in the business scenario as the business background information of the operations research optimization problem.

[0065] Among them, the business scenario of the operations research optimization problem can be obtained through an interaction device.

[0066] Among them, the interaction device can include input devices such as a mouse and a keyboard. The interaction device can also include a display device such as a monitor. An interaction interface is displayed on the display device, and the business scenario of the operations research optimization problem is obtained through interaction devices such as a mouse and a keyboard.

[0067] Obtaining the business elements related to the operational research optimization problem in the business scenario and the relationships between the business elements can be obtained based on intelligent methods such as knowledge graphs, or based on the text generation ability of large models. The text of the business scenario and the operational research optimization problem is input into the large model to obtain the relevant information.

[0068] S120: Based on a pre-trained large model, combine the optimization numerical data with the business background information to identify the business elements corresponding to the optimization numerical data.

[0069] Input the optimization numerical data and the business background information into the pre-trained large model, and utilize the language generation and understanding ability of the large model to combine the optimization numerical data with the business background information to determine the business elements corresponding to the optimization data.

[0070] Among them, the large model used in the semanticization method of operational research results driven by the large model in the embodiments of the present application can be a GPT (Generative Pretrained Transformer) model, or other large language models that can achieve corresponding functions.

[0071] In recent years, with the rapid development of NLP technology and artificial intelligence technology, natural language engines based on LLM (Large Language Model), such as GPT (Generative Pretrained Transformer) models, have performed excellently in tasks such as text generation and question answering. Generating the business interpretation corresponding to the operational research optimization problem is essentially a text generation problem. Therefore, the business interpretation corresponding to the operational research optimization problem can be generated based on LLM.

[0072] To improve the recognition effect of the large model, the operational research model corresponding to the operational research optimization problem, the optimization numerical data, and the business background information can also be input into the large model. Specifically, the decision variables, objective functions, and constraint conditions of the operational research model can be input into the large model.

[0073] S130: Generate the business interpretation corresponding to the operational research optimization problem at least according to the correspondence between the optimization numerical data and the business elements.

[0074] Generate a data description of the optimization numerical data according to the correspondence between the optimization numerical data and the business elements, convert the data description into natural language text, use the generated natural language text as the business interpretation corresponding to the operational research optimization problem, and display it on the interaction interface.

[0075] Next, a semanticization method of operational research results driven by a large model provided by the embodiments of the present application will be specifically introduced.

[0076] In some embodiments, the business interpretation corresponding to the operational research optimization problem further includes the key elements in the operational research optimization problem and the change trends of the key elements.

[0077] Figure 2 A flowchart showing an implementation method for obtaining key elements and the changing trends of key elements in operational research optimization problems based on large model recognition is shown. As Figure 2 shown, the following steps may be included:

[0078] S121: Based on a pre-trained large model and according to business background information, identify the key elements corresponding to the operational research optimization problem and the mathematical relationships between business elements and key elements.

[0079] Utilize the language generation and understanding capabilities of the large model. Based on the business elements related to the operational research optimization problem and the relationships between business elements, determine the key elements corresponding to the operational research optimization problem from the business elements related to the operational research optimization problem, and determine the mathematical relationships between the business elements corresponding to the optimization numerical data and the key elements according to the relationships between the business elements related to the operational research optimization problem. That is, how to convert business elements into key elements through mathematical calculations.

[0080] To improve the recognition effect of the large model, the operational research model corresponding to the operational research optimization problem, the optimization numerical data, and the business background information can also be input into the large model to provide data for the large model analysis. Specifically, the decision variables, objective function, and constraint conditions of the operational research model can be input into the large model.

[0081] S122: Calculate the changing trends of key elements according to the optimization numerical data and the mathematical relationships between business elements and key elements.

[0082] According to the mathematical relationships between business elements and key elements, convert the optimization numerical data into numerical data corresponding to key elements, and calculate the changing trends of key elements based on the changing trends of the numerical data corresponding to key elements and the converted numerical data in the operational research optimization problem.

[0083] In some embodiments, after obtaining the mathematical relationships between business elements and key elements and the changing trends of key elements, the changing trends of key elements can be converted into data descriptions of key elements, and the mathematical relationships between business elements and key elements can be converted into mathematical logics for determining the changing trends of key elements. The data descriptions and mathematical logics are converted into natural language texts, and the generated natural language texts are also used as part of the business explanations corresponding to the operational research optimization problem and are displayed on the interaction interface.

