Information processing method based on large model and electronic equipment

Abstract

The invention provides an information processing method based on a large model and electronic equipment. The method comprises the following steps: based on a pre-trained large model, generating question information at least according to a demand condition for establishing an operation planning optimization model; displaying the question information on an interactive interface, and obtaining answer information of a user to the question information; and analyzing the answer information, and obtaining and storing condition information corresponding to a demand condition for establishing the operation planning optimization model. According to the method, the condition information of the demand condition for establishing the operation planning optimization model is sorted based on the dialogue with the user of the large model. Through the large model, dependence on experts in the condition information carding process can be reduced, and the carding efficiency is improved. Meanwhile, compared with manual carding, based on large model carding, the consistency and stability of carding task execution can be guaranteed.

Description

Technical Field

[0001] This application relates to the field of large model technologies, and more specifically, to an information processing method based on a large model. Background Art

[0002] In modern enterprise operations research and optimization projects, the core part of modeling usually requires business consultants or algorithm engineers to deeply engage in the user's business scenarios and sort out user requirements through on-site research.

[0003] However, this process highly depends on expert experience, with low efficiency and difficulty in large-scale promotion. In addition, users often lack in-depth understanding of the terms and methods in the field of operations research and optimization, further increasing the complexity of requirement sorting. Therefore, how to use artificial intelligence technology to improve the efficiency and accuracy of requirement sorting has become an important research topic in the field of operations research and optimization. Summary of the Invention

[0004] In view of the above problems, this application aims to provide an information processing method and an electronic device based on a large model to solve the problem that in the existing technology, the process of sorting out requirements in the modeling of operations research and optimization models depends on expert experience, resulting in low sorting efficiency and difficulty in large-scale promotion.

[0005] In a first aspect, this application provides an information processing method based on a large model, and the method includes:

[0006] Based on a pre-trained large model, generate question information at least according to the requirement conditions for establishing an operations research and optimization model;

[0007] Display the question information on an interactive interface and obtain the response information of the user to the question information;

[0008] Analyze the response information to obtain and store the condition information corresponding to the requirement conditions for establishing the operations research and optimization model.

[0009] In a possible implementation manner, the generating question information based on a pre-trained large model at least according to the requirement conditions for establishing an operations research and optimization model includes:

[0010] Determine a to-be-processed requirement condition from one or more unprocessed requirement conditions according to the logical order for establishing the operations research and optimization model; each of the unprocessed requirement conditions is one of the multiple requirement conditions for establishing the operations research and optimization model;

[0011] Based on the large model, generate a question message at least according to the to-be-processed requirement condition; the question information is used to guide the user to input condition information related to the to-be-processed requirement condition;

[0012] Correspondingly, analyzing the response information to obtain and store the condition information corresponding to the requirement conditions for establishing the operational research optimization model, including:

[0013] Analyze the response information to obtain and store the condition information related to the to-be-processed requirement conditions.

[0014] In a possible implementation manner, after analyzing the response information to obtain and store the condition information related to the to-be-processed requirement conditions, it further includes:

[0015] Obtain the stored condition information related to the to-be-processed requirement conditions, and based on the to-be-processed requirement conditions, perform integrity verification on the stored condition information related to the to-be-processed requirement conditions;

[0016] In the case where the integrity verification passes, delete the to-be-processed requirement conditions from the unprocessed requirement conditions.

[0017] In a possible implementation manner, generating a question message based on the large model, at least according to the to-be-processed requirement conditions, includes:

[0018] Obtain the historical question information generated based on the large model and the historical answer information corresponding to the historical question information;

[0019] Based on the large model, generate question information according to the to-be-processed requirement conditions, and optimize the question information according to the historical question information and the historical answer information.

[0020] In a possible implementation manner, analyzing the response information to obtain and store the condition information corresponding to the requirement conditions for establishing the operational research optimization model, includes:

[0021] Analyze the response information to obtain the condition information corresponding to the requirement conditions for establishing the operational research optimization model, and convert it into structured data;

[0022] Obtain the stored historical structured data, integrate the historical structured data with the structured data obtained according to the response information, and store the integrated structured data as new historical structured data.

[0023] In a possible implementation manner, storing the integrated structured data as new historical structured data includes:

[0024] Display the integrated structured data on the interaction interface for the user to modify the integrated structured data;

[0025] In the case where the user modifies the integrated structured data, the modified integrated structured data is stored as new historical structured data.

[0026] In a possible implementation manner, the requirement conditions for establishing the operations research optimization model include any one or more of the following:

[0027] The application scenario corresponding to the operations research optimization model;

[0028] The decision variables of the operations research optimization model;

[0029] The objective function of the operations research optimization model;

[0030] The constraint conditions of the operations research optimization model;

[0031] The model parameters of the operations research optimization model.

