Product configuration recommendation method, system and equipment

Abstract

The invention relates to the technical field of computers, and discloses a product configuration recommendation method, system and equipment, and the method comprises the steps: carrying out the processing of an original configuration demand of a user through an information processing engine, and obtaining a target user demand; determining a plurality of target products from a preset product library through a configuration recommendation engine based on the target user demand, a preset recommendation scene, a preset configuration rule and a preset knowledge graph, and combining the plurality of target products to obtain an initial product configuration result; evaluating the initial product configuration result based on the historical configuration data through a self-learning engine to obtain an evaluation result; under the condition that the evaluation result meets a preset condition, determining the initial product configuration result as a target product configuration result and outputting the target product configuration result; the recommendation accuracy of the target product configuration result can be further ensured.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of computer technology, and in particular to a product configuration recommendation method, system, and device. Background Art

[0002] For 2B businesses, product configuration combinations often require consideration from multiple dimensions, such as specification and capability matching and cost, to meet user demands in different scenarios. Currently, product configuration combinations can be completed by screening matching products based on their specifications and capabilities.

[0003] However, in real-world applications, user demands are often complex. For example, beyond product specifications and capabilities, user requirements may also involve delivery cycles, costs, and other factors. This makes it difficult to adapt products based on product specifications and capabilities to meet actual needs. Summary of the Invention

[0004] The embodiments of the present application provide a product configuration recommendation method, system, and device, which can improve the accuracy of product configuration recommendations.

[0005] To achieve the above objectives, the present invention adopts the following technical solutions:

[0006] In a first aspect, an embodiment of the present application provides a product configuration recommendation method, which is applied to a product configuration recommendation system, wherein the product configuration recommendation system includes an information processing engine, a configuration recommendation engine, and a self-learning engine; the method includes: processing the user's original configuration requirements through the information processing engine to obtain the target user requirements; determining multiple target products from a preset product library based on the target user requirements, as well as preset recommendation scenarios, preset configuration rules, and preset knowledge graphs through the configuration recommendation engine, and combining the multiple target products to obtain an initial product configuration result; evaluating the initial product configuration result based on historical configuration data through the self-learning engine to obtain an evaluation result; when the evaluation result meets the preset conditions, determining the initial product configuration result as the target product configuration result and outputting it; wherein the preset conditions are used to indicate the recommendation accuracy of the initial product configuration.

[0007] Based on the above scheme, the user's original configuration requirements are processed by the information processing engine to obtain more accurate target user requirements, which can improve the recommendation accuracy of the initial product configuration results obtained subsequently by the configuration recommendation engine; in addition, the self-learning engine introduces existing successful configuration cases (historical configuration data) to evaluate the initial product configuration results. When the evaluation results corresponding to the initial product configuration results meet the preset conditions, it means that the initial product configuration results are more in line with user needs. At this time, the initial product configuration results are determined as the target product configuration results, which can further ensure the recommendation accuracy of the target product configuration results.

[0008] In another possible implementation, the self-learning engine evaluates the initial product configuration result based on the historical configuration data to obtain an evaluation result, including: determining, through the self-learning engine, from the historical configuration data, first historical configuration data whose similarity to the initial product configuration result is higher than a similarity threshold; and evaluating the initial configuration result based on the first historical configuration data by the self-learning engine to obtain an evaluation result.

[0009] Based on the above solution, after obtaining the initial product configuration, first historical configuration data similar to the initial product configuration result can be determined from the historical configuration data based on the initial product configuration. By comparing and analyzing the initial product configuration result and the first historical configuration data, the initial configuration result can be evaluated using performance, user needs, and other feedback data corresponding to the first historical configuration data as a reference to obtain an evaluation result. Based on the evaluation result corresponding to the initial configuration result, deficiencies in the initial configuration result can be identified, thereby facilitating subsequent adjustments to the product configuration recommendation strategy and improving the accuracy of product configuration recommendations.

[0010] In another possible implementation, the method further includes: when the evaluation result does not meet the preset conditions, inputting the evaluation result into the configuration recommendation engine through the self-learning engine, so that the configuration recommendation engine is iteratively trained based on the evaluation result to obtain a new configuration recommendation engine; determining a new initial product configuration result through the new configuration recommendation engine based on target user needs, preset recommendation scenarios and preset configuration rules; evaluating the new initial product configuration result based on historical configuration data through the self-learning engine to obtain a new evaluation result, and determining the new initial product configuration result as the target product configuration result when the new evaluation result meets the preset conditions.

[0011] Based on the above scheme, if the evaluation result corresponding to the initial product configuration result does not meet the preset conditions, it means that the initial product configuration result does not meet the user's needs. In this case, the evaluation result corresponding to the initial product configuration result is reversely input into the configuration recommendation engine, so that the configuration recommendation engine is iteratively trained based on the evaluation result to obtain a new configuration recommendation engine. The new configuration recommendation engine then re-produces product configuration recommendations until the initial product configuration result meets the preset conditions. In this way, iterative training of the configuration recommendation engine based on the evaluation results can make the product configuration results recommended by the configuration recommendation engine increasingly accurate, thereby improving the accuracy of product configuration recommendations.

[0012] In another possible implementation, the original configuration requirements are processed by an information processing engine to obtain target user requirements, including: performing requirement analysis on the original configuration requirements by the information processing engine to obtain initial user requirements; and performing error correction on the initial user requirements based on a preset industry knowledge base and a preset prompt word project by the information processing engine to obtain target user requirements.

[0013] Based on the above solution, the information processing engine combines the preset industry knowledge base and preset prompt word engineering with the error correction and annotation processing of the initial user needs, which can improve the accuracy and standardization of the target user needs, so that the subsequent configuration recommendation engine can quickly locate key information from the target user needs.

[0014] In another possible implementation, the multiple target products include a first product, a second product, a third product, a fourth product and a fifth product; wherein the first product, the second product, the third product, the fourth product and the fifth product are respectively different components in the computing device; by configuring the recommendation engine based on the target user needs, as well as the preset recommendation scenarios, preset configuration rules and preset knowledge graph, the multiple target products are determined from the preset product library, including: by configuring the recommendation engine based on the target user needs, the preset recommendation scenarios, the preset configuration rules and the preset knowledge graph, the first product, the second product and the third product are determined from the preset product library; by configuring the recommendation engine based on the second product, the preset recommendation scenarios, the preset configuration rules and the preset knowledge graph, the fourth product is determined from the preset product library; by configuring the recommendation engine based on the preset recommendation scenarios, the preset configuration rules and the preset knowledge graph, as well as the first product, the second product, the third product and the fourth product, the fifth product is determined.

[0015] Based on the above solution, based on the target user needs, as well as preset recommendation scenarios, preset configuration rules and preset knowledge graphs, multiple target products are determined in stages from the preset product library, and the recommendation logic can be gradually refined to ensure compatibility and performance matching between the target products.

[0016] In another possible implementation, a recommendation engine is configured to determine a first product, a second product, and a third product from a preset product library based on target user needs, preset recommendation scenarios, preset configuration rules, and a preset knowledge graph, including: decomposing the target user needs by configuring the recommendation engine to obtain target recommendation prompt words corresponding to multiple target products; configuring the recommendation engine to determine the first product, the second product, and the third product from the preset product library based on the target recommendation prompt words corresponding to multiple target products, as well as the preset recommendation scenarios, preset configuration rules, and preset knowledge graph.