[0084] In some embodiments, the business explanations corresponding to the operational research optimization problem may also include generating the operational research logics corresponding to the optimization numerical data.

[0085] Figure 3 A flowchart showing an implementation method for generating the operational research logics corresponding to the optimization numerical data is shown. As Figure 3As shown in the figure, the operational research logic for generating optimized numerical data may include the following steps:

[0086] S211: Obtain the operational research model corresponding to the operational research optimization problem.

[0087] Obtaining the operational research model corresponding to the operational research optimization problem may be to obtain relevant parameters such as the decision variables, objective function, and constraint conditions of the operational research model.

[0088] Among them, the decision variable is the unknown quantity to be determined in the problem, which is used to indicate the solutions, measures, etc. in the operational research optimization problem and can be determined and controlled by the decision maker.

[0089] The objective function is a function of the decision variables, and the goal of optimization is to find the maximum or minimum value of this function.

[0090] The constraint conditions are the constraints and restrictions on the values of the decision variables, usually expressed by equations or inequalities containing the decision variables.

[0091] S212: Based on the pre-established operational research knowledge base, identify the operational research logic for obtaining optimized numerical data according to the operational research model.

[0092] According to the decision variables, objective function, and constraint conditions of the operational research model, use the operational research knowledge in the operational research knowledge base to obtain the operational research logic for obtaining optimized numerical data according to the decision variables, objective function, and constraint conditions.

[0093] In some embodiments, the operational research knowledge base can be put into a large model, and by utilizing the language understanding ability of the large model and inputting the prompt text, the operational research logic for obtaining optimized numerical data can be identified.

[0094] In the large language model, the prompt text is the input text used to guide the model to generate specific types of text or perform specific tasks. Among them, the prompt text can be generated in various ways. It can be directly the text obtained by splicing the natural language problem text and the candidate key information as the prompt text, or predefined templates or rules can be used to generate the prompt text. Among them, the prompt template describes the structure and content of the required prompt text.

[0095] In the embodiments of this application, the text generation function of the large model can be utilized to realize the identification of the operational research logic for obtaining optimized numerical data.

[0096] Specifically, fill in the decision variables, objective function, and constraint conditions of the operational research model into the preset prompt template to generate the prompt text; the prompt template includes the operational research logic for prompting the large language model to determine the numerical values of the decision variables that make the objective function reach the extreme value under the constraint conditions based on the decision variables, objective function, and constraint conditions of the operational research model filled in the prompt template.

[0097] For example, the prompt template can be: You are a professional operations research personnel. Based on {}, establish an operations research model, solve the operations research model, and output the operations research logic for solving the operations research model. The generated prompt text can be: You are a professional operations research personnel. Based on {the decision variables, objective function, and constraints of the operations research model}, establish an operations research model, solve the operations research model, and output the operations research logic for solving the operations research model.

[0098] Among them, the operations research logic can include determining the causal relationships of the optimized numerical data, the dynamic changes of the business elements related to the operations research optimization problem, and the possible impacts of these dynamic changes on other business elements.

[0099] S213: Combine the operations research logic with the business background information to generate business logic, and generate a business interpretation corresponding to the operations research optimization problem according to the business logic.

[0100] Based on the business background information, determine the corresponding relationships between the numerical data, variables, quantities, etc. in the operations research logic and the business elements, and substitute the business elements into the operations research logic to generate business logic.

[0101] Convert the business logic into natural language text, use the generated natural language text as the business interpretation corresponding to the operations research optimization problem, and display it on the interaction interface.

[0102] Among them, converting the business logic into natural language text can be implemented based on a model. Specifically, based on a pre-trained language generation model, convert the business logic into natural language text; generate a business interpretation corresponding to the operations research optimization problem according to the natural language text.

[0103] The language generation model is a model used to convert abstract languages and professional terms into languages that the public can understand. Based on the language generation model, professional terms in the business logic can be converted into explanations that the public can understand, so that decision-makers can understand the business interpretation without relying on an in-depth mathematical background or professional knowledge.

