Computing device configuration recommendation method and electronic device

Abstract

The invention relates to the technical field of computers, and particularly provides a computing equipment configuration recommendation method and electronic equipment. According to the method, during configuration recommendation of the computing device, the database constructed according to the configuration rules of the computing device is automatically retrieved to obtain the multiple component recommendation schemes of the to-be-recommended component, automatic recommendation of the to-be-recommended component is achieved, and compared with the prior art that the recommendation schemes of the to-be-recommended component are manually obtained based on experience, the recommendation efficiency is improved. Recommendation efficiency is improved and errors are not easy to occur. Besides, when compatibility checking is carried out, the scheme of the embodiment of the invention is realized by adopting the hardware rule knowledge graph constructed based on the compatibility rule, so that compared with the prior art that a professional programmer needs to program the compatibility rule into a rule engine and the compatibility checking is realized by virtue of the rule engine, the compatibility checking is realized by adopting the hardware rule knowledge graph constructed based on the compatibility rule; the construction of the hardware rule knowledge graph in the embodiment has low professional requirements on programmers, and the knowledge graph is easy to obtain.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a method for recommending configuration of a computing device and an electronic device. Background Art

[0002] Server configuration recommendation is the process of providing users with appropriate server hardware configuration recommendations based on specific application needs, business goals, performance requirements, and budget constraints. This process aims to ensure that the server can operate efficiently and stably while meeting business needs and future expansion potential.

[0003] Related technologies rely on manual experience and server configuration manuals, which have the problems of low efficiency and prone to errors. Summary of the Invention

[0004] In view of the above problems, this application is proposed. This application provides a computing device configuration recommendation method and an electronic device.

[0005] According to a first aspect of an embodiment of the present application, a method for recommending a computing device configuration is provided, comprising:

[0006] Searching a database based on the requirement parameters of the component to be recommended to obtain a plurality of component recommendation solutions for the component to be recommended, wherein the database is constructed according to a computing device configuration rule;

[0007] Based on the demand parameters and a pre-trained large language model, selecting a first component recommendation solution from the multiple component recommendation solutions, where a matching degree between the first component recommendation solution and the demand parameters meets a preset matching degree requirement;

[0008] performing a compatibility check on the first component recommendation solution based on a hardware rule knowledge graph to obtain a compatibility check result, wherein the hardware rule knowledge graph is constructed based on compatibility rules extracted from the computing device configuration rules;

[0009] Based on the compatibility check result and the first component recommendation solution, a final recommendation solution for the component to be recommended is determined.

[0010] In the embodiment of the present application, when making a computing device configuration recommendation, multiple component recommendation schemes for the components to be recommended are obtained by automatically searching a database constructed according to the computing device configuration rules, thereby achieving automatic recommendation of the components to be recommended. Compared with the manual experience-based recommendation schemes for the components to be recommended in the related art, the recommendation efficiency is improved and it is less prone to errors. In addition, when performing a compatibility check, the solution of the embodiment of the present application is implemented using a hardware rule knowledge graph constructed based on compatibility rules. Compared with the related art that requires professional programmers to program the compatibility rules into the rule engine and use the rule engine to implement compatibility checks, the construction of the hardware rule knowledge graph in this embodiment does not require high professionalism of programmers and is easy to obtain.

[0011] In some embodiments, the computing device configuration recommendation method according to the first aspect of the embodiments of the present application, based on the requirement parameters of the to-be-recommended component input by the user, before searching the database to obtain multiple component recommendation solutions for the to-be-recommended component, further includes:

[0012] Extracting a summary of each piece of information in the computing device configuration rule to obtain an information summary of each piece of information;

[0013] The information summary is stored to obtain the database.

[0014] In an embodiment of the present application, a database is generated by performing summary extraction on the computing device configuration rules. The data in the database obtained in this way can not only represent the computing device configuration rules, but also the data after summary extraction is smaller in amount than the computing device configuration rules before extraction, which is conducive to improving data retrieval efficiency.

[0015] In some embodiments, according to the computing device configuration recommendation method of the first aspect of the embodiments of the present application, based on the requirement parameters of the component to be recommended, searching a database to obtain multiple component recommendation solutions for the component to be recommended includes:

[0016] Retrieving from the database a first information digest whose character coincidence with the requirement parameter satisfies a preset coincidence requirement; and retrieving from the database a second information digest whose semantic similarity with the requirement parameter satisfies a preset similarity requirement;

[0017] The component recommendation scheme formed by the first information digest and the component recommendation scheme formed by the second information digest are used as the multiple component recommendation schemes.

[0018] In the embodiment of the present application, a dual-path recall strategy is adopted to obtain multiple component recommendation schemes for the component to be recommended, which helps to improve the accuracy of the component recommendation schemes obtained through retrieval.

[0019] In some embodiments, the computing device configuration recommendation method according to the first aspect of the embodiments of the present application performs a compatibility check on the first component recommendation solution based on the hardware rule knowledge graph, and before obtaining the compatibility check result, further includes:

[0020] Using a large model extractor to convert the data format of the compatibility rule into a preset structured format;

[0021] performing entity naming and relationship extraction on the compatibility rules in the structured format using the large model extractor;

[0022] generating a plurality of triples based on the entity nodes obtained by the entity naming and the node relationships obtained by the relationship extraction;

[0023] The multiple triples are combined to generate the hardware rule knowledge graph.

[0024] In the embodiment of the present application, a hardware rule knowledge graph is generated by entity naming and relationship extraction. The method is simple and easy to promote.

[0025] In some embodiments, according to the computing device configuration recommendation method of the first aspect of the embodiments of the present application, a compatibility check is performed on the first component recommendation solution based on the hardware rule knowledge graph to obtain a compatibility check result, including:

[0026] Obtaining configuration constraints of the components to be recommended from the hardware rule knowledge graph;

[0027] A compatibility check is performed on the first component recommendation solution using the configuration constraint to obtain the compatibility check result.

[0028] In an embodiment of the present application, configuration constraints of the components to be recommended are extracted from the hardware rule knowledge graph to perform targeted compatibility checks on the first component recommendation solution, which helps to improve the efficiency of compatibility checks.

[0029] In some embodiments, according to the computing device configuration recommendation method of the first aspect of the embodiments of the present application, determining a final recommendation scheme for the to-be-recommended component based on the compatibility check result and the first component recommendation scheme includes:

[0030] If the compatibility check result indicates that the check is passed, using the first component recommendation solution as the final recommendation solution;

[0031] When the compatibility check result indicates that the check failed, based on the requirement parameters, the large language model and the reason for failure included in the compatibility check result, a second component recommendation scheme is screened from the multiple component recommendation schemes, whose matching degree with the requirement parameters meets the preset matching requirements, and the second component recommendation scheme is used as the final recommendation scheme.

