Operational research model recommendation method based on large language model and related equipment

Abstract

The embodiment of the invention belongs to the technical field of Internet, and relates to a big language model-based operational research model recommendation method, which comprises the following steps of: obtaining text information input in a model recommendation interface by a target user; identifying the text information, and determining model demand information corresponding to the target user; converting the model demand information into a corresponding target high-dimensional vector; performing retrieval in a preset vector database according to the target high-dimensional vector, and determining a recommended operational research model; and displaying the recommended operational research model on the model recommendation interface. The invention further provides an operational research model recommendation system based on the large language model, computer equipment and a storage medium. According to the method, deep understanding is carried out on the demand through the large language model, the user demand can be better matched with the corresponding operational research model, the problem of human intervention or large errors in a traditional recommendation method is avoided, and therefore the result accuracy is improved.

Description

Technical Field

[0001] This application relates to the field of Internet technologies and is applied to scenarios where users need to retrieve operations research models that meet certain conditions. In particular, it relates to a method for recommending operations research models based on large language models and related devices. Background Art

[0002] With the continuous maturity of deep learning, natural language processing (NLP), and operations research technologies, using large language models (LLMs) to assist in solving complex decision-making, optimization, and resource allocation problems has become a major trend in today's society.

[0003] Currently, traditional operations research models generally require domain experts to build models in combination with mathematical optimization methods, and appropriate algorithms or models need to be retrieved from professional literature or databases for application. However, in many actual scenarios, especially when dealing with complex operations research models, users often have difficulty using professional terms for retrieval, which makes it difficult for the system to understand the actual needs of users, resulting in the final operations research models not meeting the actual needs and poor user experience. Summary of the Invention

[0004] The purpose of the embodiments of this application is to propose a method for recommending operations research models based on large language models and related devices to solve the technical problem of low accuracy of the results obtained when users retrieve the required operations research models.

[0005] To solve the above technical problem, the embodiments of this application provide a method for recommending operations research models based on large language models, adopting the following technical solutions:

[0006] A method for recommending operations research models based on large language models includes the following steps:

[0007] Obtain the text information input by the target user on the model recommendation interface;

[0008] Identify the text information to determine the model requirement information corresponding to the target user;

[0009] Convert the model requirement information into a corresponding target high-dimensional vector;

[0010] Retrieve in a preset vector database according to the target high-dimensional vector to determine the recommended operations research model;

[0011] Display the recommended operations research model on the model recommendation interface.

[0012] Further, the step of identifying the text information and determining the model requirement information corresponding to the target user specifically includes:

[0013] Perform semantic understanding based on the text information to determine the requirement keywords and context information corresponding to the text information;

[0014] Perform structural transformation on the requirement keywords and the context information to obtain a target query condition;

[0015] Analyze according to the target query condition to obtain the user requirement and the description of the operations research model as the model requirement information.

[0016] Further, the step of transforming the model requirement information into a corresponding target high-dimensional vector specifically includes:

[0017] Input the user requirement and the description of the operations research model into a preset Embedding model;

[0018] Output the target high-dimensional vector through the Embedding model.

[0019] Further, the step of retrieving in a preset vector database according to the target high-dimensional vector and determining a recommended operations research model specifically includes:

[0020] Determine the similarity between the target high-dimensional vector and each preset semantic vector in the vector database according to a preset similarity algorithm;

[0021] Determine a target vector with a similarity greater than a preset threshold among each of the preset semantic vectors;

[0022] Use the operations research model corresponding to the target vector as the recommended operations research model.

[0023] Further, after the step of displaying the recommended operations research model on the model recommendation interface, it further includes:

[0024] When the number of the recommended operations research models is two or more, generate a model recommendation list on the model recommendation interface according to the recommended operations research models;

[0025] In response to a selection instruction triggered by the model recommendation interface, determine the target operations research model corresponding to the target user in the model recommendation list.

[0026] Further, after the step of displaying the recommended operations research model on the model recommendation interface, it further includes:

[0027] When the model recommendation interface receives the associated text information corresponding to the recommended operations research model, update the model requirement information according to the associated text information, and return to execute the step of converting the model requirement information into the corresponding target high-dimensional vector.

[0028] Further, after the step of displaying the recommended operations research model on the model recommendation interface, the following is also included:

[0029] Analyze the recommended operations research model to determine the model principle, application scenario, and advantages and disadvantages information corresponding to the recommended operations research model;

[0030] Display the model principle, the application scenario, and the advantages and disadvantages information on the model recommendation interface.

[0031] To solve the above technical problems, the embodiments of the present application also provide an operations research model recommendation system based on a large language model, adopting the following technical solutions:

[0032] An operations research model recommendation system based on a large language model includes:

[0033] An acquisition module, configured to acquire the text information input by the target user on the model recommendation interface;

[0034] An identification module, configured to identify the text information to determine the model requirement information corresponding to the target user;

[0035] A conversion module, configured to convert the model requirement information into the corresponding target high-dimensional vector;

[0036] A retrieval module, configured to retrieve in a preset vector database according to the target high-dimensional vector to determine a recommended operations research model;

[0037] A display module, configured to display the recommended operations research model on the model recommendation interface.