[0032] In a possible implementation manner, the method further includes:

[0033] When obtaining the condition information corresponding to the decision variables, the condition information corresponding to the objective function, and the condition information corresponding to the model parameters, by analyzing the correlation between the condition information corresponding to the decision variables, the condition information corresponding to the objective function, and the condition information corresponding to the model parameters, optimization suggestions for the operations research optimization model are generated.

[0034] In a possible implementation manner, after generating the question information based on the pre-trained large model at least according to the requirement conditions for establishing the operations research optimization model, it further includes:

[0035] Using natural language technology to parse the question information to obtain the operations research terms in the question information;

[0036] Obtain the academic explanations corresponding to the operations research terms and display the academic explanations on the interaction interface.

[0037] In a second aspect, the present application provides an information processing device based on a large model, including:

[0038] A question generation module, configured to generate question information based on the pre-trained large model at least according to the requirement conditions for establishing the operations research optimization model;

[0039] An information acquisition module, configured to display the question information on the interaction interface and obtain the answer information of the user to the question information;

[0040] An information processing module, configured to analyze the answer information to obtain and store the condition information corresponding to the requirement conditions for establishing the operations research optimization model.

[0041] 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;

[0042] The memory stores computer-executable instructions;

[0043] The processor executes the computer-executable instructions stored in the memory to implement the method in any possible implementation manner of the above first aspect.

[0044] 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 above first aspect.

[0045] An information processing method and an electronic device based on a large model provided by the present application complete the sorting of conditional information for the requirements conditions for establishing an operations research optimization model through a dialogue with the user based on the large model. Through the large model, the dependence on experts in the process of sorting conditional information can be reduced, and the sorting efficiency can be improved. At the same time, compared with manual sorting, the consistency and stability of the sorting task execution can be ensured based on the large model sorting. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] The drawings herein are incorporated into the specification and form a part of the specification, showing embodiments consistent with the present application, and are used together with the specification to explain the principles of the present application.

[0047] Figure 1 It is a schematic flowchart of an information processing method based on a large model provided by an embodiment of the present application;

[0048] Figure 2 It is a schematic flowchart of partial steps of another information processing method based on a large model provided by an embodiment of the present application;

[0049] Figure 3 It is a schematic flowchart of partial steps of another information processing method based on a large model provided by an embodiment of the present application;

[0050] Figure 4 It is a schematic flowchart of partial steps of another information processing method based on a large model provided by an embodiment of the present application;

[0051] Figure 5 It is a schematic diagram of the execution process of a specific embodiment of an information processing method based on a large model provided by an embodiment of the present application;

[0052] Figure 6 It is a schematic structural diagram of an information device based on a large model provided by an embodiment of the present application;

[0053] Figure 7 This is a hardware structure diagram of an electronic device provided by an embodiment of the present application.

[0054] Through the above-mentioned drawings, specific embodiments of the present application have been shown, and more detailed descriptions will be given hereinafter. These drawings and written 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.

[0055] Description of reference numerals:

[0056] 601 - Question generation module; 602 - Information acquisition module; 603 - Information processing module

[0057] 701 - Processor; 702 - Memory; 703 - Communication interface; 704 - Communication bus. Detailed implementation manners

[0058] Here, the exemplary embodiments will be described in detail, and the examples are shown in the drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementation manners described in the following exemplary embodiments do not represent all implementation manners 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.

[0059] In the embodiments of the present application, words such as "first" and "second" are used to distinguish the same items or similar items with basically the same functions and effects. Those skilled in the art can understand that the words "first" and "second" do not limit the quantity and execution order, and the words "first" and "second" do not necessarily mean different.

[0060] 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. Exactly speaking, 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.

[0061] It should be noted that in the embodiments of the present application, "when...", it 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.

[0062] Operational optimization is an important branch of operations research and is widely used in production planning, logistics management, resource allocation and other fields.

[0063] In traditional operations research and optimization modeling projects, there is a significant cognitive gap between customers and technical personnel. Due to the lack of an effective demand guidance mechanism, customers often find it difficult to accurately express their business needs, and technical personnel also find it difficult to quickly understand and transform these vague demand expressions. This two-way understanding barrier requires a lot of time to be invested in repeated communication and confirmation in the early stages of the project, affecting the efficiency of project advancement.

[0064] At the same time, as a highly specialized job, operations optimization modeling has extremely high requirements for talent. However, there are a limited number of algorithm engineers with relevant professional backgrounds and project experience in the market, and they often have multiple roles. This talent scarcity creates a gap with the growth of project demand, resulting in project delays or failures due to lack of suitable personnel.

