Service plan recommendation method, device, equipment and storage medium

Abstract

The application discloses a service scheme recommendation method and device, equipment and a storage medium, wherein the method comprises: acquiring the computing network resource demand information of a user; determining the service intention of the user based on the computing network resource demand information and a plurality of preset business scene matching; calculating the computing power index of all cloud resource pools according to a computing power index measurement system, and performing computing network resource search matching on the service intention and the computing power index to obtain a resource search matching result; calculating a plurality of optimal recommendation strategies for opening a requested service according to the resource search matching result and preset reference data, and determining the corresponding computing network resource recommendation party according to the plurality of optimal recommendation strategies. The application effectively solves the problem that, due to the fact that the allocation of computing network resources needs to pay attention to many points, the operation personnel is difficult to fully pay attention to these information, and thus the resources allocated to the customers have problems such as unreasonable configuration of each cloud resource node and unbalanced load.

Description

Technical Field

[0001] This invention relates to the fields of cloud computing and edge computing technologies, and in particular to service solution recommendation methods, apparatus, devices, and storage media. Background Technology

[0002] As the cornerstone and new engine of operators' digital transformation in the digital economy era, computing power has seen China Mobile, China Telecom, and China Unicom all release white papers on computing power networks. These white papers aim to build ubiquitous and converged computing power networks, creating a single-point access, on-demand, and ubiquitous computing power service capability. A computing power network is a new type of information infrastructure centered on computing power and based on the network, deeply integrating multiple elements such as network, cloud, data, intelligence, security, edge, terminal, and blockchain to provide integrated services. The goal of the computing power network is to achieve ubiquitous computing power, symbiotic computing and network, intelligent orchestration, and integrated services, gradually promoting computing power to become a social-level service, like water and electricity, accessible with a single point of contact and readily available, achieving the vision of "networks reaching everywhere, computing power ubiquitous, and intelligence pervasive." Simply put, computing power is computational ability, referring to data processing capabilities. From small mobile phones and laptops to large supercomputers, computing power exists in various intelligent hardware devices; without computing power, there would be no normal application of various software and hardware. Artificial intelligence is not without its source or foundation. Every AI-powered facial recognition and speech-to-text conversion requires the computing power of hardware chips. In existing integrated computing and network technologies, customers primarily submit computing and network resource requests in two ways: one is that the computing and network resource operator provides a list of computing and network products for the customer to choose from, leading to difficulties in product selection and comparison efficiency; the other is that the computing and network resource operator recommends resources to the customer. After the customer submits their request, the account manager needs to determine if the corresponding cloud resources are available to meet the customer's needs. Currently, this requires the account manager to contact the backend computing and network operations personnel for confirmation, which is time-consuming and results in a poor customer experience.

[0003] Currently, the requirements for government and enterprise customers are based on specific computing network resource needs, which are then allocated by backend operations personnel. Since there are many factors to consider when allocating computing network resources, such as resource utilization, resource type, resource capacity, resource load, and distance between resources and customers, it is difficult for operations personnel to fully monitor all of these factors. This leads to problems such as unreasonable configuration of cloud resource nodes and uneven load (allocation rate, utilization rate, etc.) when allocating resources to customers. Summary of the Invention

[0004] The present invention aims to at least partially solve one of the technical problems in the related art.

[0005] Therefore, the first objective of this invention is to propose a service solution recommendation method that can provide customers with a solution preview before placing an order, and can also perform intention-based intelligent recommendation of computing network solutions, accurately match customer needs, and give customers more choices and faster selection feedback results.

[0006] The second objective of this invention is to provide a service recommendation device.

[0007] The third objective of this invention is to provide a computer device.

[0008] A fourth objective of this invention is to provide a non-transitory computer-readable storage medium.

[0009] To achieve the above objectives, a first aspect of the present invention provides a service recommendation method, comprising:

[0010] Obtain information on users' computing network resource requirements;

[0011] The user's business intent is determined based on the network resource demand information and the matching of multiple preset business scenarios.

[0012] The computing power metrics of all cloud resource pools are calculated according to the computing power metric measurement system, and the business intent is matched with the computing power metrics to obtain the resource search matching result.

