Network activation method and apparatus for application launch, and device and medium

Abstract

The embodiments of the present application relate to the technical field of artificial intelligence (AI). Provided are a network activation method and apparatus for application launch, and a device and a medium. The method comprises: on the basis of a human-computer interaction mode, collecting a first natural language indicating an application launch service intent; inputting the first natural language into a large AI network model to obtain a target network activation scheme; performing simulation evaluation on the target network activation scheme; and after a simulation evaluation result satisfies a preset condition, activating a network according to the target network activation scheme. Using the technical solution provided in the embodiments of the present application, the capability for understanding user service-level intents can be enhanced, the capability for extending intent collection can be improved, and the adaptability to scenarios and networking can be boosted.

Description

A network opening method, device and equipment applied to online and a medium TECHNICAL FIELD

[0001] The present application relates to the technical field of artificial intelligence, in particular to a network opening method, device and equipment applied to online and a medium. BACKGROUND

[0002] Application online is a typical scenario in the network operation stage of the whole life cycle management service of a data center network (DCN). The implementation process of the application online function is actually the network opening process of the application online. At present, the network opening process of the application online is as follows: according to the network language of the indication intent input by a user, a target network opening scheme is determined from the pre-configured network opening scheme, and the network is opened according to the target network opening scheme, so as to realize the application online.

[0003] This application online network opening method has the following problems: the input intent is a professional network language, which is difficult for users to understand and requires professional network administrators to understand, and the user friendliness is poor; and only fixed intents can be collected and understood, when new intents and changed intents appear, the device cannot understand the new intents and changed intents, and the application online cannot be realized; in addition, the device can only use the pre-configured network opening scheme to open the network, and the pre-configured network opening scheme is limited and cannot adapt to new scenarios or networking requirements.

[0004] SUMMARY

[0005] The purpose of the embodiments of the present application is to provide a network opening method, device and equipment applied to online to increase the understanding ability of the user business level intent, improve the expansion ability of intent collection, and improve the adaptation ability to scenarios and networking. The specific technical solutions are as follows:

[0006] In a first aspect, the embodiments of the present application provide a network opening method applied to online, which comprises:

[0007] Collecting a first natural language indicating an application online business intent based on a human-computer interaction mode;

[0008] Inputting the first natural language into an artificial intelligence (AI) network large model to obtain a target network opening scheme;

[0009] Simulating and evaluating the target network opening scheme;

[0010] Opening the network according to the target network opening scheme after the simulation and evaluation result meets a preset condition.

[0011] In some embodiments, the step of collecting, based on the human-computer interaction mode, the first natural language indicating the application online service intention comprises:

[0012] displaying a human-computer interaction dialogue box;

[0013] receiving the first natural language indicating the application online service intention input by the user on the human-computer interaction dialogue box.

[0014] In some embodiments, the method further comprises:

[0015] displaying first prompt information on the human-computer interaction dialogue box, the first prompt information being used to guide the input of the application online service intention.

[0016] In some embodiments, the step of inputting the first natural language into the AI network large model to obtain the target network opening scheme comprises:

[0017] matching the first natural language with pre-stored knowledge of multiple network opening schemes to obtain multiple candidate knowledge;

[0018] inputting the first natural language and the multiple candidate knowledge into the AI network large model to obtain the target network opening scheme.

[0019] In some embodiments, the method further comprises:

[0020] collecting, based on the human-computer interaction mode, a second natural language indicating a display mode of the network opening scheme;

[0021] inputting the second natural language into the AI network large model to obtain a target display mode;

[0022] calling, by an intelligent agent, an interface corresponding to the target display mode to display the target network opening scheme.

[0023] In some embodiments, the step of collecting, based on the human-computer interaction mode, a second natural language indicating a display mode of the network opening scheme comprises:

[0024] displaying a human-computer interaction dialogue box;

[0025] receiving the second natural language indicating the display mode of the network opening scheme input by the user on the human-computer interaction dialogue box.

[0026] In some embodiments, the method further comprises:

[0027] displaying second prompt information on the human-computer interaction dialogue box, the second prompt information being used to guide the input of the display mode of the network opening scheme.

[0028] In some embodiments, the step of simulating and evaluating the target network opening scheme comprises:

[0029] Based on user operation, the target network opening scheme is adjusted;

[0030] The adjusted target network opening scheme is simulated and evaluated.

[0031] In some embodiments, the AI network large model outputs multiple target network opening schemes;

[0032] The step of simulating and evaluating the target network opening scheme comprises:

[0033] From the multiple target network opening schemes, a target network opening scheme is determined;

[0034] The determined target network opening scheme is simulated and evaluated.

[0035] In some embodiments, the step of determining a target network opening scheme from the multiple target network opening schemes comprises:

[0036] Based on the human-computer interaction mode, a third natural language indicating the determined target network opening scheme is collected;

[0037] The third natural language is input into the AI network large model to obtain the determined target network opening scheme.

[0038] In some embodiments, the step of collecting a third natural language indicating the determined target network opening scheme based on the human-computer interaction mode comprises:

[0039] A human-computer interaction dialog box is displayed;

[0040] The third natural language indicating the determined target network opening scheme input by the user on the human-computer interaction dialog box is received.

[0041] In some embodiments, the method further comprises:

[0042] Third prompt information is displayed on the human-computer interaction dialog box, and the third prompt information is used to guide the input of the determined target network opening scheme.

[0043] In some embodiments, the step of simulating and evaluating the target network opening scheme comprises:

[0044] Based on the human-computer interaction mode, a fourth natural language indicating the simulation evaluation is collected;

[0045] The fourth natural language is input into the AI network large model to obtain a simulation evaluation instruction;

[0046] The intelligent agent calls an interface corresponding to the simulation evaluation instruction to simulate and evaluate the target network opening scheme.

[0047] In some embodiments, the step of collecting the fourth natural language indicating simulation evaluation based on the human-computer interaction manner comprises:

[0048] displaying a human-computer interaction dialog box;

[0049] receiving the fourth natural language indicating simulation evaluation input by the user on the human-computer interaction dialog box.

[0050] In some embodiments, the method further comprises:

[0051] displaying fourth prompt information on the human-computer interaction dialog box, the fourth prompt information being used to guide input of execution information of simulation evaluation.