[0104] In some embodiments, the language generation model can be adaptively adjusted according to the characteristics of different industries.

[0105] Specifically, use the business background information of the operations research optimization problem to adaptively adjust the language generation model; based on the adaptively adjusted language generation model, convert the business logic into natural language text.

[0106] The industry to which the business scenario corresponding to the operations research optimization problem belongs can be determined based on the business background information of the operations research optimization problem, and the language generation model can be adaptively adjusted according to the industry characteristics of the industry to which it belongs, so that the natural language text generated by the language generation model conforms to the language habits of practitioners in this industry.

[0107] For example, industry terms of this industry and their corresponding explanations can be added to the language generation model, so that the language generation model can convert business elements in the business logic into corresponding industry terms.

[0108] In some embodiments, for different business scenarios, the details of the process of obtaining the operations research logic can be adjusted.

[0109] Figure 4 A flowchart showing an implementation manner of the adjustment is shown, such as Figure 4 shown, and may include the following steps:

[0110] S2121: Determine the business complexity corresponding to the operations research optimization problem based on the business background information, and determine the logical depth according to the business complexity.

[0111] Determine the business complexity according to the business scenario corresponding to the operations research optimization problem in the business background information, and determine the enterprise scale according to the enterprise where the operations research optimization problem is actually applied; determine the logical depth according to the business complexity and the enterprise scale, that is, the more complex the business complexity, the larger the enterprise scale, the deeper the logical depth; the simpler the business complexity, the smaller the enterprise scale, the shallower the logical depth.

[0112] S2122: Based on the pre-established knowledge base, identify and obtain the operations research logic matching the logical depth according to the operations research model.

[0113] The operations research knowledge base can be put into the large model, and using the language understanding ability of the large model, by inputting the prompt text, the operations research logic for obtaining the optimized numerical data can be identified.

[0114] Specifically, the decision variables, objective function, and constraint conditions of the operations research model and the logical depth are filled into a preset prompt template to generate the prompt text; the prompt template includes the operations research logic that prompts the large language model to determine the numerical values of the decision variables that make the objective function reach the extreme value under the constraint conditions based on the decision variables, objective function, and constraint conditions of the operations research model filled into the prompt template, and the thinking depth of this operations research logic is the logical depth.

[0115] For example, the prompt template can be: You are a professional operations research personnel. Based on {}, establish an operations research model, and based on {} in-depth thinking, solve the operations research model, and output the operations research logic for solving the operations research model. The generated prompt text can be: You are a professional operations research personnel. Based on {decision variables, objective function, and constraints of the operations research model}, establish an operations research model, and based on {logical depth} in-depth thinking, solve the operations research model, and output the operations research logic for solving the operations research model.

[0116] Among them, the deeper the logical depth, the higher the level of dynamic changes in business elements related to the operations research optimization problem is calculated, that is, not only the dynamic changes in business elements directly related to decision variables are calculated, but also the dynamic changes in business elements with a lower correlation with decision variables are calculated, and the deeper the thinking about the possible impacts of these dynamic changes on other business elements is, that is, the impacts of minor dynamic changes on business elements are not ignored.

[0117] On the contrary, the shallower the logical depth, the lower the level of dynamic changes in business elements related to the operations research optimization problem is calculated, that is, only the dynamic changes in business elements directly related to decision variables are calculated, and the shallower the thinking about the possible impacts of these dynamic changes on other business elements is, that is, the impacts of minor dynamic changes on business elements are ignored.

[0118] Figure 5 A flowchart showing another implementation method of the adjustment is shown. As Figure 5 shown, it may include the following steps:

[0119] S2123: Receive the logical suggestion sent by the user according to the interaction interface, and based on the operations research knowledge related to the logical suggestion in the pre-established operations research knowledge base, identify and obtain the operations research logic for optimizing numerical data according to the operations research model.

[0120] Receiving the logical suggestion sent by the user according to the interaction interface may be receiving the logical suggestion sent by the user according to the interaction device.

[0121] Among them, the interaction device may include input devices such as a mouse and a keyboard, and the interaction device may also include a display device such as a monitor. The interaction interface is displayed on the display device, and the user submits a logical suggestion through interaction devices such as a mouse and a keyboard.