[0038] To solve the above technical problems, the embodiments of the present application also provide a computer device, adopting the following technical solutions:

[0039] A computer device includes a memory and a processor. Computer-readable instructions are stored in the memory, and when the processor executes the computer-readable instructions, the steps of the above-mentioned operations research model recommendation method based on a large language model are implemented.

[0040] To solve the above technical problems, the embodiments of the present application also provide a computer-readable storage medium, adopting the following technical solutions:

[0041] A computer-readable storage medium stores computer-readable instructions thereon, and when the computer-readable instructions are executed by a processor, the steps of the above-mentioned operation research model recommendation method based on a large language model are implemented.

[0042] Compared with the prior art, the embodiments of the present application mainly have the following beneficial effects:

[0043] The operation research model recommendation method based on a large language model disclosed in the present application includes: obtaining text information input by a target user on a model recommendation interface; identifying the text information to determine model requirement information corresponding to the target user; converting the model requirement information into a corresponding target high-dimensional vector; retrieving in a preset vector database according to the target high-dimensional vector to determine a recommended operation research model; and displaying the recommended operation research model on the model recommendation interface. By obtaining the text information input by the user on the model recommendation interface and identifying and analyzing it, the present application can accurately understand the user's needs, automatically convert the needs into high-dimensional vectors and perform matching, and deeply understand the needs through the large language model, so as to better match the user's needs with the corresponding operation research models, avoiding the problems of human intervention or large errors in traditional recommendation methods, thereby improving the result accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] To more clearly illustrate the solutions in the present application, the following will briefly introduce the drawings required for the description of the embodiments of the present application. Obviously, the drawings in the following description are some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0045] Figure 1 is an exemplary system architecture diagram to which the present application can be applied;

[0046] Figure 2 is a flowchart of an embodiment of the operation research model recommendation method based on a large language model according to the present application;

[0047] Figure 3 is a schematic structural diagram of an embodiment of the operation research model recommendation system based on a large language model according to the present application;

[0048] Figure 4 is a schematic structural diagram of an embodiment of a computer device according to the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0049] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the technical field to which this application belongs; the terms used in the specification of this application are only for the purpose of describing specific embodiments and are not intended to limit this application; the terms "including" and "having" and any variations thereof in the specification and claims of this application and the above drawings are intended to cover non-exclusive inclusion. The terms "first", "second", etc. in the specification and claims of this application or the above drawings are used to distinguish different objects and not to describe a specific order.

[0050] Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of this application. The phrase does not necessarily refer to the same embodiment at every occurrence in the specification, nor is it an independent or alternative embodiment mutually exclusive of other embodiments. Those skilled in the art will explicitly and implicitly understand that the embodiments described herein can be combined with other embodiments.

[0051] To enable those skilled in the art to better understand the solutions of this application, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings.

[0052] As Figure 1 shown, the system architecture 100 may include terminal devices 101, 102, 103, a network 104, and a server 105. The network 104 is used to provide a medium for communication links between the terminal devices 101, 102, 103 and the server 105. The network 104 may include various connection types, such as wired, wireless communication links, or fiber optic cables, etc.

[0053] Users may use the terminal devices 101, 102, 103 to interact with the server 105 through the network 104 to receive or send messages, etc. Various communication client applications may be installed on the terminal devices 101, 102, 103, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social platform software, etc.

[0054] The terminal devices 101, 102, 103 may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, e-book readers, MP3 (Moving Picture Experts Group Audio Layer III) players, MP4 (Moving Picture Experts Group Audio Layer IV) players, laptop portable computers, and desktop computers, etc.

[0055] Server 105 can be a server that provides various services, such as a background server that supports the pages displayed on the terminal devices 101, 102, and 103.

[0056] It should be noted that the method for recommending an operations research model based on a large language model provided by the embodiments of the present application is generally executed by a terminal device. Correspondingly, a system for recommending an operations research model based on a large language model is generally set in the terminal device.

[0057] It should be understood that Figure 1 the numbers of the terminal devices, networks, and servers in are merely illustrative. According to the implementation requirements, there can be any number of terminal devices, networks, and servers.

[0058] Continuing to refer to Figure 2 , a flowchart of an embodiment of the method for recommending an operations research model based on a large language model according to the present application is shown. The method for recommending an operations research model based on a large language model includes the following steps:

[0059] Step S201, obtain the text information input by the target user on the model recommendation interface.

[0060] In this embodiment, the electronic device (such as the terminal device shown in Figure 1 ) on which the method for recommending an operations research model based on a large language model runs can send or receive data through a wired connection or a wireless connection. It should be noted that the above wireless connection methods can include, but are not limited to, 3G / 4G / 5G connections, WiFi connections, Bluetooth connections, WiMAX connections, Zigbee connections, UWB (ultra wideband) connections, and other currently known or future-developed wireless connection methods.

[0061] In this embodiment, the user can input a natural language question or requirement description through the interface to express the need for a specific operations research model. For example, the user may input "I need a model that can optimize the transportation route", which means that the user hopes to recommend an optimization model related to logistics or transportation problems. The input text information is not limited to a single requirement description, but can also include various information in multiple rounds of conversations, such as the user's refined requirements or feedback on the model. Through multiple rounds of conversations, the system can more accurately capture the user's true needs.

[0062] Step S202, identify the text information to determine the model requirement information corresponding to the target user.