[0032] In the embodiment of the present application, a specific solution for determining the final recommended solution is provided based on the compatibility check result of passing or failing the check, which helps to improve the accuracy of the final recommended solution obtained.

[0033] In some embodiments, according to the computing device configuration recommendation method of the first aspect of the embodiments of the present application, when the component to be recommended is a disk array card, the requirement parameters of the component to be recommended include the requirement parameters of the disk array card input by the user and the final recommendation solution for the hard disk;

[0034] In the case where the component to be recommended is any one of the CPU, memory or the hard disk, the requirement parameters of the component to be recommended include the requirement parameters of any one of the components input by the user and the final recommended solution of the chassis;

[0035] In a case where the component to be recommended is the chassis, the requirement parameters of the component to be recommended include the requirement parameters of the chassis input by the user.

[0036] In the embodiment of the present application, specific forms of requirement parameters are given for different components to be recommended, which helps to improve the accuracy of component recommendation solutions obtained by searching the database using the requirement parameters.

[0037] In some embodiments, according to the computing device configuration recommendation method of the first aspect of the embodiments of the present application, when the final recommendation scheme of the disk array card indicates that the type of the disk array card is a standard card, the component to be recommended also includes an expansion card, and the requirement parameters of the expansion card include the final recommendation scheme of the disk array card, the final recommendation scheme of the hard disk, and the requirement parameters of the expansion card input by the user.

[0038] In the embodiment of the present application, an expansion card is also recommended according to the type of disk array card so that the standard disk array card can be accurately combined with the computing device, which helps to improve the connection between the disk array card and the computing device.

[0039] In some embodiments, the computing device configuration recommendation method according to the first aspect of the embodiments of the present application, after determining the final recommendation scheme for the component to be recommended based on the compatibility check result and the first component recommendation scheme, further includes:

[0040] Acquire multiple computing device configuration solutions based on the final recommendation solution of the component to be recommended, each of the computing device configuration solutions includes a final recommendation solution for a chassis and final recommendation solutions for other hardware components obtained based on the final recommendation solution for the chassis, the other hardware components including at least one of the CPU, the memory, the hard disk, and the disk array card;

[0041] A large language model based on a prompt project is called to screen an optimal computing device configuration scheme from the multiple computing device configuration schemes, and the prompt words set in the prompt project are used to reflect the user's application requirements for the computing device.

[0042] In an embodiment of the present application, calling a large language model to screen the optimal computing device configuration scheme from multiple computing device configuration schemes helps to improve the adaptability of the optimal computing device configuration scheme to the user's application needs.

[0043] According to a second aspect of an embodiment of the present application, an electronic device is provided, comprising a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method described in the first aspect.

[0044] According to a third aspect of an embodiment of the present application, a computer-readable storage medium is provided, on which a computer program / instruction is stored. When the computer program / instruction is executed by a processor, the steps of the method described in the first aspect are implemented.

[0045] According to a fourth aspect of the embodiments of the present application, a computer program product is provided, comprising a computer program / instruction, which implements the steps of the method described in the first aspect when executed by a processor.

[0046] As will be described in detail below, according to the computing device configuration recommendation method of the embodiment of the present application, when making computing device configuration recommendations, multiple component recommendation schemes for the components to be recommended are obtained by automatically searching the database constructed according to the computing device configuration rules, thereby realizing automatic recommendation of the components to be recommended. Compared with the manual experience-based recommendation schemes for the components to be recommended in the related art, the recommendation efficiency is improved and it is less prone to errors. In addition, when performing compatibility checks, the solution of the embodiment of the present application is implemented using a hardware rule knowledge graph constructed based on compatibility rules. Compared with the related art that requires professional programmers to program compatibility rules into the rule engine and use the rule engine to implement compatibility checks, the construction of the hardware rule knowledge graph in this embodiment does not require high professionalism from programmers and is easy to obtain.

[0047] It is to be understood that both the foregoing general description and the following detailed description are exemplary, and are intended to provide further explanation of the technology as claimed. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] The above and other purposes, features, and advantages of the embodiments of the present application will become more apparent through a more detailed description of the embodiments of the present application in conjunction with the accompanying drawings. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the embodiments of the present application and do not constitute a limitation of the embodiments of the present application. In the drawings, the same reference numerals generally represent the same components or steps.

[0049] Figure 1 3 is a flowchart illustrating the application of a computing device configuration recommendation method according to an embodiment of the present application.

[0050] Figure 2 2 is a schematic diagram illustrating a recommended solution for recalling chassis components using a dual-path recall strategy according to an embodiment of the present application.

[0051] Figure 3 It is a schematic diagram illustrating the application of constructing a hardware rule knowledge graph according to an embodiment of the present application.

[0052] Figure 4 It is a schematic diagram illustrating the application of a large language model based on a prompt engineering guidance according to an embodiment of the present application to recommend an optimal computing device configuration solution.

[0053] Figure 5 is a schematic diagram illustrating the application of a computing device configuration recommendation according to an embodiment of the present application.

[0054] Figure 6 2 is a structural diagram illustrating an apparatus for recommending configurations of a computing device according to an embodiment of the present application.

[0055] Figure 7 FIG. 1 is a hardware block diagram illustrating an electronic device according to an embodiment of the present application.

[0056] Figure 8 is a schematic diagram illustrating a computer program product according to an embodiment of the present application. DETAILED DESCRIPTION

[0057] In order to make the purpose, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the embodiments of the present application for protection, but merely represents selected embodiments of the embodiments of the present application. Based on the embodiments of the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work are within the scope of protection of the embodiments of the present application.

[0058] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, it does not need to be further defined or explained in subsequent drawings.

[0059] The term "and / or" herein simply describes an association relationship, indicating that three relationships can exist. For example, A and / or B can represent the existence of A alone, the simultaneous existence of A and B, and the existence of B alone. In addition, the term "at least one" herein refers to any combination of at least two of any one or more of a plurality of items. For example, "at least one of A, B, and C" can represent any one or more elements selected from the set consisting of A, B, and C.

[0060] As mentioned above, the related art relies on manual experience and computing device configuration manuals to make computing device configuration recommendations, which leads to low efficiency and prone to errors.