[0065] Moreover, the demand research process often relies on the personal experience and intuition of researchers. This unstructured research method is prone to problems such as incomplete information collection and inaccurate focus, which affects the quality and efficiency of subsequent modeling work. The collation and archiving of scattered research information also increases the management cost of the project. The experience and best practices accumulated in the project are difficult to systematically preserve and pass on. This results in similar projects being unable to fully learn from existing experience, and each new project needs to conduct demand research and analysis from scratch.

[0066] These prominent technical defects not only affect the implementation efficiency of operations optimization projects, but also restrict the development speed of the entire industry. Especially in the context of the current market's growing demand for intelligent decision-making, the solution to these problems has become increasingly urgent.

[0067] In order to solve the above technical problems, the embodiments of the present application provide an information processing method based on a large model, an information processing device based on a large model, a computer-readable storage medium, and an electronic device, which completes the combing of condition information of the demand conditions for establishing an operations optimization model through dialogue with the user based on the large model. Through the large model, the reliance on experts in the process of combing condition information can be reduced, and the efficiency of combing can be improved. At the same time, compared with manual combing, the consistency and stability of combing task execution can be guaranteed based on combing based on the large model.

[0068] Figure 1 A flowchart of an information processing method based on a large model provided in an embodiment of the present application is shown in FIG. Figure 1 As shown, the specific embodiment process of the method may include the following steps:

[0069] S110: Based on the pre-trained large model, generate question information at least according to the requirements for establishing the operations research optimization model.

[0070] Among them, the pre-trained large model can be an LLM (Large Language Model).

[0071] An LLM is a deep learning model trained using a large amount of text data, which can generate natural language text or understand the meaning of language text. Its characteristics are large scale and a huge number of parameters (usually reaching above the tens of billions level), and it is usually based on deep learning architectures such as the Transformer architecture. The difference between an LLM and an ordinary pre-trained language model lies in the parameter scale. When the parameter scale exceeds a certain level, the model achieves a significant performance improvement and exhibits capabilities that small models do not have, such as in-context learning ability, being able to learn complex patterns in language, and performing a wide range of tasks, including text summarization, modification, translation, sentiment analysis, multi-turn conversations, and so on.

[0072] Common LLMs include: GPT-3 (Generative Pre-trained Transformer 3), T5 (Text-to-Text Transfer Transformer), GPT-4, PaLM (a large language model proposed by Google), LLaMA (Large Language Model Meta AI, a large language model released by Meta AI), and so on.

[0073] Among them, the large language model used in the embodiments of this application can be a GPT (Generative Pre-trained Transformer) model.

[0074] In recent years, with the rapid development of NLP (Natural Language Processing) technology and artificial intelligence technology, natural language engines based on LLMs, such as GPT models, have performed excellently in tasks such as text generation and question answering. Generating teaching problem instances is essentially a text generation problem. Therefore, information for asking questions can be generated based on an LLM.

[0075] The large model used in the embodiments of this application is set as an "operations research modeling expert", and its main responsibility is to guide the user to complete information collection in the form of a dialogue.

[0076] To establish an operations research optimization model for an operations research optimization project, it is at least necessary to determine the scenario and background information of the operations research optimization project, controllable variables, project objectives, constraint conditions, etc. The information collected by the large model guiding the user in the form of a dialogue should also at least include the scenario and background information, controllable variables, objectives, constraint conditions, etc. The requirement conditions for establishing an operations research optimization model at least include the scenario and background information, controllable variables, objectives, constraint conditions, etc.

[0077] Based on the large model, generate question information at least according to the requirement conditions for establishing the operations research optimization model, that is, generate question information for guiding the user to input scenario and background information based on the large model, generate question information for guiding the user to input relevant information of variables that can be controlled based on the large model, generate question information for guiding the user to input relevant information of the goal based on the large model, and generate question information for guiding the user to input relevant information of constraint conditions based on the large model.

[0078] S120: Display the question information on the interaction interface and obtain the user's response information to the question information.

[0079] Among them, the interaction interface can be a conversational interaction interface between the user and the large model, and the user can input response information to the question information on the interaction interface through the interaction device.

[0080] 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. The interaction interface is displayed on the display device, and the user inputs response information on the interaction interface through interaction devices such as a mouse and a keyboard.

[0081] In one implementation, the user can be prompted to answer the question information in natural language as much as possible to facilitate the analysis of the response information, and the use of mathematical formulas or LaTeX (a language for mathematical formulas) expressions is prohibited.

[0082] S130: Analyze the response information, and obtain and store the condition information corresponding to the requirement conditions for establishing the operations research optimization model.

[0083] In one implementation, analyzing the response information can be based on the large model to analyze the response information.

[0084] Utilizing the context learning ability of the large model, it is possible to learn the complex patterns in the response information, summarize and extract the response information, and obtain the condition information related to the requirement conditions for establishing the operations research optimization model from the response information.