[0013] Based on the resource search and matching results and the preset benchmark data, multiple optimal recommendation strategies for requesting service activation are calculated, and corresponding network resource recommendation schemes are determined based on the multiple optimal recommendation strategies.

[0014] The service scheme recommendation method according to embodiments of the present invention may also have the following additional technical features:

[0015] In one embodiment of the present invention, after determining the corresponding network resource recommendation scheme according to the multiple optimal recommendation strategies, the method further includes:

[0016] Obtain the resource parameters of the proposed computing network resource recommendation scheme;

[0017] A solution map is constructed based on the solution resource parameters, and the solution map is visualized to obtain a visualization result based on the cloud resource pool utilization rate.

[0018] In one embodiment of the present invention, the various business scenarios correspond to various computing network application products.

[0019] In one embodiment of the present invention, the user's computing network resource requirement information is obtained through a service terminal input interface or a human-computer interaction interface; wherein, obtaining the user's computing network resource requirement information through a service terminal input interface includes:

[0020] The parameters and values ​​of the computing network application product are used as the parameters and values ​​corresponding to the menu options in the business terminal input interface to obtain the customer's computing network resource requirements information.

[0021] In one embodiment of the present invention, obtaining a user's computing network resource requirements information through a human-computer interaction interface includes:

[0022] Voice recognition converts customer speech input into text for the human-computer interaction interface.

[0023] Intent recognition categorizes the intent of text and directs it to the corresponding domain;

[0024] By obtaining entity relationship information or word vectors of text in the corresponding domain through semantic recognition, and then processing them with the corresponding business logic, a speech synthesis result containing information on computing network resource requirements is obtained through speech synthesis.

[0025] In one embodiment of the present invention, after obtaining the user's computing network resource demand information, the method further includes:

[0026] Obtain the business metrics corresponding to the computing network resource demand information;

[0027] The business metrics are converted into computing network resource metrics according to the preset conversion rules.

[0028] In one embodiment of the present invention, determining the user's business intent based on the computing network resource demand information and a preset combination of multiple business scenarios includes:

[0029] The business scenario is determined to be cloud desktop;

[0030] The network resource requirements information is matched with all network capability templates corresponding to the cloud desktop, and the optimal network capability template is obtained based on the matching results.

[0031] The business intent of using the computing network application products associated with the optimal computing network capability template as the computing network resource requirement information.

[0032] In one embodiment of the present invention, a preset model function is used to map different types of computing power resources in the computing network to a unified dimension of measurement in order to establish the computing power index measurement system.

[0033] In one embodiment of the present invention, the computing network includes various types of computing resources such as CPU, GPU, NPU and FPGA.

[0034] In one embodiment of the present invention, the multiple optimal recommendation strategies include a comprehensive optimal recommendation strategy, a computing power optimal recommendation strategy, and a network optimal recommendation strategy.

[0035] In one embodiment of the present invention, the resource search matching results include candidate cloud resource pools; visualizing the solution map to obtain visualization results based on cloud resource pool utilization includes:

[0036] Based on the candidate cloud resource pool, a multi-layer latency circle based on the latency dimension is obtained;

[0037] The network connects the user's location with the resource pool recommended by the solution map, and the computing power distribution viewing result is obtained based on the network connection result and the computing power distribution of the multi-layer latency circle.

[0038] Based on the computing power distribution, the visualization results show the city and network domain of the user's solution map, and the capabilities of each resource pool are highlighted with tags to obtain the visualization results.

[0039] To achieve the above objectives, a second aspect of the present invention provides a service recommendation device, comprising:

[0040] The demand information acquisition module is used to acquire users' computing network resource demand information;

[0041] The business intent determination module is used to determine the user's business intent based on the computing network resource demand information and a variety of preset business scenarios.

[0042] The resource search and matching module is used to calculate the computing power indicators of all cloud resource pools according to the computing power indicator measurement system, and to perform computing network resource search and matching between the business intent and the computing power indicators to obtain the resource search and matching results.

[0043] The computing network resource recommendation module is used to calculate multiple optimal recommendation strategies for requesting service activation based on the resource search and matching results and preset benchmark data, and to determine the corresponding computing network resource recommendation scheme based on the multiple optimal recommendation strategies.