[0017] Based on the above solution, by configuring the recommendation engine to decompose the target user needs, the target user needs can be broken down into independent prompt words (i.e., target recommendation prompt words) to correspond to different product dimensions respectively, avoiding fuzzy matching caused by mixed needs when recommending products subsequently, thereby improving the recommendation accuracy of the first product, the second product, and the third product; in addition, by comprehensively combining the preset recommendation scenarios, preset configuration rules, and the preset knowledge graph, the first product, the second product, and the third product can be determined respectively from the preset product library, which can ensure the compatibility and performance synergy between the first product, the second product, and the third product.

[0018] In another possible implementation, a fourth product is determined from a preset product library by configuring a recommendation engine based on the second product, as well as preset recommendation scenarios, preset configuration rules, and a preset knowledge graph, including: determining multiple candidate products from the preset product library by configuring a recommendation engine based on the second product, as well as preset recommendation scenarios, preset configuration rules, and a preset knowledge graph; and determining the fourth product from multiple candidate products by configuring the recommendation engine based on the redundancy of each candidate product relative to the second product.

[0019] Based on the above scheme, by integrating the preset recommendation scenarios, preset configuration rules and preset knowledge graphs, multiple candidate products that match the second product are determined from the preset product library, which can ensure the compatibility and performance synergy between each candidate product and the second product; and based on the redundancy of each candidate product relative to the second product, the fourth product is determined from multiple candidate products, which can avoid over-configuration and control costs.

[0020] In another possible implementation, a fourth product is determined from multiple candidate products based on the redundancy of each candidate product relative to the second product by configuring a recommendation engine, including: configuring the recommendation engine to determine, as the fourth product, a candidate product whose redundancy relative to the second product is less than a preset redundancy threshold among the multiple candidate products.

[0021] Based on the above scheme, since the candidate products whose redundancy among multiple candidate products is greater than or exceeds the preset redundancy threshold may have configuration redundancy problems, the candidate products whose redundancy relative to the second product is less than the preset redundancy threshold among multiple candidate products are determined as the fourth product through the configuration recommendation engine, and unnecessary redundant configurations are eliminated to ensure that the obtained fourth product meets the needs without being over-configured.

[0022] In the second aspect, an embodiment of the present application also provides a product configuration recommendation system, including an information processing engine, a configuration recommendation engine and a self-learning engine module; the information processing engine is configured to: process the user's original configuration requirements to obtain target user requirements; the configuration recommendation engine is configured to: based on the target user requirements, as well as preset recommendation scenarios, preset configuration rules and preset knowledge graphs, determine multiple target products from a preset product library, and combine the multiple target products to obtain an initial product configuration result; the self-learning engine is configured to: evaluate the initial product configuration result based on historical configuration data to obtain an evaluation result; when the evaluation result meets the preset conditions, the initial product configuration result is determined as the target product configuration result and output; wherein, the preset conditions are used to indicate the recommendation accuracy of the initial product configuration.

[0023] In a third aspect, an embodiment of the present application further provides a computing device comprising: a processor and a memory; the processor and the memory are coupled; the memory is used to store program instructions; and the processor is used to execute program instructions to perform any method as described in the first aspect above.

[0024] In a fourth aspect, an embodiment of the present application provides a chip, which is used to execute any method as described in the first aspect above.

[0025] In a fifth aspect, an embodiment of the present application provides a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a computer, the method as described in any one of the first aspects is implemented.

[0026] In a sixth aspect, an embodiment of the present application provides a program product, comprising a computer program, which implements any method in the first aspect when executed by a processor. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 A flowchart of a product configuration recommendation method provided in an embodiment of the present application;

[0028] Figure 2 A schematic diagram of the process of another product configuration recommendation method provided in an embodiment of the present application;

[0029] Figure 3 A schematic diagram of a product configuration recommendation system provided in an embodiment of the present application;

[0030] Figure 4 A schematic diagram of the process of another product configuration recommendation method provided in an embodiment of the present application;

[0031] Figure 5 A schematic diagram of a configuration recommendation engine provided in an embodiment of the present application;

[0032] Figure 6 A schematic diagram of the process of another product configuration recommendation method provided in an embodiment of the present application;

[0033] Figure 7 A schematic diagram of the process of another product configuration recommendation method provided in an embodiment of the present application;

[0034] Figure 8 A schematic diagram of the process of another product configuration recommendation method provided in an embodiment of the present application;

[0035] Figure 9 A schematic diagram of another product configuration recommendation system provided in an embodiment of the present application;

[0036] Figure 10 A schematic diagram of a computing device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0037] The following will describe the technical solutions in the embodiments of the present application in conjunction with the accompanying drawings. To facilitate the clear description of the technical solutions in the embodiments of the present application, the first, second, etc. descriptions in the embodiments of the present application are only used for illustration and to distinguish the described objects. There is no order, nor does it represent a special limitation on the number of devices in the embodiments of the present application, and it does not constitute any limitation on the embodiments of the present application.

[0038] Before explaining the product configuration recommendation method provided in the embodiments of the present application, the terms involved in one or more embodiments of the present application are first explained.

[0039] Cue word engineering: Various techniques and methods for providing cue words for generative AI tools.

[0040] Optical Character Recognition (OCR): Optical character recognition can convert text on an image into text.

[0041] Knowledge graph: A series of various graphics that show the development process and structural relationship of knowledge, using visualization technology to describe knowledge resources and their carriers, and to mine, analyze, construct, draw and display knowledge and their interrelationships.

[0042] The product configuration recommendation method provided in the embodiment of the present application can be applied to a computing device, which can be a server or a terminal device. Based on the product configuration recommendation method provided in the embodiment of the present application, the user's original configuration requirements can be processed by the information processing engine to obtain more accurate target user requirements, thereby improving the accuracy of the initial product configuration results subsequently obtained by the configuration recommendation engine; it can also introduce existing successful configuration cases (historical configuration data) through the self-learning engine to evaluate the initial product configuration results, further ensuring the accuracy of the product configuration recommendation results.

[0043] Figure 1 This is a flowchart of a product configuration recommendation method provided in an embodiment of the present application. The product configuration recommendation method provided in an embodiment of the present application is applied to a product configuration recommendation system. The product configuration recommendation system includes an information processing engine, a configuration recommendation engine, and a self-learning engine. The product configuration recommendation system can be deployed on a computing device, which can be a server or a terminal device. Figure 1 As shown, the product configuration recommendation method applied to the product configuration recommendation system may include S101-1-S103.

[0044] S101. Process the user's original configuration requirements through an information processing engine to obtain target user requirements.

[0045] In some embodiments, the information processing engine is used to receive the user's original configuration requirements and process the user's original configuration requirements to obtain target user requirements. The target user requirements refer to structured original configuration requirements.

[0046] For example, since users often describe their desired component specifications and quantities in open language based on their own understanding, the original configuration requirements entered by users are typically open-ended. Therefore, upon receiving the user's original configuration requirements, the information processing engine can process them to extract the open information in the original configuration requirements and convert it into a relatively consistent structured expression, namely the target user requirement. For example, if the user's original configuration requirement is "I want two 32GB hard drives," the target user requirement can be "Specifications: 32GB, Quantity: 2."

[0047] In some embodiments, as Figure 2 As shown, S101 may include S201 - S202 .

[0048] S201: Analyze the original configuration requirements through an information processing engine to obtain initial user requirements.