[0085] In another implementation, analyzing the response information can be based on NLP technology to analyze the response information.

[0086] Specifically, NLP technology can be used to segment the response information and perform word segmentation on the segmentation results to obtain the condition information related to the requirement conditions for establishing the operations research optimization model.

[0087] After obtaining the condition information corresponding to the requirements for establishing the operations research optimization model, the condition information can be stored and displayed on the interaction interface for the user to obtain the sorting result of the operations research optimization project, that is, the information required to establish the operations research optimization model for the operations research optimization project.

[0088] The following provides a specific introduction to an information processing method based on a large model provided by an embodiment of the present application.

[0089] In one implementation, the requirement conditions for establishing an operations research optimization model may include:

[0090] Application scenario: The operations research optimization model scenario and background information.

[0091] Decision variables: The key variables controlled by the user.

[0092] Objective function: The objective for establishing the operations research optimization model.

[0093] Constraint conditions: Limiting conditions.

[0094] Model parameters: The fixed input parameters of the model.

[0095] Among them, decision variables, objective functions, and constraint conditions are the constituent elements of the operations research optimization model, jointly constituting the basic framework of the operations research optimization model for solving practical problems.

[0096] Decision variables are the unknowns to be determined in the model, representing specific solutions or measures in the planning problem. They can be continuous or discrete, but usually in operations research optimization, decision variables can be integers or real numbers.

[0097] The objective function is a function of the decision variables, used to represent the optimization objective. Its purpose is to maximize or minimize a certain index, usually economic benefits, cost minimization, or other quantitative objectives.

[0098] Constraint conditions limit the value range of the decision variables to ensure that the solution is feasible in practical applications. These constraints usually appear in the form of equations or inequalities, reflecting various limiting conditions of the problem.

[0099] In one implementation, the information processing method based on a large model provided by an embodiment of the present application can be executed multiple times, that is, the large model can have multiple rounds of conversations with the user, and the question information generated in each round of conversation can be question information generated based on a requirement condition for establishing an operations research optimization model.

[0100] In each round of conversation between the large model and the user, at least one condition information corresponding to a requirement condition is obtained. Through multiple rounds of conversation, multiple condition information corresponding to different requirement conditions is obtained.

[0101] Figure 2 Shows a flowchart of an implementation manner where each round of conversation is based on a requirement condition for establishing an operations research optimization model. As Figure 2 shown, it may include the following steps:

[0102] S111. Determine a to-be-processed requirement condition from one or more unprocessed requirement conditions according to the logical sequence of establishing an operational research optimization model; each unprocessed requirement condition is one of the multiple requirement conditions for establishing the operational research optimization model.

[0103] Among them, the logical sequence of establishing the operational research optimization model can be the sequence of scenario description, decision variables, objective function, constraint conditions, and model parameters.

[0104] In the case where this round of dialogue is the first round of dialogue between the large model and the user, the scenario description, decision variables, objective function, constraint conditions, and model parameters are all unprocessed requirement conditions.

[0105] In the case where this round of dialogue is other rounds of dialogue between the large model and the user, the requirement conditions corresponding to the condition information obtained in other rounds of dialogue before this round of dialogue are the processed requirement conditions, and the other requirement conditions except the processed requirement conditions are unprocessed requirement conditions.

[0106] S112. Based on the large model, generate a question message at least according to the to-be-processed requirement condition; the question message is used to guide the user to input condition information related to the to-be-processed requirement condition.

[0107] In one implementation, in the case where this round of dialogue is other rounds of dialogue between the large model and the user, the question message can be generated according to the historical dialogue between the large model and the user, that is, other rounds of dialogue before this round of dialogue combined with the to-be-processed requirement condition.

[0108] Specifically, obtain the historical question information generated based on the large model and the historical answer information corresponding to the historical question information; based on the large model, generate question information according to the to-be-processed requirement condition, and optimize the question information according to the historical question information and the historical answer information.

[0109] Among them, the historical question information generated based on the large model and the historical answer information corresponding to the historical question information, that is, the historical dialogue between the large model and the user, that is, other rounds of dialogue before this round of dialogue combined with the to-be-processed requirement condition to generate the question message.

[0110] Based on the large model, the specific process of generating question information according to the to-be-processed requirement condition can be as described in S110, which will not be elaborated here.

[0111] Optimization according to the historical question information and the historical answer information can be to analyze the language patterns of the historical question information and the historical answer information, obtain the language pattern that the user prefers more, and modify the question information into the corresponding language model.

[0112] For example, if the user prefers more colloquial expressions, convert the written language in the question information into colloquial expressions.