[0061] In order to alleviate the technical problems existing in the related art, the embodiment of the present application provides a method for recommending configuration of a computing device and an electronic device. When making a computing device configuration recommendation, the method automatically retrieves a database constructed according to the computing device configuration rules to obtain multiple component recommendation schemes for the components to be recommended, thereby realizing automatic recommendation of the components to be recommended. Compared with the related art in which the recommendation schemes for the components to be recommended are obtained manually based on experience, the recommendation efficiency is improved and it is less prone to errors. In addition, when performing a compatibility check, the solution of the embodiment of the present application is implemented using a hardware rule knowledge graph constructed based on compatibility rules. Compared with the related art in which professional programmers are required to program the compatibility rules into the rule engine and use the rule engine to implement compatibility checks, the construction of the hardware rule knowledge graph in this embodiment does not require high professionalism from programmers and is easy to obtain.

[0062] The following first explains the terms involved in the embodiments of the present application.

[0063] Retrieval-augmented generation: Retrieval-augmented generation (RAG) is a technique that combines information retrieval and natural language generation. Specifically, when generating text, the RAG model first retrieves relevant information from an external knowledge base and then combines this information with the model's own knowledge to generate more accurate and rich text content. This approach is particularly suitable for tasks that require specific domain knowledge.

[0064] Knowledge Graph: A knowledge graph is a graphical data structure used to represent and organize knowledge. It represents real-world entities and the relationships between them through nodes (entities) and edges (relationships). A knowledge graph contains not only the attribute information of entities but also the complex relationships between them, forming a structured knowledge network.

[0065] Recall: Recall refers to the process of extensively retrieving candidate data fragments that are semantically relevant to user queries from the knowledge base through technologies such as multi-dimensional vectorized encoding and hybrid search. Its core goal is to maximize the coverage of potential related information to ensure the completeness of knowledge supply in the generation stage.

[0066] To facilitate understanding of this embodiment, we first provide a detailed description of a computing device configuration recommendation method disclosed in an embodiment of this application. The computing device configuration recommendation method provided in this embodiment is generally executed by an electronic device with certain computing capabilities, such as a terminal device, a computing device, or other processing device. In some possible implementations, this computing device configuration recommendation method can be implemented by a processor invoking computer-readable instructions stored in a memory.

[0067] See also Figure 1 FIG. 1 is a flowchart of a method for recommending a configuration of a computing device according to an embodiment of the present application, wherein the method comprises the following steps:

[0068] Step 101: Based on the requirement parameters of the component to be recommended, a database is searched to obtain multiple component recommendation solutions for the component to be recommended. The database is constructed according to the configuration rules of the computing device.

[0069] In this embodiment, the components to be recommended are hardware components that constitute the computing device. In one example, when the computing device is a server, the hardware components include but are not limited to a chassis, a CPU, a memory, a hard disk, a network interface card, and the like.

[0070] In some embodiments, the computing device configuration rule includes but is not limited to a computing device configuration manual. In one example, when the computing device is a server, the computing device configuration rule may be a server configuration manual.

[0071] In this embodiment, the requirement parameters determine the model, type, and other attributes of the recommended component. The requirement parameters can include two types of data: user-entered parameters based on their own requirements, and system-defined associated requirement parameters. Associated requirement parameters refer to the recommended solutions for other hardware components that are required when recalling the recommended solution for the component.

[0072] It should be understood that for different components to be recommended, the requirement parameters of some components to be recommended include both the parameters entered by the user based on his or her own needs and the associated requirement parameters customized by the system, but the requirement parameters of some components to be recommended only include the parameters entered by the user based on his or her own needs.

[0073] For example, if the computing device is a server, and the recommended component is a chassis, the requirement parameters only need to include the chassis requirement parameters entered by the user. If the recommended component is a CPU, memory, or hard drive, the requirement parameters should include not only the CPU requirement parameters, memory requirement parameters, or hard drive requirement parameters entered by the user, but also the final recommended solution for the chassis. If the recommended component is a disk array card (RAID card), the requirement parameters include the user-entered disk array card requirement parameters. Furthermore, since the disk array card recommendation is closely related to the recommended hard drive, the requirement parameters also include the final recommended solution for the hard drive.

[0074] If the final recommendation for the disk array card indicates that the card is a standard card, the recommended components also include an expansion card. The expansion card requirement parameters include the final recommendation for the disk array card, the final recommendation for the hard drive, and the user-entered expansion card requirement parameters. If the disk array card is a standard card, the card requires an expansion card to be inserted into the mainboard, so in this case, an expansion card is recommended. If the disk array card is a daughter card, the card can be inserted into the mainboard without an expansion card, so in this case, no expansion card recommendation is required to insert the disk array card into the mainboard.

[0075] In this embodiment, the computing device configuration rules include multiple pieces of information, each of which includes hardware description information describing the hardware components and compatibility rules. The hardware description information includes parameters such as the model or type of the hardware component, and the compatibility rules are used to describe the configuration constraints that the hardware components need to meet. Taking the 9100F CPU as an example, the memory type it supports is DDR4 2400. If the memory frequency is higher than the frequency supported by the CPU, the actual memory frequency can only reach the frequency supported by the CPU.

[0076] In one or more embodiments, a database is generated by extracting and storing summary information from computing device configuration rules.

[0077] The application includes but is not limited to extracting a summary of each piece of information using an extractive text summarization method. Among them, the extractive text summarization method can extract key information from the original text to form a summary. This method creates a summary by identifying and selecting the most relevant sentences or fragments in the text. Extractive summarization methods are mainly divided into two categories: keyword-based extraction and sentence-based extraction. Keyword-based extraction methods generate summaries by identifying frequently appearing keywords and phrases in the text. These keywords and phrases represent the main themes and concepts of the text. The process of summarization extraction can efficiently extract key information, reduce the workload of manual processing, ensure the accuracy and completeness of the knowledge graph, and provide a reliable data source for subsequent recommendations and compatibility checks.

[0078] In one or more embodiments, to balance recall accuracy and recall rate, a dual-path recall strategy is employed to search the database and obtain multiple component recommendation solutions for the component to be recommended. Each of the dual-path recall strategies defines a database search method, and the component recommendation solutions retrieved by each search method are combined to obtain multiple component recommendation solutions for the component to be recommended.

[0079] It should be understood that in actual applications, a single-way recall strategy or more three-way or four-way recall strategies may be used to obtain multiple component recommendation solutions for the component to be recommended. A single-way recall strategy refers to searching the database using one search method, while a three-way recall strategy refers to searching the database using three search methods.

[0080] In one or more embodiments, based on obtaining a database constructed based on computing device configuration rules, the following visual steps are provided for obtaining multiple component recommendation solutions for a component to be recommended using a two-way recall strategy:

[0081] Retrieving from the database a first information digest whose character coincidence with the required parameter satisfies a preset coincidence requirement; and retrieving from the database a second information digest whose semantic similarity with the required parameter satisfies a preset similarity requirement;

[0082] The component recommendation scheme formed by the first information digest and the component recommendation scheme formed by the second information digest are used as a plurality of component recommendation schemes.