[0052] In some embodiments, the step of opening the network according to the target network opening scheme comprises:

[0053] collecting a fifth natural language indicating network opening based on a human-computer interaction manner;

[0054] inputting the fifth natural language into the AI network large model to obtain a network opening instruction;

[0055] The intelligent agent calls an interface corresponding to the network opening instruction to open the network by using the target network opening scheme.

[0056] In some embodiments, the step of collecting the fifth natural language indicating network opening based on the human-computer interaction manner comprises:

[0057] displaying a human-computer interaction dialog box;

[0058] receiving the fifth natural language indicating network opening input by the user on the human-computer interaction dialog box.

[0059] In some embodiments, the method further comprises:

[0060] displaying fifth prompt information on the human-computer interaction dialog box, the fifth prompt information being used to guide input of execution information of network opening.

[0061] In some embodiments, the method further comprises:

[0062] determining an operation to be executed;

[0063] displaying prompt information associated with the operation to be executed on the human-computer interaction dialog box.

[0064] In some embodiments, the method further comprises:

[0065] obtaining a sample natural language and a sample network opening solution corresponding to the sample natural language;

[0066] fine-tuning the AI network large model by using the sample natural language and the sample network opening solution.

[0067] In a second aspect, the embodiments of the present application provide a network opening device for application online, and the device comprises:

[0068] a collection module configured to collect a first natural language indicating a business intention of application online based on a human-computer interaction mode;

[0069] a determination module configured to input the first natural language into an AI network large model to obtain a target network opening solution;

[0070] a simulation module configured to simulate and evaluate the target network opening solution;

[0071] an opening module configured to open a network according to the target network opening solution after a simulation evaluation result meets a preset condition.

[0072] In some embodiments, the collection module is specifically configured to:

[0073] display a human-computer interaction dialogue box;

[0074] receive a first natural language indicating a business intention of application online input by a user on the human-computer interaction dialogue box.

[0075] In some embodiments, the collection module is further configured to:

[0076] display first prompt information on the human-computer interaction dialogue box, the first prompt information being used to guide input of the business intention of application online.

[0077] In some embodiments, the determination module is specifically configured to:

[0078] match the first natural language with pre-stored knowledge of multiple network opening solutions to obtain multiple candidate knowledge;

[0079] input the first natural language and the multiple candidate knowledge into an AI network large model to obtain a target network opening solution.

[0080] In some embodiments, the device further comprises a display module configured to:

[0081] collect a second natural language indicating a display mode of a network opening solution based on a human-computer interaction mode;

[0082] inputting the second natural language into the AI network large model to obtain a target display mode;

[0083] adopting an agent to call an interface corresponding to the target display mode to display the target network opening solution.

[0084] In some embodiments, the display module is specifically configured to:

[0085] display a human-computer interaction dialogue box;

[0086] receive a second natural language input by a user on the human-computer interaction dialogue box, the second natural language indicating a display mode of a network opening solution.

[0087] In some embodiments, the display module is further configured to:

[0088] display second prompt information on the human-computer interaction dialogue box, the second prompt information guiding input of the display mode of the network opening solution.

[0089] In some embodiments, the simulation module is specifically configured to:

[0090] adjust the displayed target network opening solution based on user operation;

[0091] simulate and evaluate the adjusted target network opening solution.

[0092] In some embodiments, the AI network large model outputs a plurality of target network opening solutions;

[0093] The simulation module is specifically configured to:

[0094] determine a target network opening solution from the plurality of target network opening solutions;

[0095] simulate and evaluate the determined target network opening solution.

[0096] In some embodiments, the simulation module is specifically configured to:

[0097] collect a third natural language indicating the determined target network opening solution based on a human-computer interaction mode;

[0098] input the third natural language into the AI network large model to obtain the determined target network opening solution.

[0099] In some embodiments, the simulation module is specifically configured to:

[0100] display a human-computer interaction dialogue box;

[0101] receive a third natural language input by a user on the man-machine interactive dialogue box, the third natural language indicating a determined target network opening scheme.

[0102] In some embodiments, the simulation module is further configured to:

[0103] display third prompt information on the man-machine interactive dialogue box, the third prompt information guiding input of the determined target network opening scheme.

[0104] In some embodiments, the simulation module is specifically configured to:

[0105] collect a fourth natural language indicating simulation evaluation based on the man-machine interactive mode;

[0106] input the fourth natural language into the AI network large model to obtain simulation evaluation instructions;

[0107] call an interface corresponding to the simulation evaluation instructions by using an agent to perform simulation evaluation on the target network opening scheme.

[0108] In some embodiments, the simulation module is specifically configured to:

[0109] display a man-machine interactive dialogue box;

[0110] receive a fourth natural language input by a user on the man-machine interactive dialogue box, the fourth natural language indicating simulation evaluation.

[0111] In some embodiments, the simulation module is further configured to:

[0112] display fourth prompt information on the man-machine interactive dialogue box, the fourth prompt information guiding input of execution information of simulation evaluation.

[0113] In some embodiments, the opening module is specifically configured to:

[0114] collect a fifth natural language indicating network opening based on the man-machine interactive mode;

[0115] input the fifth natural language into the AI network large model to obtain network opening instructions;

[0116] call an interface corresponding to the network opening instructions by using an agent to open a network by using the target network opening scheme.

[0117] In some embodiments, the opening module is specifically configured to:

[0118] display a man-machine interactive dialogue box;

[0119] receive a fifth natural language input by a user on the man-machine interactive dialogue box, the fifth natural language indicating network opening.

[0120] In some embodiments, the opening module is further configured to:

[0121] display fifth prompt information on the man-machine interaction dialog box, the fifth prompt information being used for guiding input of execution information of network opening.

[0122] In some embodiments, the apparatus further includes a display module configured to:

[0123] determine an operation to be performed;

[0124] display prompt information associated with the operation to be performed on the man-machine interaction dialog box.

[0125] In some embodiments, the apparatus further includes a fine-tuning module configured to:

[0126] obtain a sample natural language and a sample network opening scheme corresponding to the sample natural language;

[0127] fine-tune the AI network large model using the sample natural language and the sample network opening scheme.