[0063] In this embodiment, after obtaining the text information, the system analyzes and understands the content input by the user through natural language processing (NLP) technology, and identifies keywords or key requirements. The system can use a pre-trained large language model (such as the GPT series) to convert the text into structured requirement information. This information may include objective functions, constraints, application scenarios, fields of problem-solving, etc. The identification process can utilize text parsing and intent recognition technology, and gradually refine the user's requirements through the multi-round dialogue ability of the large language model. For example, when the user enters a vague description, the system will ask for further details until the model requirements are accurately obtained.

[0064] Step S203: Convert the model requirement information into a corresponding target high-dimensional vector.

[0065] In this embodiment, this step involves converting the user requirement information into a high-dimensional vector representation. By using Embedding technology (such as OpenAI's text-embedding-ada-002 model), the semantic information in the text is mapped into a high-dimensional vector space. This step enables the user requirements to be processed mathematically and provides a basis for subsequent retrieval and comparison. The requirement information converted into a high-dimensional vector can better capture the implicit relationships in the semantics, which helps to find model descriptions semantically similar to it in the vector database. This step enables requirements in different languages or expressions to be uniformly processed through the vector space, ensuring the accuracy of semantic matching.

[0066] Step S204: Retrieve in a preset vector database according to the target high-dimensional vector to determine a recommended operations research model.

[0067] In this embodiment, the converted high-dimensional vector is compared with the vectors in the pre-stored operations research model library, and similarity calculations (such as cosine similarity, maximum marginal relevance, etc.) are used to find the most matching operations research model. This step utilizes a vector database (such as Chroma or FAISS) to quickly retrieve the most relevant model. The retrieved model should have semantic similarity with the user requirement information to ensure that the recommended model meets the actual requirements. The retrieval process can not only match based on the basic functions of the model, but also comprehensively consider the advantages and disadvantages of the model, applicable conditions, etc. By combining multiple algorithms (for example, considering both cosine similarity and maximum marginal relevance simultaneously), it is possible to avoid recommending a single similar model and enhance the diversity and accuracy of the recommendation.

[0068] Step S205: Display the recommended operations research model on the model recommendation interface.

[0069] In this embodiment, the retrieved and screened operations research models are presented to the user. These models include information such as their names, descriptions, applicable scenarios, principles, and applicable optimization problems. The user can view the recommended models and select the model that best meets their needs for further operations. The displayed interface can not only provide the basic information of the models but also further include detailed interpretation content such as case analyses and application suggestions of the models. These contents can be generated through the interpretation ability of the large language model to help the user better understand the application scenarios and implementation steps of the recommended models and avoid complex learning curves.

[0070] By obtaining the text information input by the user on the model recommendation interface and identifying and analyzing it, this application can accurately understand the user's needs, automatically convert the needs into high-dimensional vectors for matching, and deeply understand the needs through the large language model, which can better match the user's needs with the corresponding operations research models, avoiding the problems of human intervention or large errors in traditional recommendation methods, thereby improving the accuracy of the results.

[0071] In some alternative implementation manners of this embodiment, the step of identifying the text information and determining the model requirement information corresponding to the target user specifically includes:

[0072] Perform semantic understanding based on the text information to determine the demand keywords and context information corresponding to the text information;

[0073] Perform structural transformation on the demand keywords and the context information to obtain the target query condition;

[0074] Analyze according to the target query condition to obtain the user's needs and the description of the operations research model as the model requirement information.

[0075] In this embodiment, it involves semantic analysis of the text input by the user to extract key information. Semantic understanding can be accomplished through natural language processing technologies (such as deep learning models, BERT, etc.). The system will analyze the vocabulary, sentence structure, and context in the text to identify the core requirements related to the operations research model, such as demand keywords like "optimize transportation routes" and "maximize production capacity". In addition, the system will also consider context information. For example, the requirements previously mentioned by the user or the implicit requirements inferred by the system. During the semantic understanding process, the system will not only identify single keywords but also combine the context to speculate on the user's true needs. For example, the term "transportation route" may involve multiple aspects, such as route optimization, cost control, time optimization, etc. The system can identify the most important requirements through the context to avoid misunderstandings. After identifying the demand keywords and context information, the system will convert this information into structured query conditions. Structure conversion can be achieved through techniques such as information extraction and relationship mapping to convert the unstructured data in the text into a format that can be processed. For example, converting the requirement of "optimize transportation routes" into an objective function (such as minimizing transportation costs) and constraint conditions (such as road capacity, time limit), etc. Structure conversion is not limited to a single objective function and constraint conditions. It can further extract relevant parameters and variables by analyzing the specific content of the requirements. For example, the system can automatically infer the specific fields involved in the requirements (such as logistics, production scheduling, etc.) and map them to the corresponding operations research models according to industry knowledge. Through this step, the system can process complex demand texts and generate structured information suitable for querying. After converting the demand keywords and context information into target query conditions, the system will analyze these conditions to determine the user's specific needs and select the most suitable model from the preset operations research model library according to these needs. Specifically, the system will match information such as the description, applicable scenarios, and optimization objectives of the model according to the target query conditions to ensure that the recommended model best meets the user's needs. This analysis step not only includes static matching of the query conditions but may also include a mechanism for dynamic adjustment. For example, based on real-time changes in the user input (such as parameter adjustment, modification of constraint conditions, etc.), the system will dynamically update the query conditions, re-analyze, and recommend the most suitable model. At the same time, the system can use multi-level analysis (such as classification, clustering, etc.) to comprehensively consider multiple recommendation factors to improve the accuracy and diversity of the recommendation.