[0083] In this embodiment, what is stored in the database is the information summary of each piece of information in the computing device configuration rules, that is, the key information of each piece of information. Each information summary is stored in the database in the form of characters and has corresponding semantics. On this basis, the first information summary refers to the information summary in the database whose character overlap with the requirement parameters meets the overlap requirement. The overlap requirement here includes but is not limited to the percentage of the number of identical characters being greater than the percentage threshold. The second information summary refers to the information summary in the database whose semantic similarity with the requirement parameters meets the similarity requirement. The similarity requirement here includes but is not limited to the semantic similarity being greater than the similarity threshold.

[0084] In this embodiment, the BM25 search engine in the retrieval enhancement generation technology is used, but is not limited to, to retrieve a first information summary whose character overlap with the requirement parameters satisfies the overlap requirement, and the FAISS search engine is used to retrieve a second information summary whose semantic similarity with the requirement parameters satisfies the similarity requirement. The retrieval enhancement generation technology is used to recall chassis and component information that matches the user's requirements from the database, and the recalled information is input into the large language model to generate the final recommendation. This process can quickly and accurately recall matching component information, improve the accuracy and intelligence of the recommendations, and ensure that the recommended components are compatible with the chassis and meet the user's requirements.

[0085] Among them, the BM25 search engine uses a ranking algorithm based on probability theory for retrieval. It calculates the relevance score between documents and queries through functions of term frequency (TF) and document frequency (DF). Retrieval using the FAISS search engine is divided into two stages: indexing and querying. In the indexing stage, the FAISS search engine converts all input vectors into specified dimensions and groups them using a clustering algorithm. Each cluster center is regarded as a prototype vector and used to build an index. In the query stage, the FAISS search engine quickly finds the most similar cluster center by calculating the distance between the input vector and all prototype vectors, thereby completing similar vector retrieval.

[0086] It should be noted that the data that the FAISS retriever can process is vectors, so when using the FAISS retriever for retrieval, it can be used in conjunction with an embedding layer model (embedding), which can represent data as vectors.

[0087] As an example, given Figure 2 The diagram shows a recommended solution for recalling chassis parts using a two-way recall strategy.

[0088] like Figure 2 As shown in Figure 3, the dual-path recall strategy includes a BM25 retriever, a FAISS retriever, and an embedding layer model.

[0089] The chassis requirement parameters input by the user serve as query parameters for the BM25 and FAISS search engines. The BM25 search engine performs literal recall of chassis based on the chassis requirement parameters. Specifically, by calculating the degree of character overlap with the chassis requirement parameters, it obtains a first information digest from the database that meets the preset overlap requirement. This first information digest serves as the BM25 search engine's recall result. The embedding layer model is used in conjunction with the FAISS search engine. The embedding layer model represents the chassis requirement parameters, which serve as query parameters, as vectors. The FAISS search engine performs semantic recall of chassis based on the vectorized chassis requirement parameters. Specifically, by calculating the semantic similarity with the chassis requirement parameters, it obtains a second information digest from the database that meets the preset similarity requirement. This second information digest serves as the FAISS search engine's recall result. The BM25 and FAISS search engine's recall results are combined to form the final recall result for the chassis using the dual-path recall strategy.

[0090] Step 102 : Based on the demand parameters and the pre-trained large language model, a first component recommendation solution is selected from multiple component recommendation solutions, and the matching degree between the first component recommendation solution and the demand parameters meets a preset matching degree requirement.

[0091] The large language model in this embodiment can filter component recommendation schemes for different components to be recommended. For example, the large language model can filter multiple component recommendation schemes for a dual-channel recalled CPU, and can also filter multiple component recommendation schemes for a dual-channel recalled memory.

[0092] In this embodiment, the large language model may perform similarity matching between the requirement parameters and multiple component recommendation solutions, and select a first component recommendation solution from the multiple component recommendation solutions based on the similarity matching results. The first component recommendation solution is at least one component recommendation solution from the multiple component recommendation solutions, and the degree of match between the first component recommendation solution and the requirement parameters satisfies a matching requirement, including but not limited to the matching result between the requirement parameters and the first component recommendation solution being greater than a matching threshold.

[0093] Step 103: Based on the hardware rule knowledge graph, perform a compatibility check on the first component recommendation solution to obtain a compatibility check result. The hardware rule knowledge graph is constructed based on the compatibility rules extracted from the computing device configuration rules.

[0094] In one or more embodiments, in order to ensure the feasibility of the configuration solution, after obtaining the recommended solution for the component to be recommended, it is also necessary to perform a comprehensive compatibility check on the first component recommended solution to improve the reliability of the configuration.

[0095] In specific implementation, step 103 may include the following steps:

[0096] Obtain configuration constraints of the recommended components from the hardware rule knowledge graph;

[0097] A compatibility check is performed on the first component recommendation solution using the configuration constraint to obtain a compatibility check result.

[0098] In this embodiment, the compatibility rules include configuration constraints between different hardware components. The compatibility rules are represented by a hardware rule knowledge graph, and the hardware rule knowledge graph is used to perform a compatibility check on the first component recommendation scheme. This can comprehensively cover complex compatibility relationships and special situations, and find potential problems in the configuration scheme as much as possible, avoiding the recommendation of incompatible components, and improving the reliability of the configuration as well as the stability and performance of the computing device.

[0099] In one or more embodiments, the following steps are provided for generating a hardware rule knowledge graph of compatibility rules:

[0100] Use the large model extractor to convert the data format of the compatibility rules into a preset structured format;

[0101] Utilize the large model extractor to perform entity naming and relationship extraction based on the compatibility rules of the structured format;

[0102] Generate multiple triples based on the entity nodes obtained by entity naming and the node relationships obtained by relationship extraction;

[0103] Combine multiple triplets to generate a hardware rule knowledge graph.

[0104] The pre-set structured format includes, but is not limited to, JSON. Taking JSON as an example, after the data format of the compatibility rule is converted to JSON, the keys in the JSON format represent the entities involved in the compatibility rule, i.e., the hardware components that make up the computing device, and the values ​​in the JSON format represent the relationships related to the entities.

[0105] In this embodiment, entity naming refers to identifying specific entities, namely, various hardware components, from the compatibility rules of the structured format, and relationship extraction refers to identifying the relationships between entities from the compatibility rules of the structured format to extract the semantic relationships between entities.

[0106] The hardware rule knowledge graph consists of pieces of knowledge, each piece of knowledge is represented as a triple, and each triple includes two entity nodes and an edge used to represent the relationship between the two entity nodes. Multiple triplets can be combined to generate a hardware rule knowledge graph of compatibility rules.