[0128] In a third aspect, an electronic device is provided, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory complete communication with each other through the communication bus;

[0129] The memory is configured to store a computer program.

[0130] The processor is configured to execute the program stored in the memory, and implement the method of any one of the first aspect.

[0131] In a fourth aspect, a computer readable storage medium is provided, and the computer readable storage medium stores a computer program. When the computer program is executed by a processor, the method of any one of the first aspect is implemented.

[0132] In yet another embodiment provided in the present application, a computer program product containing instructions is provided, which, when executed on a computer, causes the computer to perform the method of any one of the first aspect.

[0133] Embodiments of the present application have the following beneficial effects:

[0134] In the technical scheme provided by the embodiment, natural language is used to express the user intention, i.e., the application online service intention. The natural language is a language that is easy for users to understand and does not require professional knowledge background. Therefore, the user intention is expressed by using the natural language, the network opening for application online is realized, and the understanding ability of the user service level intention is increased. In addition, in the technical scheme provided by the embodiment, the AI network large model is used to analyze and process the natural language (i.e., the first natural language) expressing the user intention, and a target network opening scheme meeting the application online service intention is obtained. Since the AI network large model has strong natural language processing capability, it is not necessary to pre-configure the network opening scheme and intention. That is, even if new intention and changed intention appear, or the scene and networking requirement change, the AI network large model can also analyze and obtain the required target network opening scheme, improve the expansion capability of the intention collection, and improve the adaptation capability to the scene and networking. BRIEF DESCRIPTION OF DRAWINGS

[0135] The accompanying drawings, which are included to provide a further understanding of the present application, form a part of the present application and illustrate the illustrative embodiments of the present application and together with the description serve to explain the present application. The present application is not intended to be unduly limited by the illustrative embodiments and description.

[0136] FIG. 1 is a schematic diagram of a self-intelligent network hierarchical generation definition of a data center;

[0137] FIG. 2 is a schematic diagram of five general task stages of information communication network operation and management activities;

[0138] FIG. 3 is a schematic diagram of an application online processing flow;

[0139] FIG. 4 is a schematic diagram of an intention collection stage interface;

[0140] FIG. 5 is a schematic diagram of a scheme recommendation stage interface;

[0141] FIG. 6 is a schematic diagram of a scheme generation stage interface;

[0142] FIG. 7 is a flowchart of a network opening method for application online provided by the embodiment of the present application;

[0143] FIG. 8 is a detailed schematic diagram of step S72 provided by the embodiment of the present application;

[0144] FIG. 9 is a schematic diagram of a scheme knowledge base establishment flow provided by the embodiment of the present application;

[0145] FIG. 10 is a flowchart of obtaining a target network opening scheme based on an AI network large model provided by the embodiment of the present application;

[0146] FIG. 11 is a schematic diagram of intention analysis stage processing logic provided by the embodiment of the present application;

[0147] Fig. 12 is a flow diagram illustrating a network opening solution according to an embodiment of the present application;

[0148] Fig. 13 is a diagram illustrating a graphical display using an agent framework according to an embodiment of the present application;

[0149] Fig. 14 is a first flow diagram of step S73 according to an embodiment of the present application;

[0150] Fig. 15 is a detailed diagram of step S141 according to an embodiment of the present application;

[0151] Fig. 16 is a second flow diagram of step S73 according to an embodiment of the present application;

[0152] Fig. 17 is a flow diagram of step S74 according to an embodiment of the present application;

[0153] Fig. 18 is a diagram of a network opening method framework for application online according to an embodiment of the present application;

[0154] Figs. 19a-19c are first diagrams of an SDN controller interface according to an embodiment of the present application;

[0155] Figs. 20a-20e are second diagrams of an SDN controller interface according to an embodiment of the present application;

[0156] Fig. 21 is a third diagram of an SDN controller interface according to an embodiment of the present application;

[0157] Figs. 22a-22b are fourth diagrams of an SDN controller interface according to an embodiment of the present application;

[0158] Fig. 23 is a diagram of a network opening apparatus for application online according to an embodiment of the present application;

[0159] Fig. 24 is a diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0160] To make the objectives, technical solutions, and advantages of the present application clearer, further detailed descriptions will be given below with reference to the accompanying drawings and embodiments. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art belong to the scope of the present application.

[0161] For the convenience of understanding, the terms appearing in the embodiments of the present application are explained as follows.

[0162] Self-intelligent network: aims to build an automated and intelligent operation and maintenance capability throughout the life cycle of the network, provide new network and information and communication technology (ICT) services with "zero waiting, zero failure, and zero contact" to consumers and vertical industry customers, and create a "self-configuration, self-repair, and self-optimization" digital operation and maintenance capability for network intelligent operation and maintenance.

[0163] AI network large model: a machine learning model with super large-scale parameters (usually more than one billion) and super strong computing resources, capable of processing massive data and completing various complex tasks such as natural language processing and image recognition. AI network large model can also be referred to as Artificial Intelligence Generated Content (AIGC) network large model.

[0164] Data center application: a business application or application system deployed in a DCN to complete specific functions, such as online shopping applications. For ease of description, the application referred to hereinafter is the data center application.

[0165] Data center service: a functional unit within a data center application that operates independently to achieve specific business purposes. A data center application is composed of one or more data center services. For example, an online shopping application includes inventory services, flash sale services, authentication services, logistics services, and database services. For ease of description, the service referred to hereinafter is the data center service.

[0166] Application online: formally launching a data center application such as a website, application program, or other product to the public so that users can access and use it.

[0167] Currently, DCN is gradually developing towards a self-intelligent network, forming a data center self-intelligent network. The network referred to hereinafter is the data center self-intelligent network. The goal of the data center self-intelligent network is to gradually reduce and eventually eliminate human operation, guide customers through defined hierarchical capabilities, and gradually evolve towards the vision of an unattended DCN self-intelligence. As shown in Figure 1, the data center self-intelligent network is defined by hierarchical generations.

[0168] Level 1 (L1) manual processing: the data center self-intelligent network relies on experience and manual processing.

[0169] Level 2 (L2) tool collaboration: the data center self-intelligent network uses domain tools to help people achieve efficiency improvement in the domain.