[0044] The transmission line resource identification and detection method and apparatus of this invention utilizes a Swing Transformer structure. During transmission line resource detection, the model employs a sliding window attention mechanism, enabling the backbone network to more effectively identify which regions in the image are worthy of focus for transmission line resource detection. Furthermore, the use of a probabilistic sparse attention mechanism allows the model to pay more attention to foreground information resources in the transmission line resource detection scenario, thus highlighting foreground information.

[0045] To achieve the above objectives, a third aspect of this application provides a computer device, including a processor and a memory; wherein the processor reads executable program code stored in the memory to run a program corresponding to the executable program code, for implementing the service scheme recommendation method as described in the first aspect embodiment.

[0046] To achieve the above objectives, a fourth aspect of this application provides a non-transitory computer-readable storage medium having a computer program stored thereon that, when executed by a processor, implements the service scheme recommendation method as described in the first aspect embodiment.

[0047] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0048] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0049] Figure 1 A flowchart illustrating a service scheme recommendation method according to an embodiment of the present invention;

[0050] Figure 2 This is an architecture diagram of a service solution recommendation method according to an embodiment of the present invention;

[0051] Figure 3 This is a schematic diagram of the layered structure of a computing network application product related to a service according to an embodiment of the present invention;

[0052] Figure 4 This is a schematic diagram illustrating the encapsulation process of computing network capabilities related to services according to an embodiment of the present invention;

[0053] Figure 5 This is a schematic diagram illustrating the process of obtaining customer network resource requirements through human-computer dialogue in a service according to an embodiment of the present invention;

[0054] Figure 6 This is a schematic diagram of a heterogeneous computing power tiered index system related to the services provided according to an embodiment of the present invention;

[0055] Figure 7 A structural diagram of a service scheme recommendation device according to an embodiment of the present invention;

[0056] Figure 8 This is a schematic diagram of the structure of a computer device according to an embodiment of the present invention. Detailed Implementation

[0057] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0058] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0059] The following description, with reference to the accompanying drawings, outlines a service scheme recommendation method, apparatus, computer device, and storage medium based on embodiments of the present invention.

[0060] First, the technical terms used in this invention will be explained.

[0061] Computing power networks are a new type of information infrastructure that takes computing as its core, networks as its foundation, and deeply integrates multiple elements such as networks, cloud, data, intelligence, security, edge, terminals, and blockchain to provide integrated services.

[0062] The integrated computing and networking service includes: cloud resource services and network services;

[0063] Cloud resource services refer to cloud computing power support services based on cloud servers and cloud disks;

[0064] Network services refer to the network that connects user terminals to the cloud resource pool, providing network connectivity services.

[0065] Figure 1 This is a flowchart of a service scheme recommendation method according to an embodiment of the present invention.

[0066] like Figure 1 As shown, the method includes, but is not limited to, the following steps:

[0067] S1, obtain the user's computing network resource requirements;

[0068] S2 determines the user's business intent based on the information on computing network resource requirements and the matching of multiple preset business scenarios;

[0069] S3 calculates the computing power indicators of all cloud resource pools according to the computing power indicator measurement system, and performs computing network resource search matching between business intent and computing power indicators to obtain resource search matching results.

[0070] S4. Based on the resource search and matching results and the preset benchmark data, calculate multiple optimal recommendation strategies for requesting service activation, and determine the corresponding network resource recommendation scheme based on the multiple optimal recommendation strategies.

[0071] The service solution recommendation method of this invention involves the system acquiring information provided by the customer to determine the customer's computing network resource requirements. The system then searches for resources to match the customer's requirements based on data such as the capacity and utilization rate of the cloud resource pool. Based on the resource search results, multiple solutions are provided, prioritizing those that meet the user's minimum requirements: lowest cost, highest performance, best security, etc. Each solution includes network information (access method, measurement indicators) and a cloud resource pool (resource type, resource specifications). Each solution provides a map presentation, with the resource pool displayed intuitively based on its utilization rate.