[0122] The logical suggestion may be the operations research logic that is focused on. Obtain the operations research knowledge related to the operations research logic that is focused on from the operations research knowledge base, and identify and obtain the operations research logic for optimizing numerical data according to the operations research model.

[0123] Specifically, the decision variables, objective function, constraint conditions, and logical suggestions of the operations research model are filled into a preset prompt template to generate prompt text. The prompt template includes the operations research logic related to the logical suggestions for the large language model to determine the values of the decision variables that maximize or minimize the objective function under the constraint conditions based on the decision variables, objective function, and constraint conditions of the operations research model filled into the prompt template.

[0124] For example, the prompt template can be: You are a professional operations research personnel. Based on {}, establish an operations research model, solve the operations research model, and output the operations research logic related to {} for solving the operations research model. The generated prompt text can be: You are a professional operations research personnel. Based on {decision variables, objective function, and constraint conditions of the operations research model}, establish an operations research model, solve the operations research model, and output the operations research logic related to {logical suggestions} for solving the operations research model.

[0125] Since the business interpretation corresponding to the operations research optimization problem is generated according to the operations research logic, obtaining the logical suggestions of the user through the interaction interface is to obtain the suggestions of the user for the displayed business interpretation, thus enabling the user to adjust the display and generation process of the business interpretation according to their own needs.

[0126] As described above, the large model-driven operations research result semanticization method provided by the embodiments of the present application can be customized according to the specific needs of different enterprises. This includes adjusting the language output and optimization suggestions according to industry characteristics, enterprise scale, and specific decision-making needs, providing more accurate and efficient decision-making support. At the same time, this customized support reduces the additional development costs and time of enterprises in the process of tool implementation.

[0127] In some embodiments, the computational efficiency of the system can be improved by executing the large model-driven operations research result semanticization method provided by the embodiments of the present application in parallel.

[0128] The computing resources can be dynamically allocated to ensure the system response speed when processing massive data.

[0129] It can support real-time data input and update to ensure the timeliness of decision-making support.

[0130] Figure 6 It is a schematic structural diagram of a large model-driven operations research result semanticization device provided by the embodiments of the present application. The device can be in the form of software and / or hardware. Refer to Figure 6 As shown, a large model-driven operations research result semanticization device includes: a data acquisition module 601, a large model module 602, and a language generation module 603.

[0131] A data acquisition module 601, configured to acquire optimization numerical data corresponding to an operations research optimization problem and business background information of the operations research optimization problem;

[0132] A large model module 602, configured to combine the optimization numerical data with the business background information based on a pre-trained large model to identify business elements corresponding to the optimization numerical data;

[0133] A language generation module 603, configured to generate a business interpretation corresponding to the operations research optimization problem at least according to the correspondence between the optimization numerical data and the business elements.

[0134] In some embodiments, the large model module 602 is specifically configured to:

[0135] Based on a pre-trained large model, identify key elements corresponding to the operations research optimization problem and the mathematical relationship between the business elements and the key elements according to the business background information;

[0136] Calculate the change trend of the key elements according to the optimization numerical data and the mathematical relationship between the business elements and the key elements;

[0137] The language generation module 603 is specifically configured to:

[0138] Generate a business interpretation corresponding to the operations research optimization problem according to the correspondence between the optimization numerical data and the business elements and the change trend of the key elements.

[0139] In some embodiments, the large model-driven semanticization device for operations research results further includes a logic generation module, which is specifically configured to:

[0140] Acquire an operations research model corresponding to the operations research optimization problem;

[0141] Based on a pre-established operations research knowledge base, identify the operations research logic for acquiring the optimization numerical data according to the operations research model;

[0142] Combine the operations research logic with the business background information to generate a business logic, and generate a business interpretation corresponding to the operations research optimization problem according to the business logic.

[0143] In some embodiments, the logic generation module is specifically configured to:

[0144] Based on a pre-trained language generation model, convert the business logic into natural language text;

[0145] Generate a business interpretation corresponding to the operations research optimization problem according to the natural language text.

[0146] In some embodiments, the logic generation module is specifically configured to:

[0147] Adapting the language generation model using the business background information of the operations research optimization problem;

[0148] Based on the adaptively adjusted language generation model, transforming the business logic into natural language text.