[0170] Level 3 (L3) network automation: The data center self-intelligent network realizes the automation of network intent, and part of the system assists in analysis and manual decision-making.

[0171] Level 4 (L4) network intelligence: The data center self-intelligent network is integrated with applications, and AI, Machine Learning (ML), etc. are used to realize intelligent processing of application intent, system-assisted analysis, and manual decision-making.

[0172] Level 5 (L5) full self-intelligent network: The data center self-intelligent network realizes a full self-intelligent data center, and the system performs analysis and decision-making to realize automated end-to-end self-intelligence.

[0173] The development of network automation and intelligence technology is essential for the realization of the goals of the data center self-intelligent network, and the classification of the data center self-intelligent network can be divided into two dimensions.

[0174] Dimension one: DCN full life cycle management services, including planning and construction, network operation, monitoring and operation, and optimization of operation.

[0175] Dimension two: Information communication network operation and management activities, including intent management, perception, analysis, decision-making, and execution.

[0176] The data center self-intelligent network needs to perform information communication network operation and management activities at each management service stage of its full life cycle. As shown in FIG. 2, the five general task stages of information communication network operation and management activities, i.e. the above-mentioned intent management, perception, analysis, decision-making, and execution.

[0177] The intent management stage: Understand the customer's business and management and operation intent, and translate it into specific network configuration and strategy. Intent management can support low-level traditional functions, such as supporting any required orchestration to coordinate configuration operations on the network, or higher-level abstract information understanding capabilities, such as supporting open and modifiable capabilities to facilitate customers to make some adjustments to the scheme according to actual networking. Intent management also includes full recording of business intent operation processes, which are traceable and queryable.

[0178] The perception stage: Real-time monitoring and observation of the DCN to discover network business anomalies or multi-dimensional problems such as Service-Level Agreement (SLA), and trigger network analysis and positioning. Specifically, it includes collecting network raw data and performing necessary preprocessing (such as data cleaning, enhancement, statistics, etc.) on the data to achieve the purpose of monitoring and perceiving network information (including network performance, network anomalies, network events, etc.), and realizing visual presentation.

[0179] Analysis stage: analysis of the current state of the DCN, and based on historical data, combined with the customer's intention to make network analysis, generate options and suggestions that can meet the customer's intention of operation action and execution strategy.

[0180] Decision stage: through the options and suggestions given by the analysis stage, decide the most suitable, meet the customer's intention demand, executable network operation and strategy.

[0181] Execution stage: generate executable network operation and strategy for the customer's intention that has been decided, automatically implement deployment to the production network of the data center (i.e. DCN), while also containing the business verification of the intention after network implementation.

[0182] In FIG. 2, the enterprise system / customer inputs intention, the self-intelligent network manages the input intention, understands the intention, and performs perception, analysis, decision, and execution, and operates on the managed object and feeds back to the enterprise system / customer. The managed object of the DCN is various applications of the DCN, DCN devices, and network endpoint devices.

[0183] Application online as a typical scenario of network operation stage in the whole life cycle management service of DCN is a business operation commonly used by enterprise data center customers. DCN carries various applications of enterprises, and applications accessed by external customers require stable operation in DCN. With the development and innovation of business, when an enterprise customer needs to online a new application, such as an online online shopping application, from the perspective of the application department, the business scenario is: online new application adopts departmental deployment to improve the reliability of business, and multiple microservices are deployed in different regions of DCN; network appeal is: microservices within the application need to pass through load balancing, and interconnection between microservices is required, and global WEB service needs to provide access ability to external users. The network department (i.e. application department) will analyze the overall architecture of the application, including internal function services, service resources, service communication relationship, security access isolation requirements and a series of requirements, based on which the network department will allocate network resources, access the application and open the network between the applications to ensure the normal operation of the application. Application online is a complex process, including network design, simulation evaluation, configuration distribution and result acceptance, etc. The application online processing flow is shown in FIG. 3.

[0184] The application online scenario includes an external access area, a WEB area, an application (APP) area, and a database (DB) area. The external access area includes a core switching area, an Internet, and a public service area; the WEB area includes a WEB service cluster; the APP area includes a plurality of sub-services (i.e., sub-service 1, sub-service 2, …, sub-service n, etc.), which are the microservices described above; and the DB area includes a database cluster. As shown in FIG. 3, the plurality of sub-services access the database cluster and the WEB service cluster, and access the Internet through the WEB service cluster to realize external access.

[0185] The process of implementing the application online function is actually a network opening process of the application online. At present, the network opening process of the application online is as follows: through an interface entrance fixed by a software defined network (SDN) controller, a target network opening scheme is determined from a pre-configured network opening scheme according to a network language of an indication intent input by a user, and the network is opened according to the target network opening scheme to realize the application online. The network opening process of the application online consists of four stages of intent collection, scheme recommendation, scheme generation, and scheme delivery, and is as follows.

[0186] 1) Intent collection stage

[0187] A device (i.e., a DCN device) collects a network language of an indication intent input by a user, wherein the network language is a language for expressing a network architecture according to business resource arrangement, such as creating a security partition, creating an external network and a security partition including which services, interconnection relationship between services, and number of created instances, etc. In this stage, the user can arrange the intent content by himself according to the security partition, external network, service, and intercommunication relationship resources provided by the device in the resource box.

[0188] Taking that a certain customer needs to online an online shopping application as an example, the online shopping application includes inventory service, kill service, authentication service, logistics service, and database service, etc. As shown in the intent collection stage interface of FIG. 4, it includes a stage column at the top, a resource column at the left, and a display column in the middle. The stage column includes four stages of intent collection, scheme recommendation, scheme generation, and scheme delivery, and the current interface is in the intent collection stage; the resource column includes four resources of security partition, external network, service, and intercommunication relationship, and the user can create the DB area, WEB area, and APP area according to the plan, and create the corresponding services in different partitions. Among them, the APP area includes the inventory service and the kill service, the WEB area includes the authentication service and the logistics service, and the DB area includes the remote dictionary service (Redis).

[0189] Additionally, users can customize the inter-service communication relationships according to their business needs. For example, the inventory service might need to access the Redis database, while the inventory, flash sale, authentication, and logistics services might need to access the external network. The flash sale service, in turn, might need to access the inventory and authentication services. After establishing the gateway architecture using the above method and completing the input of the intent (i.e., the input of network language), the next step of solution recommendation can be performed.