[0072] Therefore, in practical application scenarios, when account managers recommend industry solutions to target customers or customers consult about industry solutions, customers only need to briefly describe their scenario requirements, and this invention can quickly match multiple scenario solutions, providing a visual presentation of cloud resource distribution and network access, allowing customers to intuitively understand and choose, thereby accurately matching customer needs, helping account managers expand their business, and guiding customers to use the cloud whenever possible.

[0073] Figure 2 This is a schematic diagram of the technical architecture of the integrated computing and network service of the present invention, as shown below. Figure 2 As shown, from bottom to top, they are: computing infrastructure layer, orchestration management layer, operation service layer, and application layer.

[0074] It is understood that the computing power network of this invention is a new type of information infrastructure that is centered on computing, based on the network, and deeply integrates network, cloud, data, intelligence, security, edge, terminal, and blockchain (ABCD-NETS) to provide integrated services. Compared with existing computing power networks, the computing power network of this invention has the following major changes:

[0075] First, there is a change in the form of computing power, from the original single computing power to the current diversified computing power cores, ubiquitous distribution, and three-dimensional products;

[0076] Second, the relationship between computing and the network has changed. Originally, the network provided computing power connection, but now it also provides sensing computing power and carrying computing power. The network is in computing and computing is in the network.

[0077] Third, the management model has changed, enabling full-service support for computing and network convergence, resource integration and scheduling across all domains, and intelligent orchestration and management;

[0078] Fourth, there is a change in service models, shifting from resource-based to task-based, and from one-stop to integrated services.

[0079] In one embodiment of the present invention, before displaying the corresponding resources to the customer, the system needs to obtain the customer's resource requirements and then search for cloud resources that match the customer's requirements. The presentation format of the cloud resource requirements information submitted by the customer is not limited; it can be in text form, in voice form, or through human-computer interaction.

[0080] Specifically, the system can obtain information by providing an input interface (either a client-side input interface or a business-side input interface) and obtaining the customer's computing network resource requirements through preset input options, or by providing a human-computer dialogue interface through which the customer submits computing network resource requirements.

[0081] For example, through a business input interface, an account manager communicates with a client via telephone or other means to obtain the client's computing network resource requirements, and then fills in the corresponding options in the input interface. For instance, the client first inputs the business scenario, and the system provides a selection of multiple industry and type business scenarios, categorized by industry, including: health, transportation, culture and tourism, computing network integration, agriculture, industry, education, finance, etc.

[0082] For example, through another business-side input interface. The account manager can... Figure 2 After entering the corresponding business scenario options on the interface shown, the system automatically renders other requirements for that business scenario, such as access method, user scenario, bandwidth, etc., based on the selected business scenario. Then, the account manager continues to communicate with the customer to improve other content and finally obtain the network resource requirement information containing the customer's detailed needs.

[0083] In one implementation of the present invention, in order to facilitate the operation of account managers and accurately match the business intentions of customers, the present invention can pre-set multiple business scenarios, wherein each business scenario corresponds to a computing network application product.

[0084] Figure 3 For the layered structure of computing network application products, such as Figure 3 As shown, the bottom layer represents the physical primitive atomic capabilities, which are encapsulated into basic units; above that, they are encapsulated into computing power domains; and at the very top are computing power scenarios, which can be sold to customers.

[0085] It is understandable that computing resources in a computing power network include central processing units (CPUs) under general-purpose server architectures, graphics processing units (GPUs) specifically designed for processing data of uniform types such as graphics and images, neural network processors (NPUs) and tensor processors (TPUs) for accelerating neural networks, CPUs widely used in edge-side embedded devices, and semi-custom processors such as field-programmable gate arrays (FPGAs). From a business perspective, based on the different algorithms the processors run and the types of data computation involved, computing power can be categorized into computing power that provides logical operations, computing power that provides parallel computing, and computing power that accelerates neural networks. Furthermore, for business operations, not only sufficient computing power is required, but also supporting storage capacity, network capabilities, and may even need encoding / decoding capabilities and throughput capacity to jointly ensure a good user experience. It is recommended to measure computing power from a microservices perspective, standardize the corresponding resource scheduling and allocation principles, reduce the complexity of business and application deployment in the computing power network, and simplify business management processes and mechanisms.