[0049] In some embodiments, the information processing engine can receive the original configuration requirements input by the user; after the information processing engine receives the original configuration requirements of the user, the information processing engine can perform demand analysis on the original configuration requirements through a large model, such as a large language model (LLM), to extract the text information in the original configuration requirements and obtain the initial user requirements.

[0050] In some embodiments, to further improve the accuracy of user requirements, the information processing engine may pre-process the user's original configuration requirements (i.e., semantic recognition) using OCR (Optical Character Recognition) technology before parsing the original configuration requirements using the large model.

[0051] S202: Using an information processing engine, based on a preset industry knowledge base and a preset prompt word project, the initial user requirements are corrected and annotated to obtain target user requirements.

[0052] In some embodiments, since misplacement, typos, or omissions may occur during the execution of S201, in order to improve the accuracy of user needs and thus improve the accuracy of subsequent product configuration recommendations, the initial user needs can be corrected by the information processing engine based on a preset industry knowledge base and a preset prompt word project to obtain the target user needs.

[0053] For example, the information processing engine performs semantic repair on the text in the initial user requirements based on a preset industry knowledge base and a preset prompt word project, extracts key requirement dimensions (such as performance requirements, budget, and delivery cycle) from the semantically repaired initial user requirements, and annotates core components (such as CPU, GPU) and specification parameters (such as "budget ≤ 120,000"). For example, if the original configuration requirement is "a CPU that supports large-scale parallel computing", after executing S201-S202, the information processing engine can annotate the original configuration requirement as "CPU type: multi-core high main frequency, number of cores ≥ 32", that is, the target user requirement is "CPU type: multi-core high main frequency, number of cores ≥ 32".

[0054] In some embodiments, both the preset industry knowledge base and the preset prompt word project are pre-configured in the product configuration recommendation system. The data included in the preset industry knowledge base can be configured based on actual needs. For example, the preset industry knowledge base can include product parameters, customer data, inventory data, contract data, etc. This embodiment of the application does not limit the data included in the preset industry library. The preset prompt word project can include a structured template. After the information processing engine performs error correction and annotation processing on the initial user requirements, the annotated information can be filled into the structured template included in the preset prompt word project to obtain the target user requirements.

[0055] It is understandable that the information processing engine, based on the preset industry knowledge base and preset prompt word engineering, combined with the error correction and annotation processing of initial user needs, can improve the accuracy and standardization of target user needs, so that the configuration recommendation engine can quickly locate key information from the target user needs.

[0056] In some embodiments, Figure 3 Figure 1 shows a schematic diagram of a product configuration recommendation system. Figure 3 As shown, after the user's original configuration requirements are input into the information processing engine, the information processing engine performs requirement analysis and core specification annotation (i.e., error correction and annotation processing) on ​​the user's original configuration requirements to obtain the target user requirements.

[0057] In some embodiments, the user can be guided to gradually refine their requirements through a guided dialogue to obtain the complete original configuration requirements. Alternatively, the user's requirements can be obtained through a questionnaire to obtain the complete original configuration requirements. This embodiment of the present application is not limited to this.

[0058] It is understandable that processing the user's original configuration requirements through the information processing engine can ensure the accuracy of understanding natural language, improve the accuracy of user requirements (target user requirements), and thus improve the accuracy of subsequent product configuration recommendations.

[0059] S102. By configuring the recommendation engine based on target user needs, preset recommendation scenarios, preset configuration rules and preset knowledge graph, multiple target products are determined from the preset product library, and the multiple target products are combined to obtain an initial product configuration result.

[0060] In some embodiments, the configuration recommendation engine is used to comprehensively consider the compatibility, performance balance and price factors among various components, and screen and combine a configuration list that meets user needs.

[0061] Exemplarily, the information processing engine sends the target user needs to the configuration recommendation engine; after receiving the target user needs, the configuration recommendation engine can determine multiple target products from the preset product library based on the target user needs, as well as preset recommendation scenarios, preset configuration rules and preset knowledge graphs, and combine the multiple target products to obtain an initial product configuration result.

[0062] In some embodiments, preset recommendation scenarios are pre-set in the product configuration recommendation system. Pre-set recommendation scenarios refer to scenarios in which multiple target products are recommended and can be set based on actual needs. The present application embodiment does not limit the data included in the preset recommendation scenarios.

[0063] For example, the preset recommendation scenario may include at least one of low cost, excellent delivery time, and excellent performance. Low cost refers to a low total cost for multiple target products; excellent delivery time refers to a short delivery cycle for multiple target products; and excellent performance refers to high performance of multiple target products in terms of functionality, efficiency, stability, user experience, and other indicators. In some application scenarios, the default recommendation scenario may be low cost. Upon receiving a selection instruction from the user, the recommendation scenario may be adjusted to excellent delivery time or excellent performance based on the user's selection instruction.

[0064] In some embodiments and application scenarios, when the information processing engine sends the target user's needs to the configuration recommendation engine, the target user's needs may include a preset recommendation scenario. In some application scenarios, the information processing engine sends the target user's needs to the configuration recommendation engine, and after receiving the target user's needs, the configuration recommendation engine may proactively obtain the preset recommendation scenario. This embodiment of the present application is not limited to this.

[0065] In some embodiments, the preset configuration rules and the preset knowledge graph are stored in a configuration rule module and a knowledge graph module, respectively. Both the configuration rule module and the knowledge graph module are pre-deployed in the product configuration recommendation system. The preset configuration rules indicate the technical parameters and configuration requirements of the knowledge product, while the preset knowledge graph indicates the constraints between products in the product configuration.

[0066] For example, a preset knowledge graph can be constructed based on the product information corresponding to each product in the preset product library, and preset configuration rules that meet the configuration requirements and budget levels can be specified. Product information can include product model, product parameters, product price, etc. The present embodiment of the application does not limit the data included in the product information.

[0067] In some embodiments, the information processing engine processes the user's original configuration requirements to obtain the target user requirements, and then the information processing engine can send the target user requirements to the configuration recommendation engine. After the configuration recommendation engine receives the target user requirements from the information processing engine, it can determine multiple target products from the preset product library based on the target user requirements, preset recommendation scenarios, preset configuration rules, and preset knowledge graphs. It is also possible to first decouple the recommendation logic into multiple stages, and then determine multiple target products from the preset product library in stages based on the target user requirements, preset recommendation scenarios, preset configuration rules, and preset knowledge graphs. This is not limited in the embodiments of the present application.

[0068] In some embodiments, the plurality of target products include a first product, a second product, a third product, a fourth product, and a fifth product, wherein the first product, the second product, the third product, the fourth product, and the fifth product are different components in a computing device.

[0069] In some embodiments, the first product, the second product, the third product, the fourth product, and the fifth product may respectively represent different parts of the hardware system of the computing device.

[0070] For example, the first product may be a processor (CPU), which is used to execute instructions and operations to control all operations of the computing device, such as data processing and program execution. The second product may be a hard disk, which is used to store the operating system, software and files. The third product may be a memory, which is used to store data and programs during the operation of the computing device. The fourth product may be a chassis, which is used to place the various components included in the computing device. The fifth product may be an IO component, which can also be called an input / output (I / O) component. The IO component may include IO devices, IO interfaces (such as USB, PCI) and IO software, etc. Among them, IO devices may include keyboards, mice, printers, etc.; IO interfaces are used to connect IO devices to the internal bus of the computer, and may include Universal Serial Bus (USB), Peripheral Component Interconnect (PCI) bus, etc.; IO software may include drivers and I / O management modules in the operating system, which are used to manage the operation of I / O devices.