[0076] Through semantic understanding and structure conversion, this application can more accurately extract the user's true needs, avoiding recommendation biases caused by vague or unclear expressions; by extracting demand keywords and context information, it further optimizes the parsing of user needs, enabling the finally recommended model to better fit the user's specific needs and improving the accuracy and applicability of the recommendation.

[0077] In some alternative implementation manners of this embodiment, the step of converting the model requirement information into a corresponding target high-dimensional vector specifically includes:

[0078] Input the user requirement and the operations research model description into a preset Embedding model;

[0079] Output the target high-dimensional vector through the Embedding model.

[0080] In this embodiment, it involves inputting the user requirement information obtained through semantic understanding and structural transformation, as well as the matching operations research model description, into a preset Embedding model. The Embedding model is generally trained through deep learning methods (such as Word2Vec, BERT, GPT, etc.). It can convert the input text data (such as user requirements, model descriptions) into low-dimensional or high-dimensional vector representations. These vectors can usually capture the semantic information of the text. For example, a user may input "Optimize the production plan and minimize the production cost", and this requirement text will be converted into a high-dimensional vector, representing the position of this requirement in the semantic space. The description of the model (such as "linear programming model" or "network flow model") will also be converted into a corresponding vector, representing the characteristics and application fields of the model. Different Embedding methods can also be used, and appropriate pre-trained models (such as BERT, GPT, etc.) can be selected for fine-tuning in specific fields to improve the accuracy of semantic understanding. For example, an Embedding model specifically trained in the field of operations research can be used, enabling the model to more accurately understand the terms and concepts in the field of operations research. The user requirement information and the operations research model description input into the Embedding model will be mapped into target high-dimensional vectors through the model. The high-dimensional vectors are the outputs of the Embedding model. They contain the semantic information of the input text and can represent complex relationships in the high-dimensional space. For example, the text of the input requirement "Minimize the transportation cost" will generate a high-dimensional vector after passing through the Embedding model, representing the semantic features of this requirement, covering information such as the optimization objective and constraints. For example, for the "network flow model", it will generate a vector representing the characteristics of this model, which may include information such as the applicable scenarios, optimization objectives, and constraint forms of the model. These target high-dimensional vectors are not just numbers. They reflect the semantic features of the input text and the relationships between various elements. For example, the distance between two vectors (such as cosine similarity) can be used to measure the matching degree between the requirement and the recommendation model. Further, the system can also sort or optimize the recommendation results according to the similarity between these vectors.

[0081] By converting the model requirement information into high-dimensional vectors and using the Embedding model, this application can effectively vectorize the user requirements and model descriptions. The representation of high-dimensional vectors can better capture and express the semantic features in the text information, improve the calculation and matching efficiency, reduce the semantic ambiguity between models, and ensure the accuracy and relevance of the recommendation results.

[0082] In some alternative implementation manners of this embodiment, the step of retrieving in the preset vector database according to the target high-dimensional vector to determine the recommended operations research model specifically includes:

[0083] Determine the similarity between the target high-dimensional vector and each preset semantic vector in the vector database according to a preset similarity algorithm;

[0084] Determine the target vectors in each of the preset semantic vectors whose similarity is greater than a preset threshold;

[0085] Use the operations research model corresponding to the target vector as the recommended operations research model.

[0086] In this embodiment, the system uses a preset similarity algorithm to calculate the similarity between the target high-dimensional vector and all the stored semantic vectors in the vector database. The similarity algorithm usually measures the similarity between two vectors by calculating the distance or similarity score between them. Common similarity algorithms include cosine similarity, Euclidean distance, Manhattan distance, and dot product. The purpose of this step is to find the semantic vectors similar to the target vector (representing the user's needs) in the preset vector database, so as to help the system find a suitable operations research model. After calculating the similarity, the system will screen out those semantic vectors whose similarity to the target high-dimensional vector is higher than the threshold according to the preset similarity threshold. This threshold is an adjustable parameter, representing the accuracy and tolerance of the system when recommending models. The setting of the threshold affects the accuracy and diversity of the recommendation system. If the threshold is too high, it may lead to too narrow a recommendation result, ignoring potential relevant models; if the threshold is too low, it may lead to too broad a recommendation result, including some irrelevant models. Suppose the target high-dimensional vector represents "an optimization model for minimizing transportation costs". The system will find all the semantic vectors in the database that are similar to this requirement according to the similarity calculation. If the similarity of these vectors is higher than the preset threshold (such as 0.8), they are considered suitable recommendation candidates. Once the target vectors with similarity greater than the threshold are determined, the system will use the operations research models corresponding to these vectors as the recommendation results. This means that the model recommendation is not only based on the similarity calculation, but also includes the operations research models already stored in the historical database, and these models may better meet the user's needs. The recommended operations research models can be linear programming models, integer programming models, dynamic programming models, network flow models, etc., depending on the types of models stored in the database. This step can also be further customized according to the user's personalized needs. For example, if the user has selected a certain specific type of model multiple times in the past, the system can increase the recommendation probability of this type of model to improve the accuracy of personalized recommendation.