[0086] Figure 4 The process of encapsulating computing network capabilities, such as Figure 4 As shown, it includes: atomic capabilities -> basic computing power units -> computing power domains -> scenario domains, which ultimately form computing network application products. Each computing network application product corresponds to a cloud access solution.

[0087] In one embodiment of the present invention, the computing network application product is a high-performance cloud desktop, and the computing network capability template corresponding to the computing network application product encapsulates the following content:

[0088] Template Name: High-Performance Cloud Desktop

[0089] Template Description: Platform Template

[0090] Related computing power template: Cloud desktop computing power template

[0091] Related products and attributes:

[0092] 5G private network (range, latency, capacity...)

[0093] CPU (number of cores, number of memory...)

[0094] Cloud dedicated line (bandwidth, access method...)

[0095] Parameter information:

[0096] Scope (province-wide, prefecture-level city, region...)

[0097] Latency (<20ms, <10ms...)

[0098] The computing power template encapsulates the following:

[0099] Capabilities: High performance, medium performance, general performance, etc.

[0100] Performance: number of cores, number of memory modules, external storage, bandwidth, etc.;

[0101] Services include: assurance, best-effort service, and flexible migration.

[0102] Types: Cloud desktop packages include the following three types of computing power templates:

[0103] High-performance cloud desktop (low latency) [Cloud desktop (16 cores 32GB, 32 cores 128GB, etc.)]

[0104] 16 cores or more, 32GB or more, <20ms response time, 50Mbps wireless / wired connection

[0105] Cloud desktop (low latency) [Cloud desktop (8 cores 16GB, 4 cores 16GB, etc.)]

[0106] 4-core to 8-core processors, less than 32GB RAM, <= 20ms response time, 50Mbps wireless / wired connection

[0107] Lightweight cloud desktop [Cloud desktops (1 core 2GB, 2 core 4GB, etc.)]

[0108] 4 cores or less, 8GB or less, 50Mbps wireless / wired

[0109] The cloud migration solutions for cloud desktop computing power products are as follows:

[0110] High-performance cloud desktop: Virtual machine + cloud desktop + PON access + campus or MEC

[0111] Cloud Desktop: Virtual Machine + Cloud Desktop + 5G Preferred Access + PON Access + Campus or MEC

[0112] Lightweight Cloud Desktop: Virtual Machine + Cloud Desktop + 4G / 5G + PON Access + Industry Cloud

[0113] Therefore, the parameters and values ​​of computing network application products can be used as the parameters and values ​​corresponding to the menu options of the human-computer interaction input interface, thereby accurately obtaining the customer's computing network resource requirements.

[0114] In one embodiment of the present invention, customer network resource requirements information can also be obtained through a human-computer interaction interface.

[0115] Specifically, speech recognition converts customer input into text through the human-computer interaction interface; intent recognition categorizes and distributes the intent of the text to the corresponding domain; semantic recognition obtains entity relationship information or word vectors of the text in the corresponding domain, and after processing with the corresponding business logic, speech synthesis results containing information on computing network resource requirements are obtained.

[0116] Figure 5 Human-computer dialogue processes, such as Figure 5 As shown, the system converts customer voice input into text through speech recognition, classifies and distributes the text intent to the corresponding domain, such as query or FAQ, through intent recognition, obtains entity relationship information or word vectors from the text through semantic recognition, and after passing through the corresponding business logic, it uses speech synthesis to provide feedback to the customer, while also displaying the customer's network resource requirements.

[0117] Furthermore, since the description of computing network resources provided by customers may differ from the description of computing network resources used by the system, it is necessary to convert the customer's requirements into metrics before conducting a search. Future computing networks often need to provide targeted services to large customers, but these customers may not be able to accurately express their needs in the language familiar to operators. Therefore, an intent engine is needed to understand customer needs and translate them into language more familiar to operators. This engine can then be used to simulate network design and specifications, breaking down customer needs into connections and breakthroughs between network resources.

[0118] In this embodiment of the invention, multiple business metrics and computing network resource metrics are set, and specific conversion rules are configured for each business metric to convert it into computing network resource metrics. For example, customer-provided business metric requirements, such as system user concurrency and system user volume, are converted into computing network resource metric requirements for unified management of the integrated computing network service, such as cloud host configuration and storage capacity. The specific conversion rules for each metric are not limited and can be set according to the actual application scenario.