[0190] 2) Solution Recommendation Stage

[0191] The device analyzes and infers the intent input by the user, and recommends one or more network provisioning schemes through the pre-built logic of the SDN controller. Different network provisioning schemes have different preferences; for example, some network provisioning schemes prioritize resource consumption, while others prioritize security.

[0192] Figure 5 shows the interface for the solution recommendation stage. The device recommends two network activation solutions to the user (i.e., Solution 1 and Solution 2). The network activation solution currently displayed on the solution recommendation stage interface is Solution 1, and the displayed content includes a graphical representation of Solution 1, the reasons for recommending Solution 1, and a text description of Solution 1 located in the lower right corner. The user can click the "Solution 2" button below the stage bar, and the device will respond to the user's click of the "Solution 2" button and display Solution 2. After the user selects a network activation solution (such as Solution 1) according to their needs, they can click the "Design" button in the upper right corner, and the device will respond to the user's click of the "Design" button and enter the solution generation stage.

[0193] 3) Solution Generation Stage

[0194] The device can perform simulation evaluation and verification of the selected network deployment scheme to ensure the reliability of online services. After the user enters this stage, the network administrator can design the scheme in detail, taking into account the physical topology and the resource usage of existing network devices (i.e., network endpoint devices). After the supplementation is completed, the device performs simulation evaluation of the selected network deployment scheme to assess its feasibility, configuration changes of existing network devices, and resource changes.

[0195] The scheme generation stage interface is shown in FIG. 6. The scheme generation stage interface displays the topology structure of the generated network opening scheme, the simulation evaluation process in the lower right corner, and the operations performed by the simulation evaluation process. The topology structure also includes a firewall (FW) module. In the simulation evaluation process: a virtual router of the authentication service is created, a virtual router of the logistics service is created, a virtual router of the inventory service is created, a virtual router of the flash sale service is created, an egress gateway (service level (SL)-GW) is bound, and the like, so as to realize instantiation of the logical network and evaluate the instantiated logical network. After the simulation evaluation is completed (i.e., the simulation evaluation process is 100%), the user can click "view report" to obtain the evaluation result of this time, and determine the next operation according to the evaluation result. If the evaluation result indicates that the scheme can be executed, the user can click the "configuration delivery" button in the upper right corner, and then the device responds to the operation of the user clicking the "configuration delivery" button to enter the scheme delivery stage. Alternatively, if the evaluation result indicates that the scheme cannot be executed, the user can return to the scheme recommendation stage, adjust the recommended scheme, and then enter the scheme generation stage again.

[0196] 4) Scheme delivery stage

[0197] The scheme delivery stage converts the network opening scheme into a network configuration, delivers the network configuration to the existing network device in the production environment, outputs a delivery report, and performs network opening to realize application online.

[0198] The application online network opening method inputs professional network language, such as the number of security partitions, external networks, services, and instances, and the interconnection relationship between services. Such network language is difficult for users to understand and requires professional network administrators to understand. However, the device does not have the understanding ability for the business-level intentions that users are really concerned about, such as the number of concurrent access amounts supported by an application composed of several services, the reliability requirements of the application on the network, and whether security protection is required for inter-application access. Therefore, the existing application online network opening method has poor user friendliness.

[0199] In addition, the existing application online network opening method can only collect and understand fixed intentions. When new intentions and changed intentions appear, the device cannot understand the new intentions and changed intentions, and cannot realize application online. For example, an online shopping application needs to increase the number of concurrent users, increase the number of application instances, and increase the bandwidth of the link. However, the device cannot directly collect these new intentions and lacks expansion capability.

[0200] In addition, the device can only open the network by using a pre-configured network opening scheme, and the pre-configured network opening scheme is limited and cannot adapt to new scenarios or networking requirements. For example, for new scenarios (such as scenarios requiring cross-DCN deployment, scenarios requiring cross-public cloud and private cloud deployment, etc.) and new networking requirements (such as DCN export gateway supporting access to a wide area network based on Segment Routing over Internet Protocol Version 6 (SRv6) tunnel of an Internet Protocol Version 6 forwarding plane), the network opening scheme needs to be reconfigured to realize the adaptation and conversion logic of the new scenarios and new networking requirements.

[0201] In order to increase the understanding ability of the user business level intention, improve the expansion ability of the intention collection, and improve the adaptation ability to the scene and networking, an application online network opening method is provided in the embodiments of the present application, which is applied to a computer, a server, an SDN controller and other electronic devices. For ease of description, the electronic device is taken as the execution subject in the following, and does not have a limiting effect.

[0202] Referring to FIG. 7, FIG. 7 is a flowchart of the application online network opening method provided by the embodiments of the present application, and the method includes the following steps.

[0203] In step S71, a first natural language indicating an application online business intention is collected based on a human-computer interaction mode.

[0204] In step S72, the first natural language is input into an AI network large model to obtain a target network opening scheme.

[0205] In step S73, the target network opening scheme is simulated and evaluated.

[0206] In step S74, after the simulation evaluation result meets a preset condition, the network is opened according to the target network opening scheme.

[0207] The technical scheme provided in the embodiments of the present application uses natural language to express the user's intention, i.e., the application online service intention. The natural language is a language that is easy for users to understand and does not require a professional knowledge background. Therefore, using natural language to express the user's intention realizes network opening of the application online, and increases the understanding ability of the user's service level intention. In addition, in the technical scheme provided in the embodiments of the present application, an AI network large model is used to analyze and process the natural language (i.e., the first natural language) expressing the user's intention, to obtain a target network opening scheme that meets the application online service intention. Since the AI network large model has strong natural language processing capability, it does not need a pre-configured network opening scheme and intention. That is, even if new intentions and changing intentions appear, or the scene and networking requirements change, the AI network large model can still analyze and obtain the required target network opening scheme, improving the expansion capability of intention collection and the adaptation capability to the scene and networking.

[0208] The network opening method for application online provided in the embodiments of the present application can include four stages: the intention perception stage corresponding to step S71, the intention analysis stage corresponding to step S72, the intention decision stage corresponding to step S73, and the intention execution stage corresponding to step S74.