[0087] This application further improves the relevance and accuracy of model recommendation by retrieving in the vector database and using the similarity algorithm. By setting the similarity threshold to screen the models that meet the user's needs, it not only improves the recommendation efficiency, but also ensures a high degree of matching between the recommendation result and the user's needs, which helps to improve the user experience.

[0088] In some alternative implementation manners of this embodiment, after the step of displaying the recommended operations research model on the model recommendation interface, the following is further included:

[0089] When the number of the recommended operations research models is two or more, generate a model recommendation list on the model recommendation interface according to the recommended operations research models;

[0090] In response to a selection instruction triggered by the model recommendation interface, determine the target operations research model corresponding to the target user in the model recommendation list.

[0091] In this embodiment, if the system recommends two or more operations research models based on the user's input, the system will generate a list containing all the recommended models on the model recommendation interface. This recommendation list is presented to the user so that the user can conveniently browse all the models that meet the requirements. During the recommendation process, the system needs to ensure that the interface is simple and the amount of information is appropriate. When generating the recommendation list, the brief information of each recommended model (such as model name, description, application scenario, etc.) can be displayed to help the user quickly understand the applicability of each recommended model. The sorting order of the recommendation list is usually based on the similarity between the model and the user's needs. The user can select the model that best meets their needs from the list. This way enables the user to make a choice among multiple recommendation results, thereby improving the personalized recommendation experience. The user can interact with the model list on the recommendation interface, such as selecting a recommended model by clicking, choosing, etc. After the system responds to the user's selection instruction, it determines the target operations research model selected by the user. The selection instruction can be triggered in various ways, such as: click selection, input box selection, sliding / paging. After the user selects a certain recommended model, the system will confirm the selection and return the selection result, and may pop up detailed information or guide the user to enter a more in-depth operation interface. By generating the recommendation list, the user can quickly compare multiple operations research models and select the model that best meets their needs. After the user selects the target model, the system can further provide detailed information about the model, usage guides, example applications, and even directly guide the user to deploy or execute the model.

[0092] This application increases the flexibility of the system by allowing the user to select the most suitable operations research model from multiple recommendation results. The user can make a choice according to their actual needs, thereby enhancing the user's sense of control and usage experience of the system.

[0093] In some alternative implementation manners of this embodiment, after the step of displaying the recommended operations research model on the model recommendation interface, the following is further included:

[0094] When the model recommendation interface receives the associated text information corresponding to the recommended operations research model, update the model requirement information according to the associated text information, and return to execute the step of converting the model requirement information into the corresponding target high-dimensional vector.

[0095] In this embodiment, in the model recommendation interface, the user may further provide text information related to the recommendation model. Such associated text information may be the user's feedback on the recommendation model, supplementary explanations, or descriptions of the user's further requirements for the model. For example, the user may enter information such as "This model is not suitable for my current scale" or "I need this model to be able to process more data" in the recommendation interface. Associated text information refers to the input, comments, or corrections made by the user based on the recommendation model. It can be feedback on the current recommendation result or a supplement to the requirements of the recommendation model. Associated text information usually includes specific modification suggestions, limiting conditions, business scenarios, or other detailed requirements. This interaction design enables the recommendation system to not only be a one-way recommendation process but also have the ability to receive user feedback and optimize the recommendation result in real time. Once the system receives the associated text information provided by the user, it analyzes this information and updates the original model requirement information. The model requirement information may be adjusted due to the user's new requirements or feedback, and the updated requirement information will more accurately reflect the user's actual needs. The updated requirement information may include requirements for aspects such as the model scale, performance, data processing ability, and operation speed. These requirements further refine the user's requirement description, enabling the system to better perform the next round of model recommendation. The updated model requirement information is converted into a new target high-dimensional vector and re-input into the system. This process ensures that the user's feedback or new requirements are effectively integrated and affect the model recommendation result. Through this step, the system can regenerate a recommendation model that better meets the user's needs based on the latest requirement information. After converting the updated requirement information into a high-dimensional vector, the system uses this new vector for model matching. This high-dimensional vector will contain all the updated requirement information and is used to improve the accuracy of the subsequent recommendation model. This process enables the recommendation system to have the ability of dynamic update, being able to continuously adjust the recommendation result according to the user's real-time feedback and changing requirements, rather than relying only on the static data input initially. Through this update mechanism, the user not only receives the recommendation result but also participates in the recommendation process and influences the result. This interactivity greatly improves the intelligence and personalization of the system, allowing the user to precisely control the recommendation result they want.

[0096] This application enhances the real-time dynamic update ability of the recommendation result. When the user further enters associated text information related to the recommendation model, it can promptly update the model requirement information and re-match, ensuring that the recommendation result is optimized as the user's needs change. This dynamic adjustment enables the recommendation system to continuously optimize and accurately reflect the user's needs, improving the adaptability and intelligence level of the system.

[0097] In some alternative implementation manners of this embodiment, after the step of presenting the recommended operations research model on the model recommendation interface, the following steps are further included:

[0098] Analyze the recommended operations research model to determine the model principle, application scenario, and advantages and disadvantages information corresponding to the recommended operations research model;

[0099] Display the model principle, the application scenario, and the advantages and disadvantages information in the model recommendation interface.