[0071] It should be noted that the number of product types included in multiple target products will vary depending on the application scenario. The data included in multiple target products can be determined based on the actual application scenario. This embodiment of the present application is not limited to this. The following embodiment uses the product configuration recommendation method provided in the embodiment of the present application as an example for illustrative purposes, applying it to the installation configuration scenario.

[0072] Exemplarily, the configuration recommendation engine can directly determine the CPU, hard disk, memory, chassis and IO components that meet the user's needs from the preset product library based on the target user's needs, preset recommendation scenarios, preset configuration rules and preset knowledge graphs. The configuration recommendation engine can also determine the CPU, hard disk, memory, chassis and IO components that meet the user's needs from the preset product library in stages. For example, the configuration recommendation engine can decouple the recommendation logic into three stages: three major component recommendations, chassis recommendations, and IO component recommendations. Among them, the three major component recommendations refer to recommending CPU, memory and hard disk; chassis recommendations refer to recommending chassis models. Among them, since the chassis is related to the hard disk and the board; for example, the xx model board needs to be installed in the XX model chassis; for another example, the hard disk position in the chassis needs to be able to accommodate an xx-inch hard disk. Therefore, the chassis recommendation needs to refer to the corresponding requirements of the hard disk and the board.

[0073] In some embodiments, a board refers to a circuit board that combines multiple electronic components to expand the hardware functions of a computing device. Boards may include graphics cards (GPUs), network cards, PCIe cards, etc., which are not limited in this embodiment of the application.

[0074] For example, in some application scenarios, the board can be determined based on the needs of the target user, that is, provided by the user; in other application scenarios, the board can also be determined based on the reasoning of the CPU, memory and hard disk. The embodiment of the present application does not limit the determination method of the board. The following embodiment is illustrated by taking the board determined based on the needs of the target user as an example. IO component recommendation refers to the recommended model and / or quantity of IO components; since IO components are related to the CPU, hard disk, memory and chassis, the IO component recommendation needs to refer to the corresponding requirements of the CPU, hard disk, memory and chassis.

[0075] It can be understood that based on the target user needs, as well as preset recommendation scenarios, preset configuration rules and preset knowledge graphs, multiple target products can be determined in stages from the preset product library, and the recommendation logic can be gradually refined to ensure compatibility and performance matching between the target products.

[0076] like Figure 4 As shown, S102 may include S401-S403.

[0077] S401. Determine a first product, a second product, and a third product from a preset product library by configuring a recommendation engine based on target user needs, preset recommendation scenarios, preset configuration rules, and a preset knowledge graph.

[0078] S402. Determine a fourth product from a preset product library by configuring a recommendation engine based on the second product, as well as a preset recommendation scenario, preset configuration rules, and a preset knowledge graph.

[0079] S403. Determine a fifth product by configuring a recommendation engine based on a preset recommendation scenario, preset configuration rules, and a preset knowledge graph, as well as the first product, the second product, the third product, and the fourth product.

[0080] In some embodiments, after the information processing engine processes the user's original configuration requirements to obtain the target user requirements, the information processing engine can send the target user requirements to the configuration recommendation engine. After receiving the target user requirements from the information processing engine, the configuration recommendation engine can first determine the first product, the second product, and the third product from the preset product library based on the target user requirements, the preset recommendation scenario, the preset configuration rules, and the preset knowledge graph. After determining the second product, the fourth product is determined from the preset product library based on the second product, the preset recommendation scenario, the preset configuration rules, and the preset knowledge graph. After determining the first product, the second product, the third product, and the fourth product, the fifth product is determined based on the first product, the second product, the third product, and the fourth product, as well as the preset recommendation scenario, the preset configuration rules, and the preset knowledge graph.

[0081] In some embodiments, the number of preset knowledge graphs may be one or more. Where there are multiple preset knowledge graphs, each of the multiple preset knowledge graphs may correspond to a different product category. In some application scenarios, each preset knowledge graph may correspond to at least one product category. The present embodiment of the application does not limit the number of knowledge graphs and the types of knowledge graphs included in the multiple preset knowledge graphs.

[0082] Exemplarily, a plurality of preset knowledge graphs may include a knowledge graph of three major components, a knowledge graph of boards and cards, and a knowledge graph of power supplies. Among them, the knowledge graph of the three major components refers to the knowledge graph corresponding to the CPU, memory, and hard disk, which is used to indicate the mutual relationship between the CPU, memory, and hard disk. For example, the knowledge graph of the three major components may include the physical connection relationship, data flow relationship, performance dependency relationship, functional collaboration relationship, etc. between the CPU, memory, and hard disk. The knowledge graph of boards and cards may include the physical connection relationship, functional collaboration relationship, performance dependency relationship, and compatibility constraint relationship corresponding to the boards and cards. The knowledge graph of power supplies may include the energy conversion relationship, interface connection relationship, power distribution relationship, security collaboration relationship, performance dependency relationship, and compatibility constraint relationship with hardware corresponding to the power supply. It should be noted that the knowledge graph of the three major components, the knowledge graph of boards and cards, and the knowledge graph of power supplies can be constructed according to actual needs, and the embodiments of this application do not limit the knowledge graph of the three major components, the knowledge graph of boards and cards, and the knowledge graph of power supplies.

[0083] In some embodiments, when the configuration recommendation engine decouples the recommendation logic into three stages: three major parts recommendation, chassis recommendation, and IO component recommendation, multiple preset knowledge graphs may include the three major parts knowledge graph, the board chassis knowledge graph, and the IO component knowledge graph.

[0084] For example, Figure 3As shown, when the information processing engine sends the target user demand to the configuration recommendation engine, the target user demand carries a preset recommendation scenario. After the configuration recommendation engine receives the target user demand, the configuration recommendation engine can determine multiple target products from the preset product library based on the preset recommendation scenario and preset configuration rules, as well as the three major parts knowledge graph, the board chassis knowledge graph and the IO component knowledge graph. That is, when executing S401, the configuration recommendation engine can determine the CPU (i.e., the first product), hard disk (i.e., the second product) and memory (i.e., the third product) from the preset product library based on the target user demand, the preset recommendation scenario, the preset configuration rules and the three major parts knowledge graph. When executing S402, the configuration recommendation engine can determine the chassis (i.e., the fourth product) from the preset product library based on the hard disk and the board provided by the user, as well as the preset recommendation scenario, the preset configuration rules and the board chassis knowledge graph. When executing S402, the configuration recommendation engine can determine the IO component (i.e., the fifth product) based on the preset recommendation scenario, the preset configuration rules and the IO component knowledge graph, as well as the CPU, hard disk, memory and chassis.

[0085] In some embodiments, S401 may include: decomposing the target user needs by configuring a recommendation engine to obtain target recommendation prompt words corresponding to multiple target products; configuring the recommendation engine to determine the first product, the second product, and the third product from the preset product library based on the target recommendation prompt words corresponding to the multiple target products, as well as the preset recommendation scenarios, preset configuration rules, and preset knowledge graphs.