[0113] Optimization based on historical question information and historical answer information can also be to obtain the information that appears in the historical question information and historical answer information, and match the information that appears with the question information. If a match can be made, it means that the information that appears in the historical question information and historical answer information can answer the question information, and the generated question information is a duplicate question information. Therefore, the question information can be regenerated.

[0114] In one implementation, the generated question information may contain professional operations research terms, and users may have insufficient understanding of operations research concepts and methods and be unable to understand the operations research terms.

[0115] Therefore, the operations research terms can be automatically analyzed and explained to lower the user's cognitive threshold and better understand the question information.

[0116] In one implementation, using natural language technology, parse the question information to obtain the operations research terms in the question information; obtain the academic explanations corresponding to the operations research terms, and display the academic explanations on the interaction interface.

[0117] Specifically, use a word segmentation tool to segment the question information to split the question information into multiple words, and use the words corresponding to the question information to match in the integrated operations research professional knowledge base (including multiple operations research terms and the academic explanations corresponding to the operations research terms). When an operations research term is matched, obtain the academic explanation corresponding to the operations research term, and display the optimized question information, the operations research term, and the corresponding academic explanation on the interaction interface.

[0118] In one implementation, the condition information related to the demand conditions may include condition information in different aspects, that is, the condition information related to the demand conditions includes multiple condition information. For example, the application scenarios corresponding to the operations research optimization model include industry characteristics, business processes, and resource allocation, etc.

[0119] Therefore, multiple rounds of conversations may be required to obtain the complete condition information related to the demand conditions through progressive questioning.

[0120] After obtaining a condition information through one round of conversation, it is necessary to perform an integrity check on the condition information that has been obtained and stored together to determine whether the complete condition information related to the demand conditions has been obtained. Only when the complete condition information related to the demand conditions has been obtained can the to-be-processed demand conditions be regarded as processed demand conditions and deleted from the unprocessed demand conditions. Figure 3 Shows a schematic flowchart of one implementation, such as Figure 3 As shown, the following steps may be included:

[0121] S113. Obtain the condition information related to the to-be-processed requirement conditions stored, and perform integrity verification on the condition information related to the to-be-processed requirement conditions stored based on the to-be-processed requirement conditions.

[0122] Among them, since this step is executed after S130, therefore, the condition information obtained according to the question information generated by S111 and S112 has also been stored.

[0123] The condition information related to the to-be-processed requirement conditions stored includes the condition information related to the to-be-processed requirement conditions obtained according to this round of conversation and other rounds of conversations before this round of conversation.

[0124] Perform integrity verification on all the condition information related to the to-be-processed requirement conditions stored. Specifically, the condition information related to the to-be-processed requirement conditions can be matched with different aspects of the to-be-processed requirement conditions to determine whether the condition information related to the to-be-processed requirement conditions stored covers all aspects of the to-be-processed requirement conditions.

[0125] S114. In the case where the integrity verification passes, delete the to-be-processed requirement conditions from the unprocessed requirement conditions.

[0126] Only in the case where the integrity verification passes can the to-be-processed requirement conditions be deleted from the unprocessed requirement conditions as processed requirement conditions.

[0127] In the case where the integrity verification fails, continue to regard the to-be-processed requirement conditions as unprocessed requirement conditions.

[0128] Since the to-be-processed requirement conditions are determined from the unprocessed requirement conditions according to the logical order of establishing the operation research optimization model, therefore, when the information processing method based on the large model provided by the embodiment of the present application is executed again, the to-be-processed requirement conditions are determined from the unprocessed requirement conditions according to the logical order of establishing the operation research optimization model, and the determined to-be-processed requirement conditions are the same as those determined this time, so that the condition information related to the to-be-processed requirement conditions can be continuously obtained until the integrity verification passes.

[0129] In one implementation, analyzing the answer information can be to convert the answer information into structured data to facilitate the storage of the condition information.

[0130] Figure 4 Shows a flow schematic diagram of an implementation manner of converting the answer information into structured data and storing the structured data, as Figure 4 shown, it may include the following steps:

[0131] S131. Analyze the answer information, obtain the condition information corresponding to the requirement conditions for establishing the operation research optimization model, and convert it into structured data.

[0132] Among them, the process of analyzing the response information to obtain the conditional information corresponding to the requirement conditions for establishing the operational research optimization model is as described in S130, which will not be elaborated here.

[0133] Converting the conditional information into structured data can be to extract the correspondence between nouns and numerical values in the submitted information, and save the nouns, numerical values, and the corresponding relationships in the form of a data table.

[0134] S132. Obtain the stored historical structured data, integrate the historical structured data with the structured data obtained according to the response information, and store the integrated structured data as the new historical structured data.

[0135] Among them, the historical structured data can be the structured data obtained and stored in other rounds of conversations before this round of conversation.