[0100] In this embodiment, after the recommendation system displays the recommended operations research models to the user, the system will further analyze these recommended models, extract their core principles, applicable business scenarios, and possible advantages and disadvantages. The purpose of this step is to enable the user to not only see the models themselves, but also clearly understand the working mechanisms, applicable scopes, and potential limitations of each model. The model principle includes the basic theoretical framework of the model, key algorithms, or mathematical foundations. For example, if the recommended is a linear programming model, the system will display the optimization principle, algorithm steps, etc. of the model. Each operations research model has applicable scenarios and conditions. The system will analyze and display the application scope of the model in real business according to the model characteristics. For example, some models may be applicable to production scheduling, logistics optimization, financial planning, etc. The advantages and limitations of the model can also be displayed to help users make better trade-offs when choosing a model. For example, a certain model may have an advantage in calculation accuracy, but may be relatively slow when dealing with large-scale data. Display the parsed model principle, application scenario, and advantages and disadvantages information on the model recommendation interface. These information helps users make more informed choices, enables users to comprehensively understand the characteristics of each recommended model, and make the most appropriate decisions. The display of this information can be in various ways such as text, diagrams, video explanations, etc. To improve the user experience, the display content needs to be intuitive and easy to understand, avoiding overly professional terms from affecting user understanding. The interface can be divided into multiple modules, such as "Model Principle", "Application Scenario", "Advantages and Disadvantages", etc. For example, display the working principle of the model in the form of information charts, brief text descriptions, or demonstration videos to help non-professional users quickly understand.

[0101] Through the analysis of the recommended operations research model and the display of its principle, application scenario, and advantages and disadvantages information in this application, users can comprehensively understand the characteristics of the recommended model, which not only helps users better understand the applicable scope of the model, but also enables users to make more informed decisions when choosing a model, reduces users' doubts about the applicability and effectiveness of the model, and improves the transparency and user trust of the model recommendation system.

[0102] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Among them, Artificial Intelligence (AI) is a theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use knowledge to obtain the best results.

[0103] Artificial intelligence basic technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, and mechatronics. Artificial intelligence software technologies mainly include several major directions such as computer vision technology, robotics, biometric technology, speech processing technology, natural language processing technology, and machine learning / deep learning.

[0104] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through computer-readable instructions. The computer-readable instructions can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above various methods. Among them, the aforementioned storage medium can be a non-volatile storage medium such as a magnetic disk, optical disk, Read-Only Memory (ROM), or a Random Access Memory (RAM), etc.

[0105] It should be understood that although the steps in the flowchart of the accompanying drawings are shown in sequence according to the indication of the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless there is a clear indication in this article, the execution of these steps has no strict order limitation, and they can be executed in other orders. Moreover, at least a part of the steps in the flowchart of the accompanying drawings may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same moment, but can be executed at different moments, and their execution order is not necessarily sequential, but can be executed alternately or in turn with at least a part of other steps or sub-steps or stages of other steps.

[0106] Further reference Figure 3 As an implementation of the method shown above Figure 2 The present application provides an embodiment of an operations research model recommendation system based on a large language model. This system embodiment corresponds to the method embodiment shown in Figure 2 and can be specifically applied to various electronic devices.

[0107] Such as Figure 3As shown in the figure, the operational research model recommendation system 300 based on the large language model described in this embodiment includes: an acquisition module 301, an identification module 302, a conversion module 303, a retrieval module 304, and a display module 305. Among them:

[0108] The acquisition module 301 is used to acquire the text information input by the target user on the model recommendation interface;

[0109] The identification module 302 is used to identify the text information and determine the model requirement information corresponding to the target user;

[0110] The conversion module 303 is used to convert the model requirement information into a corresponding target high-dimensional vector;

[0111] The retrieval module 304 is used to retrieve in a preset vector database according to the target high-dimensional vector to determine the recommended operational research model;

[0112] The display module 305 is used to display the recommended operational research model on the model recommendation interface.

[0113] The operational research model recommendation system based on the large language model provided by this application can accurately understand the user's needs by acquiring the text information input by the user on the model recommendation interface and identifying and analyzing it, automatically convert the needs into high-dimensional vectors and perform matching. Through the deep understanding of the needs by the large language model, it can better match the user's needs with the corresponding operational research models, avoiding the problems of manual intervention or large errors in traditional recommendation methods, thereby improving the result accuracy.

[0114] In some optional implementation manners of this embodiment, the identification module 302 is further used to:

[0115] Perform semantic understanding according to the text information to determine the demand keywords and context information corresponding to the text information;

[0116] Perform structural transformation on the demand keywords and the context information to obtain a target query condition;

[0117] Analyze according to the target query condition to obtain the user's needs and the description of the operational research model as the model requirement information.

[0118] The operational research model recommendation system based on the large language model provided by this application can more accurately extract the user's real needs through semantic understanding and structural transformation, avoiding recommendation deviations caused by vague or unclear expressions; by extracting demand keywords and context information, the parsing of the user's needs is further optimized, so that the finally recommended model can better fit the user's specific needs, improving the accuracy and applicability of the recommendation.

[0119] In some alternative implementation manners of this embodiment, the recognition module 302 is further configured to:

[0120] Input the user requirement and the description of the operations research model into a preset Embedding model;

[0121] Output the target high-dimensional vector through the Embedding model.