[0107] As an example, given Figure 3The diagram below shows how to build a hardware rule knowledge graph. The completed hardware rule knowledge graph details the compatibility constraints between hardware components. Nodes in the graph represent different hardware components, while edges represent the compatibility relationships and constraints between them. A large language model is used to extract compatibility rules from the configuration manual and convert them into edge and node attributes in the graph.

[0108] like Figure 3 As shown, the computing device configuration rules are input into the large language model extractor, which converts the data format of the compatibility rules in the computing device configuration rules into a preset structured format. The large language model extractor outputs the compatibility rules in the structured format. The large model extractor is used to perform entity naming and relationship extraction on the compatibility rules in the structured format. Based on the entity nodes obtained by entity naming and the node relationships obtained by relationship extraction, multiple triples are generated. The multiple triples are combined to generate a hardware rule knowledge graph for the compatibility rules. Here, the large model extractor (such as Indexify, an open source import and extraction engine) can be an extractor that can extract and return structured information from unstructured data, for example, obtaining specific information from a PDF in JSON format.

[0109] In this embodiment, based on the different types of components to be recommended, the configuration constraints of the components to be recommended obtained from the hardware rule knowledge graph can be one, multiple, or 0. Using the configuration constraints to perform a compatibility check on the first component recommendation solution refers to using all configuration constraints related to the component to be recommended to perform a compatibility check on the first component recommendation solution. When the obtained configuration constraint is 0, it means that there is no need to perform a compatibility check on the component recommendation solution for the component to be recommended. For example, the chassis is such a component to be recommended, that is, when obtaining the first component recommendation solution for the chassis, there is no need to perform a compatibility check on the first component recommendation solution for the chassis, and a compatibility check result indicating that the check has passed can be directly generated.

[0110] For example, the hardware rule knowledge graph includes configuration constraints between the chassis and CPU, as well as configuration constraints between the CPU and memory. Therefore, when the component to be recommended is the CPU, a compatibility check is performed on the first component recommendation plan. This requires a compatibility check between the CPU described in the first component recommendation plan and the recommended memory and chassis allocation.

[0111] Step 104: Determine a final recommendation for the component to be recommended based on the compatibility check result and the first component recommendation.

[0112] In one or more embodiments, when the compatibility check result indicates that the check is passed, the first component recommendation solution is used as the final recommendation solution for the component to be recommended;

[0113] If the compatibility check result indicates a failure, a second component recommendation is selected from the multiple component recommendations based on the requirement parameters, the large language model, and the failure reason included in the compatibility check result. The second component recommendation is selected as the final recommendation. Alternatively, the first component recommendation is manually revised, and the revised component recommendation is selected as the final recommendation.

[0114] In this embodiment, for component combinations that do not meet the compatibility check, the system automatically recalls suitable components and repeats the above steps until all components meet the configuration requirements. This process dynamically optimizes the recommended results, ensuring that the final configuration meets all compatibility requirements, and improving the accuracy and reliability of the configuration.

[0115] The following describes the process of determining the final recommended solution based on the compatibility check results, using the computer chassis, CPU, memory, and hard drive as an example of the hardware components of the computing device that needs to be recalled.

[0116] Check the constraints between chassis and CPU:

[0117] Use the hardware rule knowledge graph to obtain the configuration constraints between the chassis and CPU. Check whether the recommended chassis and CPU meet the configuration guidelines. If not, re-recommend the CPU; if so, proceed to the overall compatibility check.

[0118] Chassis and memory constraint check:

[0119] Using the hardware rule knowledge graph, we obtain the configuration constraints between the chassis and memory. We then check whether the recommended chassis and memory comply with the configuration guidelines. If not, we re-recommend the memory; if not, we proceed to the overall compatibility check.

[0120] Check the constraints between the chassis, hard disk, and PCIE (Peripheral Component Interconnect Express, a high-speed serial bus interface):

[0121] Using the hardware rule knowledge graph, we identify configuration constraints between the chassis, hard drives, and PCIE cards. We then check whether the recommended chassis, hard drives, and PCIE cards comply with these configuration guidelines. If any of these recommendations do not meet these requirements, we re-recommend hard drives, then recommend RAID cards, and finally PCIE cards. If they do meet these requirements, we proceed to a comprehensive compatibility check.

[0122] It's important to note that compatibility checks not only include checking compatibility between hardware components—such as the chassis and CPU, the chassis and memory, and the chassis and hard drive—but also include compatibility checks between multiple components, such as the overall compatibility between the chassis, CPU, memory, hard drive, and PCIE card. The system ensures that all components are compatible not only with the chassis but also with each other.

[0123] In this embodiment, as previously described, taking the hardware components of a computing device as a chassis, CPU, memory, hard disk, and disk array card as an example, the final recommendation results for the CPU, memory, and hard disk depend on the final recommendation result for the chassis, and the final recommendation result for the disk array card depends on the final recommendation result for the hard disk. Since the final recommendation result for the hard disk is related to the final recommendation result for the chassis, the final recommendation result for the disk array card is also related to the final recommendation result for the chassis. Therefore, when the final recommendation result for the chassis changes, the final recommendation results for the CPU, memory, hard disk, and disk array card will also change. In this way, multiple computing device configuration plans can be obtained. By calling a large language model based on prompt engineering, the large model can be guided to recommend the optimal configuration plan, understand complex configuration requirements, and take into account budget and scalability as well as specific application scenarios, thereby improving the intelligence level of the recommendation.

[0124] In a specific implementation, in an optional embodiment, after step 104, the method may further include the following steps:

[0125] Obtaining multiple computing device configuration solutions based on the final recommended solutions for the components to be recommended, each computing device configuration solution including a final recommended solution for a chassis and final recommended solutions for other hardware components based on the final recommended solution for the chassis, the other hardware components including at least one of a CPU, memory, hard disk, and disk array card;

[0126] A large language model based on the prompt project is called to screen the optimal computing device configuration scheme from multiple computing device configuration schemes. The prompt words set in the prompt project are used to reflect the user's application needs for the computing device.

[0127] In some embodiments, after obtaining the optimal computing device configuration solution, a detailed configuration report can be generated. The configuration report can include information such as recommended component solutions, compatibility check results, performance evaluation, cost budget, etc. A user feedback mechanism is also provided to collect user evaluations and suggestions on the configuration solution to further optimize the configuration recommendation algorithm.