[0136] Integrating the historical structured data with the structured data obtained according to the response information can be to update the historical structured data using the structured data obtained according to the response information. If the same structured data is stored, it only needs to be saved once.

[0137] Storing the integrated structured data as the new historical structured data can ensure that the structured data obtained each time is the integrated structured data, and there are no duplicate or conflicting data.

[0138] After obtaining and storing the integrated structured data, display the integrated structured data on the interaction interface for the user to modify the integrated structured data; in the case where the user modifies the integrated structured data, store the modified integrated structured data as the new historical structured data.

[0139] That is to say, after each round of conversation between the large model and the user ends, the obtained integrated structured data is fed back to the user to ensure that the user can supplement or correct it in a timely manner. After the user makes supplements and corrections, save the user's supplements and corrections.

[0140] In one implementation, after multiple rounds of conversations to obtain all the requirement conditions, that is, the conditional information corresponding to the application scenario, decision variables, objective function, constraints, and model parameters, by analyzing the relevance between the conditional information corresponding to the decision variables, the conditional information corresponding to the objective function, and the conditional information corresponding to the model parameters, generate optimization suggestions for the operational research optimization model.

[0141] Specifically, based on the professional knowledge base and existing model cases, generate detailed modeling suggestions, including possible optimization directions, potential problems, and avoidance strategies.

[0142] As described above, the information processing method based on the large model provided by the embodiments of the present application has the following advantages:

[0143] By integrating the semantic understanding ability of the large model, the complex requirement sorting task is adaptively decomposed into five core modules: application scenarios, decision variables, objective functions, constraint conditions, and model parameters. This semantic-based intelligent decomposition not only ensures the accurate positioning and comprehensive coverage of requirement collection, but also establishes a complete requirement verification system through the logical association analysis between modules. On this basis, the solution realizes a progressive requirement refinement process, supporting users to continuously deepen their understanding and definition of problems during the interaction process.

[0144] Through real-time semantic understanding technology, the system can accurately parse user input, identify key information points and potential requirements. Notably, its context-aware optimization ability enables the system to not only intelligently track the conversation progress, dynamically adjust the prompting strategy, but also optimize the subsequent conversation direction based on historical interaction records, automatically identify information gaps, and actively guide users to supplement key elements. This multi-dimensional quality control mechanism ensures the accuracy and integrity of the requirement sorting process.

[0145] By organically combining the professional knowledge of operations research with the conversation ability of the large model, while maintaining professionalism, it ensures that the expression is easy to understand, effectively reducing the cognitive threshold of users. The system can dynamically generate modeling suggestions based on the collected information, provide multi-level optimization solutions, and predict potential modeling pitfalls to provide avoidance strategies. This intelligent professional guidance significantly improves the quality and efficiency of requirement sorting.

[0146] Through the decoupled design of the prompt word framework and the specific large model, true model independence is achieved, supporting cross-model migration and adaptation, and ensuring the stability of core functions. At the same time, the solution has excellent scenario adaptability, can support multi-domain requirement sorting scenarios, dynamically adjust the interaction strategy according to different business characteristics, and flexibly adapt to problems of different complexities. In terms of capacity expansion, the system supports the rapid integration of new knowledge domains, allows customization of professional term libraries, and reserves rich tool chain expansion interfaces.

[0147] The solution significantly reduces the cognitive burden of users through the design of an intelligent workflow. The multi-level information verification system, intelligent consistency check, and automated integrity assessment jointly build an all-round quality assurance mechanism. Through the reuse of prompt word templates, context information management using deep learning models, and optimized conversation turn control, the optimal utilization of system resources is achieved, ensuring the efficiency and reliability of the requirement sorting process.

[0148] The following uses a specific embodiment to illustrate the information processing method based on the large model provided by the embodiments of the present application.

[0149] As Figure 5 shown, after the system starts, it is customized as the "Operations Research Modeling Expert" and clearly states the task objective to the user, that is, to structurally sort out the modeling requirements of operations research optimization problems. The system first guides the user to describe the scenario and background information, including industry characteristics, business processes, and resource allocation. Through progressive questioning, the integrity of the collected information is ensured.

[0150] In accordance with the logical order of the modeling requirements, the following information is collected one by one:

[0151] Decision variables: Identify the core variables controlled by the user.

[0152] Objective function: Collect the modeling objectives and their priorities.

[0153] Constraints: Collect business restrictions and rules and verify their rationality.

[0154] Model parameters: List the required fixed input variables and their sources.

[0155] After each round of conversation ends, the system integrates the collected information and feedbacks it to the user in a structured manner to ensure that the user can supplement or correct it in a timely manner.

[0156] After the collection is completed, the system generates detailed modeling suggestions based on the professional knowledge base and existing model cases, including possible optimization directions, potential problems, and avoidance strategies.