[0209] In the above step S71, the first natural language is any natural language used to indicate the application online service intention. The first natural language can be text or voice, and the form of the first natural language is not limited herein.

[0210] The electronic device can provide multiple human-computer interaction modes for the user, and the user inputs the first natural language to the electronic device through multiple human-computer interaction modes. Then, the electronic device collects the first natural language input by the user based on multiple human-computer interaction modes.

[0211] In one example, the electronic device provides a human-computer interaction dialog box. The above step S71 can be: displaying the human-computer interaction dialog box; and receiving the first natural language indicating the application online service intention input by the user on the human-computer interaction dialog box.

[0212] For example, the electronic device displays the human-computer interaction dialog box, the user inputs text on the human-computer interaction dialog box, and the electronic device detects the text on the human-computer interaction dialog box and takes the text as the first natural language.

[0213] For another example, the electronic device displays the human-computer interaction dialog box, the user inputs voice on the human-computer interaction dialog box, and the electronic device can receive the voice input by the user through a voice receiver (such as a microphone) and directly take the voice as the first natural language; or, the electronic device recognizes the voice, converts the voice into text, and then takes the text as the first natural language.

[0214] Through the man-machine interactive dialogue box, the interaction between the user and the electronic device can be intuitively displayed, and the friendliness of the network opening mode of the application online is improved.

[0215] When the man-machine interaction is performed by using the man-machine interactive dialogue box, the electronic device can further display prompt information (i.e., first prompt information) for guiding the input of the application online service intention on the man-machine interactive dialogue box. The first prompt information can include specific content that needs to be input by the user to the man-machine interactive dialogue box, for example, the first prompt information can include "application name", "application is composed of several service parts", etc. The first prompt information can also include requirements for the content input by the user, for example, the first prompt information can include "please use natural language to input the demand of the application online on the network", etc. The content of the first prompt information is not limited herein. The electronic device can guide the user to input the service intention by displaying the first prompt information on the man-machine interactive dialogue box, so as to more accurately meet the user demand, and facilitate the user operation.

[0216] In another example, the electronic device can further receive the natural language input by the user on the other device, and take the natural language as the first natural language. Taking the electronic device as a personal computer (PC) terminal and the other device as a mobile phone terminal as an example, the user can input the natural language on the mobile phone terminal, the mobile phone terminal sends the natural language to the PC terminal, and the PC terminal takes the received natural language as the first natural language. The man-machine interaction mode between the electronic device and the user is not limited herein.

[0217] In the above step S72, the AI network large model can be a general AI network large model, or an AI network large model obtained by pre-training a general AI network large model based on a large amount of DCN corpus, which is not limited herein. For the AI network large model, the training process is: inputting the DCN corpus into the AI network large model; after the AI network large model obtains the DCN corpus, the context knowledge in the DCN corpus is learned.

[0218] In the embodiment of the present application, the DCN corpus can include DCN networking information, topology information, device information, logical network information, application network and logical network mapping relationship information, logical network and physical network mapping relationship information, link information, port information, security information, routing information, network policy information, reliability information, consistency information, stability information, performance load information, capacity information, system information and resource information, etc. The DCN corpus is not limited herein.

[0219] The electronic device inputs the collected first natural language into the AI network large model, so that the AI network large model outputs a network opening scheme meeting the application online business intention of the user based on the first natural language, and the electronic device outputs the network opening scheme output by the AI network large model as a target network opening scheme.

[0220] In the embodiments of the present application, the electronic device can call an interface (i.e., a first interface) corresponding to the AI network large model and input the first natural language into the AI network large model. The interface can be an application programming interface (API), and the first interface is an API corresponding to the AI network large model. The manner in which the electronic device inputs the first natural language into the AI network large model is not limited herein.

[0221] In the above step S73, after obtaining the target network opening scheme, the electronic device performs simulation evaluation on the target network opening scheme and obtains a simulation evaluation result, to ensure the reliability of the application online business based on the target network opening scheme and to ensure that the target network opening scheme can meet the expected operation.

[0222] In the embodiments of the present application, the electronic device can perform simulation evaluation on the target network opening scheme based on a simulation platform. For example, the electronic device can simulate opening a network according to the target network opening scheme to obtain specific conditions after the application is online, perform comprehensive simulation evaluation on the target network opening scheme from multiple aspects such as the satisfaction degree of the existing network resources, the connectivity between the expected businesses, and the impact on the existing network, and confirm the simulation evaluation result by checking a simulation report.

[0223] In the above step S74, the preset condition can be a simulation evaluation verification pass result in multiple aspects, which can be set according to actual conditions and is not limited herein. The target network opening scheme meeting the preset condition is a target network opening scheme that successfully passes the simulation evaluation verification.

[0224] After the target network opening scheme successfully passes the simulation evaluation verification, the electronic device opens a network according to the target network opening scheme, distributes network configurations corresponding to the target network opening scheme to network devices (i.e., existing network devices) in a production environment, and outputs a distribution report, to realize network opening and application online.

[0225] In the embodiments of the present application, if the target network opening scheme does not meet the preset condition (i.e., the target network opening scheme does not pass the simulation evaluation verification), the electronic device can feed back a message that the simulation evaluation does not pass to the user, and return to the intention analysis stage (i.e., step S72), adjust the target network opening scheme, obtain a new target network opening scheme, and then perform simulation evaluation on the new target network opening scheme. This cycle of adjustment is repeated until the target network opening scheme meets the preset condition.

[0226] In order to complete the network opening and application online as soon as possible, the simulation evaluation result can include the factors that cause the simulation evaluation to fail, so that the user can adjust the target network opening scheme and overcome the problem of failing simulation evaluation.

[0227] In some embodiments, referring to FIG. 8, which is a detailed schematic diagram of step S72 provided by the embodiments of the present application, the above-mentioned step S72 can include the following steps.

[0228] Step S81: match the first natural language with the pre-stored knowledge of multiple network opening schemes to obtain multiple candidate knowledge.

[0229] Step S82: input the first natural language and the multiple candidate knowledge into the AI network large model to obtain the target network opening scheme.