[0119] Furthermore, through the above indicator conversion, the total computing power requirements for video cloud transmission and storage are calculated from the input parameters provided by the customer for the integrated computing and network service, such as the number of video access channels, QPS, access bandwidth requirements, latency requirements, and access scale.

[0120] Specifically, the client submitted the following computing network service requirements: deploy game servers in the cloud to meet immersive experience requirements, provide rendering services, ensure low latency and high bandwidth, guarantee reliability, and also provide big data computing services. The description of these computing network service requirements does not correspond to the metrics used by the system, therefore, a conversion of business metrics is necessary. The converted computing network requirements are: computing power requirements + capability requirements + network requirements. Computing power requirements include central cloud parameters (cloud host configuration) and edge cloud parameters (cloud host configuration). Capabilities include big data applications, facial recognition, etc. Network requirements include bandwidth parameters, latency parameters, isolation parameters, and primary / backup link parameters.

[0121] Furthermore, the above methods can accurately match the customer's business intent and ultimately output the matching results.

[0122] In one embodiment of the present invention, the business scenario selected by the customer is cloud desktop, and the corresponding resource range, number of CPU cores, and network latency are shown in Table 1:

[0123] Table 1

[0124] scope The whole province prefecture-level cities area Core count 4 cores or less 16 cores or less 32 cores or less Delay <20ms <40ms <100ms

[0125] By matching all computing and network capability templates corresponding to the cloud desktop, the best computing and network capability template is finally matched, and the product associated with this template is taken as the customer's computing and network resource requirements, as shown in Table 2:

[0126] Table 2

[0127] 5G private network Private network type: Dedicated | Range: Province-wide | Latency: <20ms CPU CPU: X86 | Number of cores: 16 Cloud dedicated line Bandwidth: 1T | Access method: PON access

[0128] Furthermore, given the diverse core characteristics of computing power (e.g., CPU, GPU, NPU, FPGA), its ubiquitous distribution (e.g., centralized computing power, city-level shared edge cloud), near-field computing power (city-level edge cloud), on-site computing power (campus edge cloud), and edge computing power (edge ​​gateway, monitoring terminal), and its multi-dimensional product characteristics (e.g., IaaS, PaaS, SaaS), a standardized computing power metric system can be established based on real-time sensing capabilities and dimensions such as capacity and load to accurately realize business scenarios such as computing power scheduling, allocation, and collaboration. This achieves standardized and unified capability quantification. For ubiquitous computing power resources, different types of computing power resources can be mapped to a unified dimensional dimension through model functions to establish a standardized computing power metric system and achieve standardized and unified capability quantification.

[0129] Figure 6 This is a schematic diagram of a hierarchical index system for heterogeneous computing power, such as... Figure 6 As shown, the integrated computing and network system calculates the computing power indicators of all cloud resource pools based on the aforementioned computing power resource capability quantification standards, and matches them one by one with the customer's computing and network resource requirements, outputting the matched cloud resource pools and the corresponding cloud access solutions for the cloud resource pools, which is the result of the resource search.

[0130] Understandably, the results of a typical resource search will contain a large number of candidate resources that meet the customer's needs.

[0131] In one embodiment of the present invention, based on the results of resource search and taking the minimum requirements of the customer as the benchmark, the system simulates and calculates multiple recommended solutions for the customer's request to activate services, such as comprehensive optimal, computing power optimal, network optimal, etc., and presents the specific resource parameters of the solutions.

[0132] The embodiments of the present invention are pre-configured with multiple scheme recommendation strategies to give customers more choices, preferably including the following three types of strategies:

[0133] Comprehensive optimal recommendation strategy: Under this strategy, considering factors such as cost, computing power, and network, the best computing and network resource in terms of overall performance across all dimensions is selected as the recommended solution by comparing the candidate resources.

[0134] Optimal computing power recommendation strategy: Under this strategy, only the computing power dimension is considered (such as the number of CPU or GPU cores, memory capacity, etc.). By comparing the computing power of each candidate resource, the computing network resource with the highest computing power is selected as the recommended solution of this strategy.