[0122] The operations research model recommendation system based on the large language model provided by this application can effectively vectorize the user requirement and the model description by converting the model requirement information into a high-dimensional vector and using the Embedding model. The representation method of the high-dimensional vector can better capture and express the semantic features in the text information, improve the calculation and matching efficiency, reduce the semantic ambiguity between models at the same time, and ensure the accuracy and relevance of the recommendation result.

[0123] In some alternative implementation manners of this embodiment, the retrieval module 304 is further configured to:

[0124] Determine the similarity between the target high-dimensional vector and each preset semantic vector in the vector database according to a preset similarity algorithm;

[0125] Determine a target vector with a similarity greater than a preset threshold among each of the preset semantic vectors;

[0126] Use the operations research model corresponding to the target vector as the recommended operations research model.

[0127] The operations research model recommendation system based on the large language model provided by this application further improves the relevance and accuracy of model recommendation by retrieving in the vector database and using the similarity algorithm. By setting the similarity threshold to screen the models that meet the user requirements, it not only improves the recommendation efficiency, but also ensures a high degree of matching between the recommendation result and the user requirements, which helps to improve the user experience.

[0128] In some alternative implementation manners of this embodiment, the display module 305 is further configured to:

[0129] When the number of the recommended operations research models is two or more, generate a model recommendation list according to the recommended operations research models on the model recommendation interface;

[0130] In response to a selection instruction triggered by the model recommendation interface, determine the target operations research model corresponding to the target user in the model recommendation list.

[0131] The operational research model recommendation system based on a large language model provided by this application increases the flexibility of the system by allowing users to select the most suitable operational research model from multiple recommended results. Users can make selections according to their actual needs, thereby enhancing the users' sense of control over the system and the usage experience.

[0132] In some alternative implementation manners of this embodiment, the display module 305 is further configured to:

[0133] When the model recommendation interface receives the associated text information corresponding to the recommended operational research model, update the model requirement information according to the associated text information, and return to execute the step of converting the model requirement information into the corresponding target high-dimensional vector.

[0134] The operational research model recommendation system based on a large language model provided by this application, by increasing the real-time dynamic update ability of the recommended results, when the user further inputs the associated text information related to the recommended model, can timely update the model requirement information and re-match, ensuring that the recommended results are optimized as the user's needs change. This dynamic adjustment enables the recommendation system to continuously optimize and accurately reflect the user's needs, improving the adaptability and intelligence level of the system.

[0135] In some alternative implementation manners of this embodiment, the display module 305 is further configured to:

[0136] Analyze the recommended operational research model to determine the model principle, application scenario, and advantages and disadvantages information corresponding to the recommended operational research model;

[0137] Display the model principle, the application scenario, and the advantages and disadvantages information in the model recommendation interface.

[0138] The operational research model recommendation system based on a large language model provided by this application, by analyzing the recommended operational research model and displaying its principle, application scenario, and advantages and disadvantages information, enables users to comprehensively understand the characteristics of the recommended model, not only helps users better understand the applicable scope of the model, but also enables users to make more informed decisions when selecting the model, reducing the users' doubts about the applicability and effect of the model, and enhancing the transparency and user trust of the model recommendation system.

[0139] To solve the above technical problems, an embodiment of this application also provides a computer device. For details, please refer to Figure 4 , Figure 4 which is the basic structural block diagram of the computer device in this embodiment.

[0140] The computer device 4 includes a memory 41, a processor 42, and a network interface 43 that are communicatively connected to each other via a system bus. It should be noted that only the computer device 4 with components 41-43 is shown in the figure, but it should be understood that it is not required to implement all the shown components, and more or fewer components can be alternatively implemented. Among them, those skilled in the art of the present technology can understand that the computer device here is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to microprocessors, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital signal processors (DSPs), embedded devices, etc.

[0141] The computer device can be a computing device such as a desktop computer, a notebook, a palm computer, and a cloud server. The computer device can perform human-computer interaction with the user through a keyboard, a mouse, a remote control, a touchpad, a voice control device, etc.

[0142] The memory 41 includes at least one type of readable storage medium, and the readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (such as SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 41 can be an internal storage unit of the computer device 4, such as the hard disk or memory of the computer device 4. In other embodiments, the memory 41 can also be an external storage device of the computer device 4, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the computer device 4. Of course, the memory 41 can also include both the internal storage unit and the external storage device of the computer device 4. In this embodiment, the memory 41 is generally used to store the operating system and various application software installed on the computer device 4, such as computer-readable instructions of the operational research model recommendation method based on a large language model. In addition, the memory 41 can also be used to temporarily store various data that have been output or will be output.

[0143] In some embodiments, the processor 42 may be a Central Processing Unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chips. The processor 42 is generally used to control the overall operation of the computer device 4. In this embodiment, the processor 42 is used to run the computer-readable instructions stored in the memory 41 or process data, such as running the computer-readable instructions of the operation research model recommendation method based on the large language model.

[0144] The network interface 43 may include a wireless network interface or a wired network interface, and this network interface 43 is generally used to establish a communication connection between the computer device 4 and other electronic devices.