[0149] In some embodiments, the logic generation module is specifically configured to:

[0150] Determine the business complexity corresponding to the operations research optimization problem based on the business background information, and determine the logic depth according to the business complexity;

[0151] Based on the pre-established knowledge base, identify and obtain the operations research logic matching the logic depth according to the operations research model.

[0152] In some embodiments, the logic generation module is specifically configured to:

[0153] Receive the logic suggestion sent by the user according to the interaction interface, and based on the operations research knowledge related to the logic suggestion in the pre-established operations research knowledge base, identify and obtain the operations research logic for optimizing numerical data according to the operations research model.

[0154] Figure 7 The hardware structure diagram of an electronic device provided by an embodiment of the present application. This embodiment provides an electronic device, including: at least one processor 701, and a memory 702 communicatively connected to the at least one processor 701; the memory 702 stores computer-executable instructions; the processor 701 executes the computer-executable instructions stored in the memory 702 to implement the method for semanticizing the operations research results driven by the large model described in any of the foregoing embodiments.

[0155] Figure 7 The electronic device shown further includes a communication interface 703 and a communication bus 704. Among them, the processor 701, the memory 702, and the communication interface 703 are connected to each other through the communication bus 704. The communication bus 704 can be divided into an address bus, a data bus, a control bus, etc. For the sake of representation, Figure 7 only a thick line is used to represent the communication bus 704 in the figure, but it does not mean that there is only one communication bus 704 or one type of communication bus 704. The processor 701 can also be called a controller, and there is no restriction on the name.

[0156] In the embodiment of the present application, the memory 702 stores instructions executable by at least one processor 701. By executing the instructions stored in the memory 702, the at least one processor 701 can execute the method for semanticizing the operations research results driven by the large model described above. The processor 701 can implement Figure 7 the functions of each module in the device shown.

[0157] Among them, the processor 701 is the control center of the device, which can connect various parts of the entire control device through various interfaces and lines. By running or executing instructions stored in the memory 702 and invoking data stored in the memory 702, various functions of the device and data processing are carried out, so as to monitor the device as a whole.

[0158] In a possible design, the processor 701 may include one or more processing units. The processor 701 may integrate an application processor and a modem processor. Among them, the application processor mainly processes the operating system, user interface, application programs, etc., and the modem processor mainly processes wireless communication. It can be understood that the above-mentioned modem processor may not be integrated into the processor 701 either. In some embodiments, the processor 701 and the memory 702 may be implemented on the same chip or separately on independent chips.

[0159] The processor 701 may be a general-purpose processor, such as a central processing unit (CPU), a digital signal processor, an application-specific integrated circuit, a field programmable gate array, or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, which can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of the present application. The general-purpose processor may be a microprocessor or any conventional processor, etc. The steps of the semanticization method of the operation research results driven by the large model disclosed in combination with the embodiments of the present application can be directly embodied as being completed by the execution of the hardware processor, or completed by the combination of the hardware and software modules in the processor.

[0160] The memory 702, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs, non-volatile computer-executable programs, and modules. The memory 702 may include at least one type of storage medium. For example, it may include flash memory, hard disk, multimedia card, card-type memory, random access memory (RAM), static random access memory (SRAM), programmable read-only memory (PROM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), magnetic memory, magnetic disk, optical disc, etc. The memory 702 is any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory 702 in the embodiments of the present application may also be a circuit or any other device capable of implementing a storage function, for storing program instructions and / or data.

[0161] By programming the processor 701, the code corresponding to the large model-driven semanticization method of operations research results introduced in the foregoing embodiments can be solidified into the chip, so that the chip can execute Figures 1 to 6 the steps of the large model-driven semanticization method of operations research results shown in the embodiments. How to program the processor 701 is a well-known technology to those skilled in the art and will not be elaborated here.

[0162] The embodiments of the present application also provide a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement the large model-driven semanticization method of operations research results described in any of the foregoing embodiments. Therefore, it will not be elaborated here. In addition, the description of the beneficial effects of the same method will not be elaborated either. For the technical details not disclosed in the embodiments of the computer storage medium involved in the present invention, please refer to the description of the method embodiments of the present invention.