[0135] Network Optimal Recommendation Strategy: Under this strategy, only the network dimension (such as latency, bandwidth, network reliability, etc.) is considered. By comparing the network attributes of each candidate resource, the computing network resource with the optimal network is selected as the recommended solution.

[0136] Furthermore, the system will present recommended solutions to customers, including network information (access method, measurement metrics) and cloud resource pools (resource type, resource specifications) for each solution. Each solution provides a map presentation, and the resource pools are displayed intuitively based on usage rates.

[0137] In one embodiment of the present invention, all searched candidate resource pools are included in different latency zones (e.g., 100ms, 50ms, 20ms, 10ms, and 5ms) based on latency. Simultaneously, it supports connecting to recommended resource pools via network from the customer's location and viewing the computing power distribution across multiple latency zones from the network to the cloud pool. The link connecting the user's city and network domain is visualized, and the capabilities of each resource pool (e.g., supercomputing capabilities, security capabilities, high availability capabilities, stability capabilities, and AI capabilities) are highlighted with tags, allowing customers to intuitively view the content and make purchasing decisions for integrated computing and network services.

[0138] In one embodiment of the present invention, the recommended scheme includes network and resource pool related information, and supports both scheme adjustment and scheme selection functions. If the customer selects scheme adjustment, they can re-edit the network and resource pool in the selected scheme, such as making changes, deletions, or additions. If the customer selects a particular scheme, such as the optimal scheme, the system automatically generates an order corresponding to that scheme.

[0139] Therefore, this invention provides a computing network demand identification method based on customer intent, which allows customers to communicate accurately with service providers without needing to understand computing power technology, thereby reducing the difficulty for customers to participate in computing power services and enabling account managers to quickly get started in promoting the business.

[0140] The service solution recommendation method of this invention allows customers to quickly match multiple scenario solutions by simply describing their scenario needs. It provides a visual presentation of cloud resource distribution and network access, allowing customers to intuitively understand and choose, thereby accurately matching customer needs, assisting account managers in business expansion, guiding customers to use the cloud whenever possible, realizing pre-sales experiential marketing, giving customers more choices, and improving the success rate of business transactions.

[0141] To achieve the above embodiments, such as Figure 7 As shown, this embodiment also provides a service solution recommendation device 10, which includes: a demand information acquisition module 100, a business intent determination module 200, a resource search and matching module 300, and a computing network resource recommendation module 400;

[0142] The demand information acquisition module 100 is used to acquire users' computing network resource demand information;

[0143] The business intent determination module 200 is used to determine the user's business intent based on the computing network resource demand information and a variety of preset business scenarios.

[0144] The resource search and matching module 300 is used to calculate the computing power indicators of all cloud resource pools according to the computing power indicator measurement system, and to perform computing network resource search and matching between the business intent and the computing power indicators to obtain the resource search and matching results.

[0145] The computing network resource recommendation module 400 is used to calculate multiple optimal recommendation strategies for requesting service activation based on the resource search matching results and preset benchmark data, and to determine the corresponding computing network resource recommendation scheme based on the multiple optimal recommendation strategies.

[0146] The service solution recommendation device of this invention allows customers to quickly match multiple scenario solutions by simply describing their scenario needs. It provides a visual presentation of cloud resource distribution and network access, allowing customers to intuitively understand and choose, thereby accurately matching customer needs, assisting account managers in business expansion, guiding customers to use the cloud whenever possible, realizing pre-sales experiential marketing, giving customers more choices, and improving the success rate of business transactions.

[0147] To implement the methods of the above embodiments, the present invention also provides a computer device, such as... Figure 8 As shown, the computer device 600 includes a memory 601 and a processor 602; wherein, the processor 602 reads executable program code stored in the memory 601 to run a program corresponding to the executable program code, so as to implement the various steps of the method described above.

[0148] To implement the above embodiments, this application also proposes a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the method described in the foregoing embodiments.

[0149] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0150] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified.

Claims

1. A service plan recommendation method, characterized in that, The method includes: Obtain information on users' computing network resource requirements; The user's business intent is determined based on the network resource demand information and the matching of multiple preset business scenarios. The computing power metrics of all cloud resource pools are calculated according to the computing power metric measurement system, and the business intent is matched with the computing power metrics to obtain the resource search matching result. Based on the resource search and matching results and the preset benchmark data, multiple optimal recommendation strategies for requesting service activation are calculated, and corresponding network resource recommendation schemes are determined based on the multiple optimal recommendation strategies.