[0157] If the user is not satisfied with the collected information, return to the information collection step; if the user is satisfied with the collected information, end the process.

[0158] Figure 6 The structural schematic diagram of an information processing device based on a large model provided by an embodiment of the present application, and the device can be in the form of software and / or hardware. Refer to Figure 6 shown, an information processing device based on a large model includes: a question generation module 601, an information acquisition module 602, and an information processing module 603.

[0159] The question generation module 601 is configured to generate question information based on a pre-trained large model and at least according to the requirement conditions for establishing an operations research optimization model;

[0160] The information acquisition module 602 is configured to display the question information on the interaction interface and obtain the answer information of the user to the question information;

[0161] The information processing module 603 is configured to analyze the answer information and obtain and store the condition information corresponding to the requirement conditions for establishing an operations research optimization model.

[0162] In a possible implementation manner, the question generation module 601 is specifically configured as:

[0163] Determine a to-be-processed requirement condition from one or more unprocessed requirement conditions according to the logical order of establishing an operational research optimization model; each unprocessed requirement condition is one of the multiple requirement conditions for establishing the operational research optimization model.

[0164] Based on the large model, generate a question message at least according to the to-be-processed requirement condition; the question message is used to guide the user to input condition information related to the to-be-processed requirement condition.

[0165] Correspondingly, the information processing module 603 is configured to: analyze the answer information, and obtain and store the condition information related to the to-be-processed requirement condition.

[0166] In a possible implementation manner, the information processing device based on the large model further includes an inspection module, which is specifically configured to:

[0167] Obtain the condition information related to the to-be-processed requirement condition stored, and perform integrity verification on the stored condition information related to the to-be-processed requirement condition based on the to-be-processed requirement condition.

[0168] In the case where the integrity verification passes, delete the to-be-processed requirement condition from the unprocessed requirement conditions.

[0169] In a possible implementation manner, the question generation module 601 is configured to:

[0170] Obtain the historical question information generated based on the large model and the historical answer information corresponding to the historical question information;

[0171] Based on the large model, generate question information according to the to-be-processed requirement condition, and optimize the question information according to the historical question information and the historical answer information.

[0172] In a possible implementation manner, the information processing module 603 is configured to:

[0173] Analyze the answer information, obtain the condition information corresponding to the requirement condition for establishing the operational research optimization model, and convert it into structured data;

[0174] Obtain the stored historical structured data, integrate the historical structured data with the structured data obtained according to the answer information, and store the integrated structured data as the new historical structured data.

[0175] In a possible implementation manner, the information processing module 603 is configured to:

[0176] Display the integrated structured data on the interaction interface for the user to modify the integrated structured data.

[0177] In the case where the user modifies the integrated structured data, the modified integrated structured data is stored as new historical structured data.

[0178] In a possible implementation, the requirements for establishing an operations research optimization model include any one or more of the following:

[0179] The application scenario corresponding to the operations research optimization model;

[0180] The decision variables of the operations research optimization model;

[0181] The objective function of the operations research optimization model;

[0182] The constraint conditions of the operations research optimization model;

[0183] The model parameters of the operations research optimization model.

[0184] In a possible implementation, the information processing device based on a large model further includes a suggestion module, which is specifically configured to:

[0185] When obtaining the condition information corresponding to the decision variables, the condition information corresponding to the objective function, and the condition information corresponding to the model parameters, by analyzing the relevance among the condition information corresponding to the decision variables, the condition information corresponding to the objective function, and the condition information corresponding to the model parameters, generate optimization suggestions for the operations research optimization model.

[0186] In a possible implementation, the information processing device based on a large model further includes an explanation module, which is specifically configured to:

[0187] Use natural language technology to parse the question information and obtain the operations research terms in the question information;

[0188] Obtain the academic explanations corresponding to the operations research terms and display the academic explanations on the interaction interface.

[0189] Figure 7 This is a 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 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 information processing method based on a large model described in any of the previous embodiments.

[0190] Figure 7The electronic device shown also 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 convenience of representation, Figure 7 only a thick line is used to represent the communication bus 704 in Figure 7 , 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.

[0191] 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, at least one processor 701 can execute the information processing method based on the large model described above. The processor 701 can implement Figure 7 the functions of each module in the device shown.

[0192] Among them, the processor 701 is the control center of the device. It can use various interfaces and lines to connect all parts of the entire control device. By running or executing the instructions stored in the memory 702 and calling the data stored in the memory 702, various functions of the device and process data, so as to monitor the device as a whole.

[0193] 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 modem processor may not be integrated into the processor 701. In some embodiments, the processor 701 and the memory 702 can be implemented on the same chip or separately on independent chips.