[0086] For example, Figure 5 As shown, the configuration recommendation engine may include a decomposer and multiple recommendation nodes. The multiple recommendation nodes may include a first recommendation node, a second recommendation node, and a third recommendation node. In the installation configuration scenario, the information processing engine processes the user's original configuration requirements to obtain target user requirements. The information processing engine may then send the target user requirements to the configuration recommendation engine. After receiving the target user requirements from the information processing engine, the configuration recommendation engine may send the target user requirements to the decomposer within the configuration recommendation engine. The decomposer decomposes the target user requirements to obtain target recommendation prompts corresponding to the CPU (i.e., the first product), the hard disk (i.e., the second product), and the memory (i.e., the third product). The target recommendation prompts corresponding to the CPU, memory, and hard disk are then input into three parallel nodes: the first recommendation node, the second recommendation node, and the third recommendation node. The first recommendation node, the second recommendation node, and the third recommendation node generate recommendations for the CPU, hard disk, and memory, respectively, resulting in the first, second, and third products.

[0087] It should be noted that multiple recommendation nodes in the configuration recommendation engine are parallel nodes. In different application scenarios, the number of recommendation nodes can be the same or different. The number of recommendation nodes can be set according to actual needs and is not limited in this embodiment of the application.

[0088] It is understandable that by configuring the recommendation engine to decompose the target user needs, the target user needs can be broken down into independent prompt words (i.e., target recommendation prompt words) to correspond to different product dimensions respectively, so as to avoid fuzzy matching caused by mixed needs when recommending products subsequently. This can improve the recommendation accuracy of the first product, the second product, and the third product; in addition, by comprehensively considering the preset recommendation scenarios, preset configuration rules, and preset knowledge graphs, the first product, the second product, and the third product can be determined respectively from the preset product library, which can ensure the compatibility and performance synergy between the first product, the second product, and the third product.

[0089] In some embodiments, as Figure 6 As shown, S402 may include S601-S603.

[0090] S601. Determine multiple candidate products from a preset product library by configuring a recommendation engine based on the second product, as well as preset recommendation scenarios, preset configuration rules, and preset knowledge graphs.

[0091] S602: Determine a fourth product from multiple candidate products by configuring a recommendation engine based on the redundancy of each candidate product relative to the second product.

[0092] In some embodiments, the redundancy relative to the second product indicates the degree to which each candidate product exceeds the configuration requirements of the second product. The higher the degree to which a candidate product exceeds the configuration requirements of the second product, the higher the redundancy of the candidate product relative to the second product; the lower the degree to which a candidate product exceeds the configuration requirements of the second product, the lower the redundancy of the candidate product relative to the second product.

[0093] For example, the redundancy of the chassis (candidate product) relative to the hard disk (second product) may refer to the capacity of the chassis exceeding the number or volume of the hard disks required to be configured.

[0094] In some embodiments, by configuring a recommendation engine based on the second product, as well as preset recommendation scenarios, preset configuration rules and preset knowledge graphs, multiple products in a preset product library are screened to determine multiple candidate products that best match the hard drive (second product) from a chassis vector database in the preset product library, and then determine the redundancy of each candidate product (chassis) relative to the second product, and determine a fourth product from multiple candidate products based on the redundancy of each candidate product relative to the second product.

[0095] It can be understood that by comprehensively considering the preset recommendation scenarios, preset configuration rules and preset knowledge graphs, and determining multiple candidate products that match the second product from the preset product library, the compatibility and performance synergy between each candidate product and the second product can be ensured; and determining the fourth product from multiple candidate products based on the redundancy of each candidate product relative to the second product can avoid over-configuration and control costs.

[0096] In some embodiments, determining the fourth product from multiple candidate products based on the redundancy of each candidate product relative to the second product by configuring the recommendation engine may include: determining, by configuring the recommendation engine, a candidate product among the multiple candidate products whose redundancy relative to the second product is less than a preset redundancy threshold as the fourth product.

[0097] In some embodiments, the preset redundancy threshold is a preset value that can be pre-set in the configuration recommendation engine. The value of the preset redundancy threshold can be set according to actual needs. The embodiment of the present application does not limit the value of the preset redundancy threshold.

[0098] For example, in an installation configuration scenario, since the selection of a chassis (i.e., the fourth product) is limited by the hard drive (i.e., the second product), when determining 10 candidate chassis (i.e., candidate products) that best match the hard drive from a preset product library based on the hard drive, as well as a preset recommendation scenario, preset configuration rules, and preset knowledge graph, through a configuration recommendation engine, the chassis can be determined from the 10 candidate chassis based on the redundancy of each chassis relative to the hard drive. For example, there may be 6 candidate chassis among the 10 candidate chassis, and their redundancy relative to the hard drive is greater than or equal to the preset redundancy threshold, that is, the above 6 candidate chassis may have configuration redundancy issues relative to the hard drive. At this time, in order to eliminate unnecessary redundant configurations and ensure that the recommended chassis meets the requirements without being over-configured, the 4 candidate chassis among the 10 candidate chassis whose redundancy relative to the hard drive is less than the preset redundancy threshold can be determined as the chassis, that is, the 4 candidate chassis among the 10 candidate chassis other than the above 6 candidate chassis are determined as the chassis.

[0099] It is understandable that since the candidate products whose redundancy among multiple candidate products is greater than or exceeds the preset redundancy threshold may have configuration redundancy problems, the candidate product whose redundancy relative to the second product among multiple candidate products is less than the preset redundancy threshold is determined as the fourth product through the configuration recommendation engine, eliminating unnecessary redundant configurations, ensuring that the obtained fourth product meets the needs without being over-configured.

[0100] S103. Evaluate the initial product configuration result based on historical configuration data through a self-learning engine to obtain an evaluation result. If the evaluation result meets a preset condition, determine the initial product configuration result as a target product configuration result and output it.

[0101] In some embodiments, the preset condition is used to indicate the recommendation accuracy of the initial product configuration.

[0102] Exemplarily, an evaluation result that satisfies a preset condition means that the recommendation accuracy of the initial product configuration is high, and an evaluation result that does not satisfy the preset condition means that the recommendation accuracy of the initial product configuration is low.

[0103] In some embodiments, historical configuration data refers to product configuration data generated by the product configuration recommendation system within a preset time range. The preset time range can be set according to user needs, such as a preset time range of 7 days, 1 month, etc., which is not limited in this embodiment of the present application.

[0104] Exemplarily, after the configuration recommendation engine obtains the initial product configuration, it sends the initial product configuration to the self-learning engine. Upon receiving the initial product configuration from the configuration recommendation engine, the self-learning engine retrieves historical configuration data from a pre-set case library and evaluates the initial product configuration using the historical configuration data as a reference to obtain an evaluation result.

[0105] In some embodiments, historical configuration data may include product configuration information recommended based on historical user needs within a preset time period, user feedback information on the product configuration, and transaction information corresponding to the product configuration. Among them, product configuration information includes user information, product information, cost information, delivery cycle, etc. User information is used to indicate the user's basic information; for example, user name, user type, etc.; product information may include product name, product model, etc.; cost information is used to indicate the price of the product; and the delivery cycle is used to indicate the complete time span for the product to be actually delivered to the customer. Feedback information is used to indicate the user's evaluation of the product configuration output by the product configuration recommendation system. The evaluation information includes satisfaction or dissatisfaction, or the degree of satisfaction, etc. Transaction information is used to indicate whether the user adopts the product configuration. The transaction information includes transaction or non-transaction, etc.

[0106] In some embodiments, in order to improve the accuracy of the product configuration output by the product configuration recommendation system, after the self-learning engine receives the initial product configuration sent by the configuration recommendation engine, it can obtain historical configuration data with transaction information as the transaction from a preset case library, and use the historical configuration data with transaction information as the transaction as a reference to evaluate the initial product configuration result and obtain an evaluation result.