[0230] In the technical solutions provided by the embodiments of the present application, the electronic device pre-establishes a scheme knowledge base including the knowledge of multiple network opening schemes, and uses the scheme knowledge base with rich network experience as an effective supplement for the AI network large model to recommend the network opening scheme in the application online scenario. Therefore, the AI network large model can recommend the target network opening scheme according to the candidate knowledge selected from the scheme knowledge base and the application online business intention, and the accuracy of recommending the network opening scheme is improved.

[0231] In the above-mentioned step S81, the knowledge of the network opening scheme is related to the application online scenario, and can be text information of an actual network opening scheme or text information used to generate the network opening scheme, and no limitation is made thereto.

[0232] In the embodiments of the present application, the electronic device can establish the scheme knowledge base in the following manner: obtaining experience and guidance documents related to application online scene as original knowledge, cutting the original knowledge into multiple text segments, and performing vectorization processing on each text segment, storing each vectorized text segment as knowledge of network opening scheme in the database, obtaining knowledge of multiple network opening schemes, and thus forming a scheme knowledge base, as shown in one of the schematic diagrams of the scheme knowledge base establishment process in FIG. 9. The original knowledge can include scheme scene, business requirement, design principle, design experience, landing case, and scheme value, and the content of the original knowledge and the manner in which the electronic device establishes the scheme knowledge base are not limited herein.

[0233] In the embodiments of the present application, the manner of vectorization processing can be: determining the feature value corresponding to each character in the text segment according to the pre-stored correspondence between characters and feature values, and concatenating the feature values corresponding to the characters in sequence to obtain the vector corresponding to the text segment, that is, the vectorization processing of the text segment is realized.

[0234] The manner of vectorization processing can also be: preprocessing the text segment to remove invalid characters, conjunctions and other useless strings in the text segment, such as commas, spaces, and, but, etc.; determining the feature value corresponding to each character in the preprocessed text segment according to the pre-stored correspondence between characters and feature values, and concatenating the feature values corresponding to the characters in sequence to obtain the vector corresponding to the text segment, that is, the vectorization processing of the text segment is realized.

[0235] In the embodiments of the present application, the vectorization processing can also be realized in other manners, and it is only required to convert the text segment into a vector represented by 0 and 1.

[0236] After the electronic device collects the first natural language, the electronic device performs vectorization processing on the first natural language. Here, the electronic device can take the first natural language as a text segment, convert the first natural language into a vector represented by 0 and 1 by using the above-mentioned vectorization processing manner of the text segment, and obtain the vectorized first natural language.

[0237] The electronic device can use similarity algorithms such as cosine similarity and Euclidean distance to perform similarity matching between the vectorized first natural language and the knowledge of multiple network opening schemes in the scheme knowledge base, and according to the matching result, obtain the first preset number of vectors most similar to the first natural language, that is, obtain the knowledge of multiple network opening schemes most similar to the application online business intent, and the multiple network opening scheme knowledge is the multiple candidate knowledge, and the value of the preset number is not limited herein.

[0238] In the embodiments of the present application, the plurality of candidate knowledge obtained by the electronic device corresponding to the knowledge of the network opening scheme in the scheme knowledge base can be the text information of the actual network opening scheme, or can be the text information according to which the network opening scheme is generated, and the present application is not limited in this regard.

[0239] In the embodiments of the present application, the electronic device can continuously update the scheme knowledge base, that is, the electronic device can obtain new experience and guidance documents related to the application online scene (i.e., original knowledge), and store the new original knowledge after cutting and vectorization processing in the database to update the scheme knowledge base. Further, the electronic device can use the updated scheme knowledge base for matching to obtain more rich and accurate plurality of candidate knowledge, and the present application is not limited in this regard.

[0240] In the above step S82, the electronic device adds the plurality of candidate knowledge obtained as context to the prompt template together with the first natural language, inputs the added prompt template into the AI network large model, so that the AI network large model performs context reasoning such as information combination and exclusion based on the plurality of candidate knowledge according to the application online business intent, obtains and outputs the recommended network opening scheme, and the electronic device takes the network opening scheme output by the AI network large model as the target network opening scheme.

[0241] In the embodiments of the present application, the electronic device can call the request interface to input the first natural language and the plurality of candidate knowledge (i.e., the added prompt template) into the AI network large model, and the present application is not limited in this regard.

[0242] The complete flow of the electronic device obtaining the target network opening scheme based on the AI network large model is shown in FIG. 10. The electronic device collects the application online business intent (i.e., the first natural language) input by the user, vectorizes the application online business intent, and performs information retrieval in the scheme knowledge base, that is, performs similarity matching between the vectorized application online business intent and the vectorized text fragments in the scheme knowledge base to obtain the top N (i.e., a preset number) network opening scheme knowledge (i.e., candidate knowledge) with the highest similarity to the application online business intent. The electronic device calls the request API (i.e., the request interface) to input the obtained top N scheme knowledge and the application online business intent into the AI network large model, so that the AI network large model performs context reasoning according to the top N network opening scheme with the highest similarity and the application online business intent, and further outputs the recommended scheme, i.e., the target network opening scheme.

[0243] In the embodiment of the present application, if the knowledge of the plurality of network opening schemes fails to match the first natural language (i.e., no candidate knowledge is obtained), it indicates that the similarity between the knowledge of the plurality of network opening schemes and the first natural language is too low. The electronic device can directly input the first natural language (i.e., the application online service intent input by the user) into the AI network large model. The AI network large model only performs context reasoning according to the first natural language and the pre-learned knowledge to obtain the target network opening scheme, thereby further improving the adaptation capability of the application online network opening method provided in the embodiment of the present application to the scene and networking.

[0244] Here, the pre-learned knowledge can be: the knowledge input by the electronic device into the AI large language model in the training stage or the pre-training stage.

[0245] In some embodiments, the electronic device can fine-tune the AI network large model. The electronic device obtains a sample natural language and a sample network opening scheme corresponding to the sample natural language; and fine-tunes the AI network large model by using the sample natural language and the sample network opening scheme. The sample natural language and the corresponding sample network opening scheme are samples of the application online scene, and one sample is one labeled data. The content of the sample is the sample natural language, and the label of the sample is the real network opening scheme corresponding to the sample natural language, i.e., the sample network opening scheme.