[0194] The processor 701 can 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, and can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of the present application. The general-purpose processor can be a microprocessor or any conventional processor, etc. The steps of the information processing method based on the large model disclosed in combination with the embodiments of the present application can be directly embodied as being executed by a hardware processor, or executed by a combination of hardware and software modules in the processor.

[0195] 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 can include at least one type of storage medium. For example, it can 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 can also be a circuit or any other device capable of implementing a storage function, for storing program instructions and / or data.

[0196] By programming the design of the processor 701, the code corresponding to the information processing method based on the large model introduced in the foregoing embodiments can be solidified into the chip, so that the chip can execute Figures 1 to 5 the steps of the information processing method based on the large model shown in the embodiments. How to program the design of the processor 701 is a well-known technology to those skilled in the art and will not be elaborated here.

[0197] The embodiments of the present application also provide a computer-readable storage medium. The computer-readable storage medium stores computer-executable instructions. When the computer-executable instructions are executed by a processor, they are used to implement the information processing method based on the large model described in any of the foregoing embodiments. Therefore, it will not be elaborated here. In addition, the description of the beneficial effects of using 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.

[0198] In some possible implementation manners, various aspects of the information processing method based on the large model 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 information processing method based on the large model according to various exemplary embodiments of the present application described above in this specification.

[0199] 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.

[0200] 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 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 a device for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0201] 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 implements the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0202] 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 implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0203] 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 large model-based information processing method, characterized in that: The method comprises: Based on the pre-trained large model, at least generate question information according to the requirements for establishing the operations optimization model; Displaying the question information on the interactive interface and obtaining the user's answer information to the question information; The answer information is analyzed to obtain and store condition information corresponding to the requirement conditions for establishing the operations research optimization model.

2. The method according to claim 1, characterized in that: The pre-trained large model generates question information at least according to the requirements for establishing the operations research optimization model, including: According to the logical order of establishing the operations research optimization model, a to-be-processed demand condition is determined from one or more unprocessed demand conditions; each of the unprocessed demand conditions is one of the multiple demand conditions for establishing the operations research optimization model; Based on the large model, a question message is generated at least according to the demand condition to be processed; the question message is used to guide the user to input condition information related to the demand condition to be processed; The step of analyzing the answer information to obtain and store condition information corresponding to the requirement conditions for establishing the operations research optimization model includes: The answer information is analyzed to obtain and store condition information related to the requirement condition to be processed.

3. The method according to claim 2, characterized in that After analyzing the answer information and obtaining and storing the condition information related to the requirement condition to be processed, the method further includes: Acquire the stored condition information related to the pending demand condition, and perform integrity check on the stored condition information related to the pending demand condition based on the pending demand condition; When the integrity check passes, the to-be-processed requirement condition is deleted from the unprocessed requirement conditions.

4. The method according to claim 2, characterized in that: The step of generating a question message based on the large model at least according to the requirement to be processed includes: Acquire historical question information generated based on the large model and historical answer information corresponding to the historical question information; Based on the large model, question information is generated according to the demand conditions to be processed, and the question information is optimized according to the historical question information and the historical answer information.

5. The method according to claim 1, characterized in that The step of analyzing the answer information to obtain and store condition information corresponding to the requirement conditions for establishing the operations research optimization model includes: Analyze the answer information, obtain condition information corresponding to the demand conditions for establishing the operations optimization model, and convert it into structured data; The stored historical structured data is acquired, the historical structured data is integrated with the structured data acquired according to the answer information, and the integrated structured data is stored as new historical structured data.

6. The method according to claim 5, characterized in that The storing of the integrated structured data as new historical structured data includes: Displaying the integrated structured data on the interactive interface so that the user can modify the integrated structured data; In the case where the user modifies the integrated structured data, the modified integrated structured data is stored as new historical structured data.

7. The method according to claim 1, characterized in that The requirements for establishing the operations research optimization model include any one or more of the following: Application scenarios corresponding to the operations research optimization model; Decision variables of the operations research optimization model; The objective function of the operations research optimization model; Constraints of the operations research optimization model; Model parameters of the operations research optimization model.

8. The method according to claim 7, characterized in that The method further comprises: When the condition information corresponding to the decision variables, the condition information corresponding to the objective function, and the condition information corresponding to the model parameters are obtained, optimization suggestions for the operations research optimization model are generated by analyzing the correlation between the condition information corresponding to the decision variables, the condition information corresponding to the objective function, and the condition information corresponding to the model parameters.

9. The method according to claim 1, characterized in that: After generating the question information based on the pre-trained large model at least according to the requirement conditions for establishing the operations research optimization model, the method further includes: Analyzing the question information by using natural language technology to obtain operations research terms in the question information; The academic explanation corresponding to the operations research term is obtained, and the academic explanation is displayed on the interactive interface.

10. 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 9.

Images