2. The method according to claim 1, characterized in that, After determining the corresponding network resource recommendation scheme based on the multiple optimal recommendation strategies, the method further includes: Obtain the resource parameters of the proposed computing network resource recommendation scheme; A solution map is constructed based on the solution resource parameters, and the solution map is visualized to obtain a visualization result based on the cloud resource pool utilization rate.

3. The method according to claim 1, characterized in that, The various business scenarios correspond to various computing network application products.

4. The method according to claim 3, characterized in that, The user's computing network resource requirements are obtained through the business terminal input interface or the human-computer interaction interface; wherein, obtaining the user's computing network resource requirements through the business terminal input interface includes: The parameters and values ​​of the computing network application product are used as the parameters and values ​​corresponding to the menu options in the business terminal input interface to obtain the customer's computing network resource requirements information.

5. The method according to claim 4, characterized in that, Information on users' network resource requirements is obtained through a human-computer interaction interface, including: Voice recognition converts customer speech input into text for the human-computer interaction interface. Intent recognition categorizes the intent of text and directs it to the corresponding domain; By obtaining entity relationship information or word vectors of text in the corresponding domain through semantic recognition, and then processing them with the corresponding business logic, a speech synthesis result containing information on computing network resource requirements is obtained through speech synthesis.

6. The method according to claim 1, characterized in that, After obtaining the user's computing network resource requirements information, the method further includes: Obtain the business metrics corresponding to the computing network resource demand information; The business metrics are converted into computing network resource metrics according to the preset conversion rules.

7. The method according to claim 3, characterized in that, Based on the network resource demand information and the matching of various preset business scenarios, the user's business intent is determined, including: The business scenario is determined to be cloud desktop; The network resource requirements information is matched with all network capability templates corresponding to the cloud desktop, and the optimal network capability template is obtained based on the matching results. The business intent of using the computing network application products associated with the optimal computing network capability template as the computing network resource requirement information.

8. The method according to claim 1, characterized in that, By using a preset model function, different types of computing power resources in the computing network are mapped to a unified dimension to establish the computing power index measurement system.

9. The method according to claim 8, characterized in that, The computing network includes various types of computing resources, including CPUs, GPUs, NPUs, and FPGAs.

10. The method according to claim 1, characterized in that, The various optimal recommendation strategies include a comprehensive optimal recommendation strategy, a computing power optimal recommendation strategy, and a network optimal recommendation strategy.

11. The method according to claim 2, characterized in that, The resource search and matching results include candidate cloud resource pools; visualizing the solution map to obtain visualization results based on cloud resource pool utilization includes: Based on the candidate cloud resource pool, a multi-layer latency circle based on the latency dimension is obtained; The network connects the user's location with the resource pool recommended by the solution map, and the computing power distribution viewing result is obtained based on the network connection result and the computing power distribution of the multi-layer latency circle. Based on the computing power distribution, the visualization results show the city and network domain of the user's solution map, and the capabilities of each resource pool are highlighted with tags to obtain the visualization results.

12. A service plan recommendation device, characterized in that, include: The demand information acquisition module is used to acquire users' computing network resource demand information; The business intent determination module is used to determine the user's business intent based on the computing network resource demand information and a variety of preset business scenarios. The resource search and matching module is used to calculate the computing power indicators of all cloud resource pools according to the computing power indicator measurement system, and to perform computing network resource search and matching between the business intent and the computing power indicators to obtain the resource search and matching results. The computing network resource recommendation module is used to calculate multiple optimal recommendation strategies for requesting service activation based on the resource search and matching results and preset benchmark data, and to determine the corresponding computing network resource recommendation scheme based on the multiple optimal recommendation strategies.

13. A computer device, characterized in that, Including processor and memory; The processor reads executable program code stored in the memory to run a program corresponding to the executable program code, so as to implement the service scheme recommendation method as described in any one of claims 1-11.

14. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the service scheme recommendation method as described in any one of claims 1-11.

Images