[0145] The computer device provided in this application can accurately understand the user's needs by obtaining the text information input by the user on the model recommendation interface, identifying and analyzing it, automatically converting the needs into high-dimensional vectors and performing matching, and deeply understanding the needs through the large language model, so as to better match the user's needs with the corresponding operation research model, avoiding the problems of human intervention or large errors in traditional recommendation methods, and thus improving the result accuracy.

[0146] This application also provides another implementation manner, that is, to provide a computer-readable storage medium storing computer-readable instructions, and the computer-readable instructions can be executed by at least one processor to enable the at least one processor to execute the steps of the operation research model recommendation method based on the large language model as described above.

[0147] The computer-readable storage medium provided in this application can accurately understand the user's needs by obtaining the text information input by the user on the model recommendation interface, identifying and analyzing it, automatically converting the needs into high-dimensional vectors and performing matching, and deeply understanding the needs through the large language model, so as to better match the user's needs with the corresponding operation research model, avoiding the problems of human intervention or large errors in traditional recommendation methods, and thus improving the result accuracy.

[0148] Through the description of the above embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus a necessary general hardware platform. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes several instructions for causing a terminal device (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the methods described in various embodiments of the present application.

[0149] Obviously, the above-described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The accompanying drawings show the preferred embodiments of the present application, but do not limit the patent scope of the present application. The present application can be implemented in many different forms. On the contrary, the purpose of providing these embodiments is to make the understanding of the disclosure content of the present application more thorough and comprehensive. Although the present application has been described in detail with reference to the foregoing embodiments, for those skilled in the art, they can still modify the technical solutions recorded in the foregoing specific embodiments, or perform equivalent replacements on some of the technical features. Any equivalent structure directly or indirectly using the content of the specification and drawings of the present application in other related technical fields is equally within the scope of the patent protection of the present application.

Claims

1. An operations research model recommendation method based on a large language model, characterized in that: The steps include: Obtain the text information entered by the target user on the model recommendation interface; Identify the text information to determine the model requirement information corresponding to the target user; Convert the model requirement information into a corresponding target high-dimensional vector; Searching a preset vector database according to the target high-dimensional vector to determine a recommended operations research model; The recommended operations research model is displayed on the model recommendation interface.

2. The operations research model recommendation method based on a large language model according to claim 1, characterized in that: The step of identifying the text information and determining the model requirement information corresponding to the target user specifically includes: Perform semantic understanding based on the text information to determine the demand keywords and context information corresponding to the text information; Performing structural transformation on the demand keywords and the context information to obtain target query conditions; According to the target query conditions, analysis is performed to obtain user requirements and operations research model descriptions as the model requirement information.

3. The operations research model recommendation method based on a large language model according to claim 2, characterized in that: The step of converting the model requirement information into a corresponding target high-dimensional vector specifically includes: Inputting the user requirements and the operations research model description into a preset embedding model; The target high-dimensional vector is outputted through the Embedding model.

4. The operations research model recommendation method based on a large language model according to claim 1, characterized in that: The step of searching a preset vector database according to the target high-dimensional vector to determine a recommended operations research model specifically includes: Determine the similarity between the target high-dimensional vector and each preset semantic vector in the vector database according to a preset similarity algorithm; Determine a target vector whose similarity is greater than a preset threshold value among the preset semantic vectors; The operations research model corresponding to the target vector is used as the recommended operations research model.

5. The operations research model recommendation method based on a large language model according to claim 1, characterized in that: After the step of displaying the recommended operations research model on the model recommendation interface, the method further includes: When the number of the recommended operations research models is two or more, generating a model recommendation list on the model recommendation interface according to the recommended operations research models; In response to a selection instruction triggered by the model recommendation interface, a target operations research model corresponding to the target user is determined in the model recommendation list.

6. The operations research model recommendation method based on a large language model according to claim 1, characterized in that: After the step of displaying the recommended operations research model on the model recommendation interface, the method further includes: When the model recommendation interface receives the associated text information corresponding to the recommended operations research model, the model requirement information is updated according to the associated text information, and the step of converting the model requirement information into a corresponding target high-dimensional vector is returned to be executed.

7. The operations research model recommendation method based on a large language model according to any one of claims 1 to 6, characterized in that: After the step of displaying the recommended operations research model on the model recommendation interface, the method further includes: Analyze the recommended operations research model to determine the model principle, application scenario, advantages and disadvantages information corresponding to the recommended operations research model; The model principle, the application scenario and the advantages and disadvantages information are displayed in the model recommendation interface.

8. An operations research model recommendation system based on a large language model, characterized in that: include: The acquisition module is used to obtain the text information entered by the target user in the model recommendation interface; An identification module, used to identify the text information and determine the model requirement information corresponding to the target user; A conversion module, used to convert the model requirement information into a corresponding target high-dimensional vector; A retrieval module, used to search in a preset vector database according to the target high-dimensional vector to determine a recommended operations research model; A display module is used to display the recommended operations research model on the model recommendation interface.

9. A computer device, characterized in that: The invention comprises a memory and a processor, wherein the memory stores computer-readable instructions, and when the processor executes the computer-readable instructions, the steps of the operations research model recommendation method based on a large language model as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-readable instructions, which, when executed by a processor, implement the steps of the operations research model recommendation method based on a large language model as described in any one of claims 1 to 7.

Images