[0107] In some embodiments, the evaluation results may include a score corresponding to the initial product configuration and may also include comments on the initial product configuration. The present embodiment of the application does not limit the data included in the evaluation results.

[0108] For example, the preset condition may be that the score corresponding to the evaluation result is greater than or equal to a preset score threshold. The higher the score corresponding to the evaluation result, the higher the accuracy of the recommendation of the initial product configuration. The lower the score corresponding to the evaluation result, the lower the accuracy of the recommendation of the initial product configuration.

[0109] In some embodiments, if the evaluation results meet preset conditions, the initial product configuration result is determined as the target product configuration result and output. A user may send a request to the product configuration recommendation system to generate a recommended configuration. Upon receiving the request, the product configuration recommendation system may perform a preset action in response to the request. For example, it may generate a list corresponding to the target product configuration.

[0110] For example, Figure 3 As shown, the self-learning engine evaluates the initial product configuration result based on historical configuration data to obtain an evaluation result. When the evaluation result meets the preset conditions, the initial product configuration result is determined as the target product configuration result and output. At this time, if the product configuration recommendation system receives a request to generate a recommended configuration, the product configuration recommendation system can obtain real-time information (such as promotion, control, etc.) corresponding to each target product in the target product configuration in response to the request to generate a recommended configuration, and generate a recommended configuration based on the acquired real-time information. The recommended configuration includes a bill of materials and a product description. In some application scenarios, the user can manually optimize the recommended configuration (such as replacing hardware models, adjusting parameters). After the user completes the manual adjustment, the hardware catalog can be generated based on the final recommended configuration in response to the generation operation input by the user. Among them, the product configuration recommendation system can also support saving configurations.

[0111] In some embodiments, when the evaluation result does not meet the preset conditions, the product configuration recommendation method provided by the present application may further include: inputting the evaluation result into the configuration recommendation engine through the self-learning engine, so that the configuration recommendation engine is iteratively trained based on the evaluation result to obtain a new configuration recommendation engine. The new configuration recommendation engine is used to determine a new initial product configuration result based on the target user needs, as well as the preset recommendation scenario and the preset configuration rules. The self-learning engine is used to evaluate the new initial product configuration result based on the historical configuration data to obtain a new evaluation result, and when the new evaluation result meets the preset conditions, the new initial product configuration result is determined as the target product configuration result.

[0112] In some embodiments, if the evaluation result does not meet the preset conditions, the self-learning engine inputs the evaluation result into the configuration recommendation engine. After the configuration recommendation engine receives the evaluation result input by the self-learning engine, it will perform iterative training based on the received evaluation result to obtain a new configuration recommendation engine. Thereafter, the new configuration recommendation engine will re-recommend product configurations based on the target user needs to obtain a new initial product configuration result. If the evaluation result corresponding to the new initial product configuration result meets the preset conditions, the new initial product configuration result will be determined as the target product configuration and output. If the evaluation result corresponding to the new initial product configuration result does not meet the preset conditions, the evaluation result corresponding to the new initial product configuration result will be input into the new configuration recommendation engine for a new round of iterative training.

[0113] Exemplarily, after the configuration recommendation engine receives an evaluation result from the self-learning engine that does not meet the preset conditions, it can construct a negative sample based on the evaluation result and perform iterative training based on the negative sample to adjust the model parameters in the configuration recommendation engine to obtain a new configuration recommendation engine.

[0114] In some embodiments, a large language model (such as an LLM) module and a retrieval-augmented generation (RAG) module are deployed in the product configuration recommendation system. The large language model module and the retrieval-augmented generation module can be deployed on the self-learning engine or on the configuration recommendation module. This embodiment of the present application is not limited to this.

[0115] In some embodiments, when the evaluation results do not meet the preset conditions, the evaluation results can be structured by the large language model module and the retrieval enhancement generation module to obtain the corresponding structured query language (SQL), so as to configure the recommendation engine to perform iterative training based on the evaluation results.

[0116] It is understood that if the evaluation result corresponding to the initial product configuration result does not meet the preset conditions, it means that the initial product configuration result does not meet the user's needs. In this case, the evaluation result corresponding to the initial product configuration result is reversely input into the configuration recommendation engine, so that the configuration recommendation engine is iteratively trained based on the evaluation result to obtain a new configuration recommendation engine. The new configuration recommendation engine then re-produces product configuration recommendations until the initial product configuration result meets the preset conditions. In this way, iterative training of the configuration recommendation engine based on the evaluation results can make the product configuration results recommended by the configuration recommendation engine increasingly accurate, thereby improving the accuracy of product configuration recommendation results.

[0117] In some embodiments, as Figure 7As shown, the self-learning engine evaluates the initial product configuration result based on historical configuration data, and the evaluation result obtained may include S701-S702.

[0118] S701: Determine, through a self-learning engine, from historical configuration data, first historical configuration data whose similarity to an initial product configuration result is higher than a similarity threshold.

[0119] In some embodiments, the similarity threshold is a value pre-configured in the product configuration recommendation system. The similarity threshold can be set according to actual needs. This embodiment of the present application is not limited to this.

[0120] S702: Evaluate the initial configuration result based on the first historical configuration data through a self-learning engine to obtain an evaluation result.

[0121] In some embodiments, after receiving the initial product configuration result, the self-learning engine may determine the similarity between the initial product configuration result and the historical configuration data. Based on the similarity between the initial product configuration result and the historical configuration data, the self-learning engine may determine at least one first historical configuration data having a similarity to the initial product configuration result that exceeds a similarity threshold. The self-learning engine may then evaluate the initial configuration result based on the first historical configuration data to obtain an evaluation result. The historical configuration data may include the first historical configuration data.

[0122] For example, after receiving the initial product configuration result, the self-learning engine may determine the similarity between the initial product configuration result and the historical configuration data based on the product configuration information corresponding to the historical configuration data, and identify at least one product configuration information whose similarity to the initial product configuration result exceeds a similarity threshold. The self-learning engine may then evaluate the initial configuration result based on the product configuration information corresponding to the first historical configuration data, the feedback information, and the transaction information to obtain an evaluation result.

[0123] In some embodiments, the self-learning engine can evaluate the initial configuration results from dimensions such as performance matching, user satisfaction, and transaction probability based on the first historical configuration data to obtain corresponding evaluation results.

[0124] Exemplarily, the self-learning engine can compare the performance (such as running score data) between the first historical configuration data and the initial configuration result based on the product configuration information of at least one first historical configuration data to evaluate the performance matching degree of the initial configuration result; it can evaluate the user satisfaction of the initial configuration result based on the evaluation information (such as the satisfaction rate) of at least one first historical configuration data; and it can evaluate the transaction probability of the initial configuration result based on the transaction information (such as the transaction rate) of at least one first historical configuration data.

[0125] It should be noted that the evaluation dimensions of the self-learning engine for the initial product configuration can be set according to actual needs, and the embodiments of the present application do not limit this.

[0126] Figure 8 A flowchart of another product configuration recommendation method provided in an embodiment of the present application is shown as follows: Figure 8 As shown, an embodiment of the present application also provides a product configuration recommendation method, which includes three parts: original demand processing, product configuration and configuration confirmation.