[0128] This embodiment uses specific prompts to guide the large model in generating optimized configuration solutions. This report then generates a detailed configuration report, including recommended component solutions, compatibility check results, performance evaluation, cost estimates, and other information. It also provides a user feedback mechanism. This process ensures that the configuration is neither overly redundant nor scalable, while also taking budgets and specific application scenarios into consideration. This improves the intelligence of the recommendations and enhances user satisfaction.

[0129] In this embodiment, the prompt words set in the prompt project include but are not limited to computing device application scenarios, computing device budget requirements and / or computing device scalability.

[0130] As an example, given Figure 4 The diagram shown is a diagram of a large language model based on a prompt engineering to recommend the optimal computing device configuration solution. Figure 4 As shown, three computing device configuration plans were obtained based on the final recommendation plan of the components to be recommended. The large language model based on the prompt words of the computing device application scenario, computing device budget requirements, and computing device scalability was called to select the second computing device configuration plan from the three computing device configuration plans as the optimal computing device configuration plan.

[0131] In the solution provided by this embodiment, when making a computing device configuration recommendation, multiple component recommendation schemes for the components to be recommended are obtained by automatically searching a database constructed according to computing device configuration rules, thereby achieving automatic recommendation of the components to be recommended. Compared with the manual experience-based recommendation schemes for the components to be recommended in the related art, the recommendation efficiency is improved and it is less prone to errors. In addition, when performing a compatibility check, the solution of the embodiment of the present application is implemented using a hardware rule knowledge graph constructed based on compatibility rules. Compared with the related art that requires professional programmers to program compatibility rules into the rule engine and use the rule engine to implement compatibility checks, the construction of the hardware rule knowledge graph in this embodiment does not require high professionalism from programmers and is easy to obtain.

[0132] The following introduces the configuration recommendation method for a computing device in an embodiment of the present application by taking hardware components including a chassis, CPU, memory, hard disk, disk array card and expansion card as an example.

[0133] like Figure 5 As shown, first, a chassis recommendation is performed. Specifically, the user inputs the required parameters of the chassis. Through a two-way recall strategy, a database obtained from the computing device configuration manual is retrieved based on the required parameters of the chassis to obtain multiple component recommendation schemes for the chassis. The pre-trained large language model is called to obtain the final component recommendation scheme for the chassis from the multiple component recommendation schemes.

[0134] Then, when making recommendations for each component in the CPU, memory, and hard disk, the user inputs the required parameters of each component and the final recommendation results of the chassis, and follows the two-way recall strategy to search the database to obtain multiple component recommendation schemes for each component. The pre-trained large language model is called to obtain the final component recommendation scheme for each component from the multiple component recommendation schemes.

[0135] When recommending disk array cards, the user inputs the required parameters of the disk array card and the final recommendation results of the hard disk, and according to the two-way recall strategy, the database is searched to obtain multiple component recommendation schemes for the disk array card. The pre-trained large language model is called to obtain the final component recommendation scheme for the disk array card from the multiple component recommendation schemes.

[0136] When the final component recommendation for a disk array card indicates that the card type is a standard card, expansion cards are also recommended. This is also achieved using a two-way recall strategy and a pre-trained large language model.

[0137] Then, a compatibility check is performed on the final recommended solution for each component to be recommended. If the compatibility check passes, the final recommended solution is output. Otherwise, the requirement parameters are updated based on the reason for the compatibility check failure, and the recommended components are recommended again.

[0138] As another example, taking the computing device as a server as an example, the following embodiment is given to introduce the computing device configuration recommendation method of the embodiment of the present application.

[0139] User demand input:

[0140] The user enters their requirements, including parameters such as the desired CPU, memory, and hard drive type. Using RAG and the large language model, a list of chassis matching the user's requirements is retrieved, such as models with 8 x 2.5-inch hard drives and models with 12 x 3.5-inch hard drives with direct access. This information is then fed into the large language model, which generates the final recommendation: the 8 x 2.5-inch hard drive model.

[0141] Initial recall of matching parts:

[0142] Use RAG to initially recall CPUs, memory, and hard drives that match the 8x2.5-inch hard drive model from the server configuration manual. Input the recalled component information and user requirements into a large language model, allowing the model to recommend the appropriate CPU, memory, hard drive, disk array card, and expansion cards.

[0143] Constraint Checking:

[0144] By writing a knowledge graph query statement, each recommended component is checked against the chassis for constraints. If the constraints are not met, a new recommendation is made. For components that meet the constraints, a comprehensive compatibility check is performed, also using a knowledge graph query statement. For example:

[0145] MATCH(server:`HARD_DRIVE_AND_IO_MODULE_SUPPORT_SPEC`{name:

[0146] '8*2.5-inch hard drive model')

[0147] -[:CPU number]->(cpu:`HARD_DRIVE_AND_IO_MODULE_SUPPORT_SPEC`)

[0148] RETURN cpu

[0149] Final recommendation solution generation:

[0150] Solutions that pass the compatibility check are fed into a large language model, which then selects the most appropriate solution based on the prompt word engineering. The resulting recommendations include the specific models and configurations of the CPU, memory, hard drive, disk array card, and expansion cards.

[0151] The present application also provides a computing device configuration recommendation apparatus for executing the computing device configuration recommendation method provided in any of the above embodiments. Figure 6 As shown, the device includes:

[0152] An acquisition module 61 is configured to retrieve a database based on the requirement parameters of the component to be recommended, and obtain a plurality of component recommendation solutions for the component to be recommended, wherein the database is constructed according to a computing device configuration rule;

[0153] a screening module 62 for screening a first component recommendation from the plurality of component recommendation solutions based on the demand parameters and a pre-trained large language model, wherein a matching degree between the first component recommendation and the demand parameters satisfies a preset matching degree requirement;

[0154] a checking module 63 configured to perform a compatibility check on the first component recommendation solution based on a hardware rule knowledge graph constructed based on compatibility rules extracted from the computing device configuration rules, and obtain a compatibility check result;

[0155] The determination module 64 is configured to determine a final recommendation scheme for the component to be recommended based on the compatibility check result and the first component recommendation scheme.

[0156] In one or more embodiments, the apparatus is further configured to:

[0157] Based on the requirement parameters of the component to be recommended input by the user, before searching the database to obtain multiple component recommendation solutions for the component to be recommended, extracting a summary of each piece of information in the computing device configuration rule to obtain an information summary of each piece of information;

[0158] The information summary is stored to obtain the database.

[0159] In one or more embodiments, the obtaining module 61 is configured to:

[0160] Retrieving from the database a first information digest whose character coincidence with the requirement parameter satisfies a preset coincidence requirement; and retrieving from the database a second information digest whose semantic similarity with the requirement parameter satisfies a preset similarity requirement;

[0161] The component recommendation scheme formed by the first information digest and the component recommendation scheme formed by the second information digest are used as the multiple component recommendation schemes.