[0163] In some possible implementation manners, various aspects of the large model-driven semanticization method of operations research results provided in the present application can also be implemented in the form of a program product, which includes program code. When the program product runs on a device, the program code is used to cause the control device to execute the steps in the large model-driven semanticization method of operations research results according to various exemplary implementation manners of the present application described above in this specification.

[0164] 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-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0165] 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 flowchart and / or block diagram, as well as the combination of flows and / or blocks in the flowchart and / or block diagram, can be realized 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 a device for realizing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0166] 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 an instruction device that realizes the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0167] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, so that the instructions executed on the computer or other programmable device provide steps for realizing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0168] Obviously, those skilled in the art can make various modifications and variations to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalent technologies, the present application is also intended to include these modifications and variations.

Claims

1. A semantic method for operations research results driven by a large model, characterized in that: The method comprises: Obtaining optimized numerical data corresponding to the operations research optimization problem and business background information of the operations research optimization problem; Based on the pre-trained large model, the optimized numerical data is combined with the business background information to identify the business elements corresponding to the optimized numerical data; A business explanation corresponding to the operations optimization problem is generated based at least on the corresponding relationship between the optimized numerical data and the business elements.

2. The method according to claim 1, characterized in that The method of combining the optimized numerical data with the business background information based on the pre-trained large model to identify the business elements corresponding to the optimized numerical data includes: Based on the pre-trained large model, according to the business background information, identify the key elements corresponding to the operations optimization problem, and the mathematical relationship between the business elements and the key elements; Calculating the change trend of the key element according to the optimized numerical data and the mathematical relationship between the business element and the key element; The generating of the business explanation corresponding to the operations optimization problem at least according to the correspondence between the optimized numerical data and the business elements includes: A business explanation corresponding to the operations optimization problem is generated according to the corresponding relationship between the optimized numerical data and the business elements, and the change trend of the key elements.

3. The method according to claim 1, characterized in that: The method further comprises: Obtaining an operations research model corresponding to the operations optimization problem; Based on a pre-established operations research knowledge base and according to the operations research model, identifying the operations research logic for obtaining the optimized numerical data; The operations research logic is combined with the business background information to generate business logic, and a business explanation corresponding to the operations optimization problem is generated according to the business logic.

4. The method according to claim 3, characterized in that: Generating the business explanation corresponding to the operations optimization problem according to the business logic includes: Based on a pre-trained language generation model, convert the business logic into natural language text; A business explanation corresponding to the operations optimization problem is generated according to the natural language text.

5. The method according to claim 4, characterized in that The converting the business logic into natural language text based on the pre-trained language generation model includes: Adaptively adjusting the language generation model using the business background information of the operations research optimization problem; Based on the adaptively adjusted language generation model, the business logic is converted into natural language text.

6. The method according to claim 3, characterized in that The step of identifying and acquiring the operations research logic for optimizing the numerical data based on the pre-established operations research knowledge base and according to the operations research model comprises: Determine the business complexity corresponding to the operations optimization problem based on the business background information, and determine the logic depth according to the business complexity; Based on a pre-established knowledge base and according to the operations research model, operations research logic deeply matching the logic is identified and acquired.

7. The method according to claim 3, characterized in that The step of identifying and acquiring the operations research logic for optimizing the numerical data based on the pre-established operations research knowledge base and according to the operations research model comprises: The logic suggestion sent by the user is received according to the interactive interface, and based on the operations research knowledge related to the logic suggestion in a pre-established operations research knowledge base and according to the operations research model, the operations research logic for optimizing the numerical data is identified and acquired.

8. A semantic device for operations research results driven by a large model, characterized in that: include: A data acquisition module, used to acquire the optimization numerical data corresponding to the operations optimization problem and the business background information of the operations optimization problem; A large model module, used to combine the optimized numerical data with the business background information based on a pre-trained large model, and identify the business elements corresponding to the optimized numerical data; A language generation module is used to generate a business explanation corresponding to the operations optimization problem based at least on the corresponding relationship between the optimization numerical data and the business elements.

9. An electronic device, characterized in that: include: A processor, and a memory communicatively connected to the processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory, so that the processor performs the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, which are used to implement the method according to any one of claims 1 to 7 when executed by a processor.

Images