[0127] In some embodiments, the original demand processing includes: first obtaining the original configuration demand of the multimodal. The original configuration demand of the multimodal means that the form of the original configuration demand can be pictures, voice, text, documents, etc. After obtaining the original configuration demand of the multimodal, semantic services can be performed on the original configuration demand (for example, guiding the user to gradually refine the demand in a guiding manner) to obtain the initial user demand. Thereafter, the initial user demand is input into the multimodal information processing large model (AI-Generated Content, AIGC), that is, the information processing engine. The data input into the multimodal information processing large model together with the target user demand can also include at least one of product specifications, customer data, inventory data, and contract data. The AIGC large model processes the initial user demand to extract data such as product hardware and software, compatibility, reliability indicators, etc., and integrates product specifications / customer historical transaction records, inventory and other information to obtain structured data corresponding to the original configuration demand (i.e., target user demand).

[0128] In some embodiments, product configuration includes a configuration model (i.e., a configuration recommendation engine) that receives structured data corresponding to the original configuration requirements sent by the AIGC large model, and based on scenarios (preset recommendation scenarios), customer information, different product combination combinations, transaction configuration information, inventory and other information, obtains an algorithm model of the relationship between market volume, customers, different product configuration combinations or different configurations of the same product and transactions, and constructs an output configuration result (i.e., the initial configuration result).

[0129] In some embodiments, configuration confirmation includes a configuration evaluation model receiving a configuration result sent by a configuration model, and evaluating the configuration result from dimensions such as configuration result evaluation, feedback, and transaction based on historical transaction data in the case library to obtain an evaluation result. When the evaluation result corresponding to the configuration result is unreliable, the configuration result is intercepted, and the evaluation result is reversely input into the configuration model, so that the configuration model performs iterative learning based on the evaluation result. When the evaluation result corresponding to the configuration result is credible, the configuration result is determined as a configuration recommendation result (target product configuration result) and output. After outputting the configuration recommendation result, the transaction data of the configuration recommendation result is obtained, and the transaction data and the configuration recommendation result are stored in the case library.

[0130] Corresponding to the aforementioned embodiment of the product configuration recommendation method, the present application also provides an embodiment of a product configuration system. Figure 9 This is a schematic diagram of a product configuration recommendation system provided in an embodiment of the present application. Figure 9 As shown, the product configuration recommendation system 900 includes an information processing engine 910, a configuration recommendation engine 920 and a self-learning engine 930; wherein,

[0131] The information processing engine 910 is configured to: process the user's original configuration requirements to obtain target user requirements;

[0132] The configuration recommendation engine 920 is configured to: determine multiple target products from a preset product library based on the target user's needs, preset recommendation scenarios, preset configuration rules, and preset knowledge graph, and combine the multiple target products to obtain an initial product configuration result;

[0133] The self-learning engine 930 is configured to: evaluate the initial product configuration result based on historical configuration data to obtain an evaluation result; if the evaluation result meets a preset condition, determine the initial product configuration result as the target product configuration result and output it; wherein the preset condition is used to indicate the recommendation accuracy of the initial product configuration

[0134] In some embodiments, the self-learning engine 930 is configured to: determine first historical configuration data from the historical configuration data whose similarity with the initial product configuration result is higher than a similarity threshold; and evaluate the initial configuration result based on the first historical configuration data to obtain the evaluation result.

[0135] In some embodiments, the self-learning engine 930 is configured to: if the evaluation result does not meet the preset conditions, the self-learning engine 930 is configured to: input the evaluation result into the configuration recommendation engine, so that the configuration recommendation engine performs iterative training based on the evaluation result to obtain a new configuration recommendation engine;

[0136] The new configuration recommendation engine is configured to: determine a new initial product configuration result based on target user needs, preset recommendation scenarios, and preset configuration rules;

[0137] The self-learning engine 930 is further configured to: evaluate the new initial product configuration result based on historical configuration data to obtain a new evaluation result, and determine the new initial product configuration result as the target product configuration result if the new evaluation result meets the preset conditions.

[0138] In some embodiments, the information processing engine 910 is configured to: perform demand analysis on the original configuration requirements to obtain initial user requirements; and perform error correction on the initial user requirements based on a preset industry knowledge base and a preset prompt word project to obtain target user requirements.

[0139] In some embodiments, the multiple target products include a first product, a second product, a third product, a fourth product and a fifth product; wherein the first product, the second product, the third product, the fourth product and the fifth product are respectively different components in the computing device; the configuration recommendation engine 920 is configured to: determine the first product, the second product and the third product from the preset product library based on the target user needs, the preset recommendation scenarios, the preset configuration rules and the preset knowledge graph; determine the fourth product from the preset product library based on the second product, the preset recommendation scenarios, the preset configuration rules and the preset knowledge graph; determine the fifth product based on the preset recommendation scenarios, the preset configuration rules and the preset knowledge graph, as well as the first product, the second product, the third product and the fourth product.

[0140] In some embodiments, the configuration recommendation engine 920 is configured to: decompose the target user needs to obtain target recommendation prompt words corresponding to multiple target products; based on the target recommendation prompt words corresponding to multiple target products, as well as preset recommendation scenarios, preset configuration rules and preset knowledge graphs, determine the first product, the second product and the third product from the preset product library respectively.

[0141] In some embodiments, the configuration recommendation engine 920 is configured to: determine multiple candidate products from a preset product library based on the second product, as well as preset recommendation scenarios, preset configuration rules and preset knowledge graphs; and determine a fourth product from multiple candidate products based on the redundancy of each candidate product relative to the second product.

[0142] In some embodiments, the configuration recommendation engine 920 is configured to: determine, by configuring the recommendation engine, a candidate product among the plurality of candidate products whose redundancy with respect to the second product is less than a preset redundancy threshold as the fourth product.

[0143] Figure 10 A schematic diagram of a computing device provided in an embodiment of the present application. In some embodiments, the computing device may be a server, a terminal device, or the like. Server types may include cabinet servers, rack servers, high-density servers, graphics processing unit (GPU) servers, tower servers, blade servers, artificial intelligence (AI) servers, and the like. The present embodiment of the application does not limit the server type.

[0144] In some embodiments, a computing device includes one or more processors and a memory. The memory is configured to store one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors implement the product configuration recommendation method of the above-described embodiment.

[0145] like Figure 10 As shown, the computing device 1000 includes a processor 1001 and a memory 1002. By way of example, the computing device 1000 may further include a communication interface 1003 and a communication bus 1004.

[0146] The processor 1001, the memory 1002 and the communication interface 1003 communicate with each other via a communication bus 1004. The communication interface 1003 is used to communicate with other devices such as a client or other server network elements.

[0147] In some embodiments, the processor 1001 is configured to execute a program 1005, specifically, to execute the relevant steps in the above-mentioned product configuration recommendation method embodiment. Specifically, the program 1005 may include program code, which includes computer-executable instructions.

[0148] For example, the processor 1001 may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement some embodiments of the present application. The computing device 1000 may include one or more processors of the same type, such as one or more CPUs, or different types of processors, such as one or more CPUs and one or more ASICs.

[0149] In some embodiments, the memory 1002 is used to store the program 1005. The memory 1002 may include a high-speed RAM memory, and may also include a non-volatile memory (NVM), such as at least one disk storage.

[0150] Program 1005 can be specifically called by processor 1001 to enable computing device 1000 to perform product configuration recommendation operations.