[0162] In one or more embodiments, the apparatus is further configured to:

[0163] Based on the hardware rule knowledge graph, a compatibility check is performed on the first component recommendation solution. Before obtaining the compatibility check result, a large model extractor is used to convert the data format of the compatibility rule into a preset structured format;

[0164] performing entity naming and relationship extraction on the compatibility rules in the structured format using the large model extractor;

[0165] generating a plurality of triples based on the entity nodes obtained by the entity naming and the node relationships obtained by the relationship extraction;

[0166] The multiple triples are combined to generate the hardware rule knowledge graph.

[0167] In one or more embodiments, the inspection module 63 is configured to:

[0168] Obtaining configuration constraints of the components to be recommended from the hardware rule knowledge graph;

[0169] A compatibility check is performed on the first component recommendation solution using the configuration constraint to obtain the compatibility check result.

[0170] In one or more embodiments, the inspection module 63 is configured to:

[0171] If the compatibility check result indicates that the check is passed, using the first component recommendation solution as the final recommendation solution;

[0172] When the compatibility check result indicates that the check failed, based on the requirement parameters, the large language model and the reason for failure included in the compatibility check result, a second component recommendation scheme is screened from the multiple component recommendation schemes, whose matching degree with the requirement parameters meets the preset matching requirements, and the second component recommendation scheme is used as the final recommendation scheme.

[0173] In one or more embodiments, when the component to be recommended is a disk array card, the requirement parameters of the component to be recommended include the requirement parameters of the disk array card input by the user and the final recommendation solution of the hard disk;

[0174] In the case where the component to be recommended is any one of the CPU, memory or the hard disk, the requirement parameters of the component to be recommended include the requirement parameters of any one of the components input by the user and the final recommended solution of the chassis;

[0175] In a case where the component to be recommended is the chassis, the requirement parameters of the component to be recommended include the requirement parameters of the chassis input by the user.

[0176] In one or more embodiments, when the final recommendation scheme for the disk array card indicates that the type of the disk array card is a standard card, the component to be recommended also includes an expansion card, and the requirement parameters of the expansion card include the final recommendation scheme for the disk array card, the final recommendation scheme for the hard disk, and the requirement parameters of the expansion card input by the user.

[0177] In one or more embodiments, the apparatus is further configured to:

[0178] After determining a final recommendation for the component to be recommended based on the compatibility check result and the first component recommendation, obtain multiple computing device configuration plans based on the final recommendation for the component to be recommended, each computing device configuration plan includes a final recommendation for a chassis and final recommendation plans for other hardware components based on the final recommendation for the chassis, the other hardware components including at least one of the CPU, the memory, the hard disk, and the disk array card;

[0179] A large language model based on a prompt project is called to screen an optimal computing device configuration scheme from the multiple computing device configuration schemes, and the prompt words set in the prompt project are used to reflect the user's application requirements for the computing device.

[0180] The computing device configuration recommendation apparatus provided in the embodiment of the present application and the computing device configuration recommendation method provided in the embodiment of the present application are based on the same inventive concept and have the same beneficial effects as the methods adopted, operated or implemented therein.

[0181] The embodiment of the present application also provides an electronic device to perform the above-mentioned computing device configuration recommendation method. Figure 7 It shows a schematic diagram of an electronic device provided by some implementation methods of the present application. Figure 7 As shown, the electronic device 7 includes: a processor 700, a memory 701, a bus 702 and a communication interface 703, and the processor 700, the communication interface 703 and the memory 701 are connected via the bus 702; the memory 701 stores a computer program that can be run on the processor 700, and when the processor 700 runs the computer program, it executes the computing device configuration recommendation method provided by any of the aforementioned embodiments of the present application.

[0182] The memory 701 may include high-speed random access memory (RAM) and may also include non-volatile memory, such as at least one disk storage. The communication connection between the device network element and at least one other network element is achieved through at least one communication interface 703 (which may be wired or wireless), and may use the Internet, a wide area network, a local area network, a metropolitan area network, etc.

[0183] Bus 702 may be an ISA bus, a PCI bus, or an EISA bus. The bus may be divided into an address bus, a data bus, a control bus, and the like. Memory 701 is used to store programs, and processor 700 executes the programs upon receiving execution instructions. The computing device configuration recommendation method disclosed in any of the aforementioned embodiments of the present application may be applied to or implemented by processor 700.

[0184] The processor 700 may be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above method can be completed by hardware integrated logic circuits in the processor 700 or by software instructions. The processor 700 may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), an off-the-shelf field programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor may be a microprocessor or any conventional processor. The steps of the method disclosed in conjunction with the embodiments of this application can be directly implemented and executed by a hardware decoding processor, or by a combination of hardware and software modules in the decoding processor. The software module can be located in a storage medium mature in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, etc. The storage medium is located in the memory 701 , and the processor 700 reads the information in the memory 701 and completes the steps of the above method in combination with its hardware.

[0185] The electronic device provided in the embodiment of the present application and the computing device configuration recommendation method provided in the embodiment of the present application are based on the same inventive concept and have the same beneficial effects as the methods adopted, operated or implemented by them.

[0186] An embodiment of the present application also provides a computer-readable storage medium corresponding to the computing device configuration recommendation method provided in the aforementioned embodiment. The computer-readable storage medium is a CD on which a computer program (i.e., a computer program product) is stored. When the computer program is run by a processor, it will execute the computing device configuration recommendation method provided in any of the aforementioned embodiments.

[0187] It should be noted that examples of the computer-readable storage medium may also include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other optical or magnetic storage media, which are not listed here one by one.

[0188] The computer-readable storage medium provided in the above-mentioned embodiments of the present application and the computing device configuration recommendation method provided in the embodiments of the present application are based on the same inventive concept and have the same beneficial effects as the methods adopted, run or implemented by the application programs stored therein.

[0189] The present application also provides a computer program product. Figure 8 The computer program product 800 carries a program code, namely a computer program 801. The instructions included in the computer program 801 can be used to execute the steps of the computing device configuration recommendation method described in the above method embodiment. For details, please refer to the above method embodiment and will not be repeated here.

[0190] The computer program product may be implemented in hardware, software, or a combination thereof. In one embodiment, the computer program product is implemented as a computer storage medium. In another embodiment, the computer program product is implemented as a software product, such as a software development kit (SDK).

[0191] The basic principles of the present application have been described above in conjunction with specific embodiments. However, it should be noted that the advantages, strengths, and effects mentioned in this application are merely illustrative and not restrictive, and it should not be assumed that these advantages, strengths, and effects are required of each embodiment of this application. In addition, the specific details disclosed above are merely illustrative and facilitating understanding, and are not restrictive. The above details do not limit this application to necessarily being implemented using the above specific details.