[0151] An embodiment of the present application provides a computer-readable storage medium, which stores at least one executable instruction. When the executable instruction is executed on a computing device 1000, the computing device 1000 executes the product configuration recommendation method in the above embodiment.

[0152] The executable instructions can be specifically used to enable the computing device 1000 to perform product configuration recommendation method operations.

[0153] For example, the computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a compact disc read-only memory (CD-ROM), a magnetic tape, a floppy disk, an optical data storage device, and the like.

[0154] The beneficial effects that can be achieved by the readable storage medium provided in the embodiment of the present application can be referred to the beneficial effects in the corresponding product configuration recommendation method provided above, and will not be repeated here.

[0155] It should be noted that, in the application, relational terms such as first and second, etc. are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or apparatus comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a..." does not exclude the presence of other identical elements in the process, method, article, or apparatus comprising the element.

[0156] Each embodiment in this specification is described in a related manner. Similar portions between the embodiments can be referred to in conjunction with each other. Each embodiment focuses on the differences from other embodiments. In particular, the device embodiments are generally similar to the method embodiments, so their description is relatively simple. For related portions, refer to the description of the method embodiments.

[0157] The logic and / or steps represented in the flowchart or otherwise described herein may be considered, for example, as an ordered list of executable instructions for implementing logical functions, and may be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device).

[0158] For the purposes of this specification, a "computer-readable medium" can be any apparatus that can contain, store, communicate, propagate, or transport the program for use by or in connection with an instruction execution system, apparatus, or device.

[0159] More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection having one or more wires (electronic device), a portable computer disk cartridge (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and programmable read-only memory (EPROM or flash memory), fiber optic device, and a portable compact disc read-only memory (CDROM).

[0160] In addition, the computer readable medium can even be paper or other suitable medium on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other medium, and then editing, interpreting or processing it in other suitable ways as necessary, and then storing it in the computer memory. It should be understood that various parts of the present application can be implemented in hardware, software, firmware or a combination thereof.

[0161] In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used: a discrete logic circuit having logic gate circuits for implementing logic functions on data signals, an application-specific integrated circuit having suitable combinational logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0162] The implementation methods described above are only specific implementation methods of the present application and are not intended to limit the scope of protection of the present application. Any modifications, equivalent replacements, improvements, etc. made on the basis of the technical solution of the present application should be included in the scope of protection of the present application.

Claims

1. A product configuration recommendation method, characterized in that: Applied to a product configuration recommendation system, the product configuration recommendation system includes an information processing engine, a configuration recommendation engine, and a self-learning engine; The method comprises: The information processing engine processes the user's original configuration requirements to obtain target user requirements; The configuration recommendation engine determines multiple target products from a preset product library based on the target user's needs, preset recommendation scenarios, preset configuration rules, and preset knowledge graph, and combines the multiple target products to obtain an initial product configuration result; The self-learning engine evaluates the initial product configuration result based on historical configuration data to obtain an evaluation result; when the evaluation result meets a preset condition, the initial product configuration result is determined as the target product configuration result and output; wherein the preset condition is used to indicate the recommendation accuracy of the initial product configuration.

2. The method according to claim 1, characterized in that The self-learning engine evaluates the initial product configuration result based on historical configuration data to obtain an evaluation result, including: Determining, by the self-learning engine, from the historical configuration data, first historical configuration data having a similarity with the initial product configuration result that is higher than a similarity threshold; The self-learning engine evaluates the initial configuration result based on the first historical configuration data to obtain the evaluation result.

3. The method according to claim 1, characterized in that The method further comprises: If the evaluation result does not meet the preset condition, inputting the evaluation result into the configuration recommendation engine through the self-learning engine, so that the configuration recommendation engine performs iterative training based on the evaluation result to obtain a new configuration recommendation engine; Determining a new initial product configuration result based on the target user's needs, the preset recommendation scenario, and the preset configuration rules through the new configuration recommendation engine; The self-learning engine evaluates the new initial product configuration result based on the historical configuration data to obtain a new evaluation result, and when the new evaluation result meets the preset conditions, the new initial product configuration result is determined as the target product configuration result.

4. The method according to claim 1, wherein The processing of the original configuration requirements by the information processing engine to obtain target user requirements includes: Performing demand analysis on the original configuration requirements by the information processing engine to obtain initial user requirements; The information processing engine performs error correction processing on the initial user needs based on a preset industry knowledge base and a preset prompt word project to obtain the target user needs.

5. The method according to claim 1, wherein The plurality of target products include a first product, a second product, a third product, a fourth product, and a fifth product; wherein the first product, the second product, the third product, the fourth product, and the fifth product are different components in a computing device; The configuration recommendation engine determines multiple target products from a preset product library based on the target user's needs, preset recommendation scenarios, preset configuration rules, and preset knowledge graph, including: Determining, by the configuration recommendation engine, the first product, the second product, and the third product from the preset product library based on the target user needs, the preset recommendation scenario, the preset configuration rules, and the preset knowledge graph; Determining, by the configuration recommendation engine, the fourth product from the preset product library based on the second product, the preset recommendation scenario, the preset configuration rule, and the preset knowledge graph; The fifth product is determined by the configuration recommendation engine based on the preset recommendation scenario, the preset configuration rules and the preset knowledge graph, as well as the first product, the second product, the third product and the fourth product.

6. The method according to claim 5, characterized in that The determining, by the configuration recommendation engine based on the target user needs, the preset recommendation scenario, the preset configuration rules, and the preset knowledge graph, the first product, the second product, and the third product from the preset product library includes: Decomposing the target user's needs by the configuration recommendation engine to obtain target recommendation prompt words corresponding to each of the multiple target products; The configuration recommendation engine determines the first product, the second product, and the third product from the preset product library based on the target recommendation prompt words corresponding to each of the multiple target products, the preset recommendation scenario, the preset configuration rules, and the preset knowledge graph.

7. The method according to claim 5, characterized in that The determining, by the configuration recommendation engine, the fourth product from the preset product library based on the second product, the preset recommendation scenario, the preset configuration rule, and the preset knowledge graph includes: Determining, by the configuration recommendation engine, a plurality of candidate products from the preset product library based on the second product, the preset recommendation scenario, the preset configuration rules, and the preset knowledge graph; The configuration recommendation engine determines the fourth product from the plurality of candidate products based on the redundancy of each candidate product relative to the second product.

8. The method according to claim 7, characterized in that Determining the fourth product from the plurality of candidate products based on the redundancy of each candidate product relative to the second product by the configuration recommendation engine includes: The configuration recommendation engine determines, among the multiple candidate products, a candidate product whose redundancy with respect to the second product is less than a preset redundancy threshold as the fourth product.

9. A product configuration recommendation system, characterized in that: Includes information processing engine, configuration recommendation engine and self-learning engine modules; The information processing engine is configured to: process the user's original configuration requirements to obtain target user requirements; The configuration recommendation engine is configured to: determine multiple target products from a preset product library based on the target user's needs, preset recommendation scenarios, preset configuration rules, and preset knowledge graph, and combine the multiple target products to obtain an initial product configuration result; The self-learning engine is configured to: evaluate the initial product configuration result based on historical configuration data to obtain an evaluation result; if the evaluation result meets a preset condition, determine the initial product configuration result as the target product configuration result and output it; wherein the preset condition is used to indicate the recommendation accuracy of the initial product configuration.

10. A computing device, characterized in that include: one or more processors; and a memory configured to: store one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 8.

Images