[0192] The block diagrams of the devices, devices, equipment, and systems involved in this application are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As will be appreciated by those skilled in the art, these devices, devices, equipment, and systems can be connected, arranged, or configured in any manner. Words such as "include," "comprise," "have," and the like are open-ended words, meaning "including but not limited to," and can be used interchangeably therewith. The words "or" and "and" used herein refer to the words "and / or" and can be used interchangeably therewith, unless the context clearly indicates otherwise. The word "such as" used herein refers to the phrase "such as but not limited to," and can be used interchangeably therewith.

[0193] Additionally, as used herein, "or" used in a list of items beginning with "at least one" indicates a separate list, so that, for example, a list of "at least one of A, B, or C" means A or B or C, or AB or AC or BC, or ABC (i.e., A and B and C). Furthermore, the word "exemplary" does not mean that the example described is preferred or better than other examples.

[0194] It should also be noted that in the system and method of the present application, each component or each step can be decomposed and / or recombined. Such decomposition and / or recombination should be regarded as equivalent solutions of the present application.

[0195] Various changes, substitutions, and modifications of the technology described herein may be made without departing from the teachings defined by the appended claims. Moreover, the scope of the claims herein is not limited to the specific aspects of the processes, machines, manufactures, compositions of things, means, methods, and actions described above. Currently existing or later developed processes, machines, manufactures, compositions of things, means, methods, or actions that perform substantially the same functions or achieve substantially the same results as the corresponding aspects described herein may be utilized. Accordingly, the appended claims include within their scope such processes, machines, manufactures, compositions of things, means, methods, or actions.

[0196] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of the present application. Therefore, the present application is not intended to be limited to the aspects shown herein, but rather to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0197] The above description has been provided for the purpose of illustration and description. Furthermore, this description is not intended to limit the embodiments of the present application to the forms disclosed herein. Although a number of example aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations thereof.

Claims

1. A method for recommending configuration of a computing device, characterized in that: include: Searching a database based on the requirement parameters of the component to be recommended to obtain a plurality of component recommendation solutions for the component to be recommended, wherein the database is constructed according to a computing device configuration rule; Based on the demand parameters and a pre-trained large language model, selecting a first component recommendation solution from the multiple component recommendation solutions, where a matching degree between the first component recommendation solution and the demand parameters meets a preset matching degree requirement; performing a compatibility check on the first component recommendation solution based on a hardware rule knowledge graph to obtain a compatibility check result, wherein the hardware rule knowledge graph is constructed based on compatibility rules extracted from the computing device configuration rules; Based on the compatibility check result and the first component recommendation solution, a final recommendation solution for the component to be recommended is determined.

2. The method according to claim 1, characterized in that Before searching a database based on the requirement parameters of the component to be recommended input by the user to obtain a plurality of component recommendation solutions for the component to be recommended, the method further includes: Extracting a summary of each piece of information in the computing device configuration rule to obtain an information summary of each piece of information; The information summary is stored to obtain the database.

3. The method according to claim 2, characterized in that Based on the demand parameters of the component to be recommended, a database is searched to obtain multiple component recommendation solutions for the component to be recommended, including: Retrieving from the database a first information digest whose character coincidence with the requirement parameter satisfies a preset coincidence requirement; and retrieving from the database a second information digest whose semantic similarity with the requirement parameter satisfies a preset similarity requirement; The component recommendation scheme formed by the first information digest and the component recommendation scheme formed by the second information digest are used as the multiple component recommendation schemes.

4. The method according to claim 1, wherein Based on the hardware rule knowledge graph, a compatibility check is performed on the first component recommendation solution. Before obtaining the compatibility check result, the method further includes: Using a large model extractor to convert the data format of the compatibility rule into a preset structured format; performing entity naming and relationship extraction on the compatibility rules in the structured format using the large model extractor; generating a plurality of triples based on the entity nodes obtained by the entity naming and the node relationships obtained by the relationship extraction; The multiple triples are combined to generate the hardware rule knowledge graph.

5. The method according to claim 1, characterized in that: Based on the hardware rule knowledge graph, a compatibility check is performed on the first component recommendation solution to obtain a compatibility check result, including: Obtaining configuration constraints of the components to be recommended from the hardware rule knowledge graph; A compatibility check is performed on the first component recommendation solution using the configuration constraint to obtain the compatibility check result.

6. The method according to claim 1, wherein Determining a final recommendation for the component to be recommended based on the compatibility check result and the first component recommendation, including: If the compatibility check result indicates that the check is passed, using the first component recommendation solution as the final recommendation solution; When the compatibility check result indicates that the check failed, based on the requirement parameters, the large language model and the reason for failure included in the compatibility check result, a second component recommendation scheme is screened from the multiple component recommendation schemes, whose matching degree with the requirement parameters meets the preset matching requirements, and the second component recommendation scheme is used as the final recommendation scheme.

7. The method according to claim 1, characterized in that In the case where the component to be recommended is a disk array card, the requirement parameters of the component to be recommended include the requirement parameters of the disk array card input by the user and the final recommendation solution of the hard disk; In the case where the component to be recommended is any one of the CPU, memory or the hard disk, the requirement parameters of the component to be recommended include the requirement parameters of any one of the components input by the user and the final recommended solution of the chassis; In a case where the component to be recommended is the chassis, the requirement parameters of the component to be recommended include the requirement parameters of the chassis input by the user.

8. The method according to claim 7, characterized in that When the final recommendation scheme for the disk array card indicates that the type of the disk array card is a standard card, the component to be recommended further includes an expansion card, and the requirement parameters of the expansion card include the final recommendation scheme for the disk array card, the final recommendation scheme for the hard disk, and the requirement parameters of the expansion card input by the user.

9. The method according to claim 7, characterized in that After determining a final recommendation for the component to be recommended based on the compatibility check result and the first component recommendation, the method further includes: Acquire multiple computing device configuration solutions based on the final recommendation solution of the component to be recommended, each of the computing device configuration solutions includes a final recommendation solution for a chassis and final recommendation solutions for other hardware components obtained based on the final recommendation solution for the chassis, the other hardware components including at least one of the CPU, the memory, the hard disk, and the disk array card; A large language model based on a prompt project is called to screen an optimal computing device configuration scheme from the multiple computing device configuration schemes, and the prompt words set in the prompt project are used to reflect the user's application requirements for the computing device.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory, wherein: The processor executes the computer program to implement the steps of the method according to any one of claims 1 to 9.

Images