[0246] The electronic device can pre-adopt samples (i.e., a plurality of sample natural languages and corresponding sample network opening schemes) of a plurality of application online scenes to construct a sample set (i.e., a labeled data set) of the application online scene. The sample set is actually a plurality of question and answer pairs. The question is the sample natural language, and the answer is the sample network opening scheme. The electronic device inputs the sample set into the AI network large model to realize the learning of the AI network large model on these question and answer pairs, and complete the fine-tuning of the AI network large model.

[0247] In the embodiment of the present application, the electronic device can continuously update the sample set of the application online scene, that is, the electronic device can continuously obtain new sample natural languages and corresponding sample network opening schemes to fine-tune the AI network large model.

[0248] The electronic device can also continuously update the DCN corpus. The electronic device can train and update the AI network large model by using the updated DCN corpus, rich application online scenes, and a continuously updated scheme knowledge base. The manner of training and updating the AI network large model is not limited herein.

[0249] By adopting the technical solutions provided in the embodiments of the present application, the AI network large model is fine-tuned by using the annotated data set of the application online scene, so that the AI network large model is more suitable for the application online scene, and then the electronic device can realize self-evolution and evolution when opening the network in the application online scene. In addition, the electronic device can also use the continuously updated scheme knowledge base, new DCN corpus and rich application online scene to train and update the AI network large model, realize self-evolution and evolution of the recommendation and generation ability of the network opening scheme in the application online scene, and further improve the ability of the electronic device to realize self-evolution and evolution.

[0250] As shown in FIG. 11, an example of the intention analysis stage processing logic, taking an electronic device as an application online intention management system for example. In the data center self-intelligent network, the application online intention management system receives the application online business intention input by the customer (i.e. user) (i.e. the first natural language), and performs application online intention perception, application online intention analysis, application online intention decision and application online intention execution. In the intention analysis stage, the application online intention management system analyzes and extracts features (i.e. vectorization) of the application online business intention, obtains the application online intention features (i.e. the vectorized first natural language), and inputs the application online intention features into the network large model (such as the AI network large model).

[0251] In the embodiments of the present application, when deploying the AI network large model, the scheme knowledge base and the agent (AI Agent) framework that assist the AI network large model can also be deployed. The application online intention management system outputs the network opening scheme (i.e. the target network opening scheme) based on the scheme knowledge base, the agent and the AI network large model.

[0252] Among them, the scheme knowledge base can assist the AI network large model to perform context reasoning in less but most relevant knowledge to obtain the target network opening scheme. The agent framework can call the rich API of the SDN controller based on the tool set, can realize the graph presentation ability and other abilities of the application online scene scheme, and the agent framework can continuously enrich the tool set for the interaction between the application online scene and the SDN controller, and improve the intention execution ability of the application online scene.

[0253] As shown in FIG. 11, the large model training method can be: training the general large model by using the DCN corpus, fine-tuning the large model obtained by training by using the annotated data set of the application online scene, finally obtaining the reliable AI network large model, and publishing.

[0254] In some embodiments, the electronic device can display the network opening scheme based on the human-computer interaction mode. Referring to FIG. 12, FIG. 12 is a flowchart of displaying the network opening scheme provided by the embodiments of the present application, which can include the following steps.

[0255] In step S121, the second natural language indicating the display mode of the network opening solution is collected based on the human-computer interaction mode.

[0256] In step S122, the second natural language is input into the AI network large model to obtain a target display mode.

[0257] In step S123, the interface corresponding to the target display mode is called by the intelligent agent to display the target network opening solution.

[0258] In the technical scheme provided by the embodiments of the present application, based on the strong natural language processing capability of the AI network large model, the natural language (i.e., the second natural language) of the display mode expressing the user's demand is analyzed and processed by using the AI network large model to obtain the target display mode, which improves the interactivity, and the corresponding API is called by the intelligent agent to realize the display of the target network opening solution in the target display mode.

[0259] In the above step S121, the second natural language is any natural language used to indicate the display mode of the network opening solution, and the second natural language can be text or voice, etc. The display mode of the network opening solution can include display in the form of text and display in the form of graphics, etc. Here, the form of the second natural language and the display mode of the network opening solution are not limited.

[0260] The electronic device can provide multiple human-computer interaction modes for the user, and the user inputs the second natural language to the electronic device through multiple human-computer interaction modes, and then the electronic device collects the second natural language input by the user based on multiple human-computer interaction modes.

[0261] When the electronic device provides a human-computer interaction dialog box, the above step S121 can be: displaying the human-computer interaction dialog box; receiving the second natural language input by the user in the human-computer interaction dialog box to indicate the display mode of the network opening solution. In this case, the electronic device can also display prompt information (i.e., second prompt information) for guiding the input of the display mode of the network opening solution on the human-computer interaction dialog box, such as the second prompt information can be "please enter the display mode", etc. Here, the content of the second prompt information is not limited. Through the human-computer interaction dialog box, the interaction between the user and the electronic device can be directly displayed, and by displaying the prompt information on the human-computer interaction dialog box, the user's input can be guided, the user's demand can be met, and the user's operation is facilitated.

[0262] In the embodiments of the present application, the way in which the electronic device collects the second natural language based on the human-computer interaction mode is similar to the way in which the first natural language is collected, and specific details can be referred to the related description of step S71.

[0263] In step S122, the electronic device inputs the collected second natural language into the AI network large model, so that the AI network large model outputs a display mode meeting the user's demand based on the second natural language, and the electronic device takes the display mode as a target display mode.

[0264] In the embodiments of the present application, the electronic device can call an interface (i.e., a first interface) corresponding to the AI network large model, and input the second natural language into the AI network large model, which is not limited herein.

[0265] In step S123, the electronic device adopts an agent, calls an interface (i.e., a second interface) corresponding to the target display mode, and realizes display of the target network opening scheme based on the target display mode.

[0266] For example, if the target display mode is to display in the form of graphics (i.e., graphical display), the electronic device adopts an agent, calls an interface corresponding to the graphical display, and displays the target network opening scheme in the form of graphics; if the target display mode is to display in the form of text, the electronic device adopts an agent, calls an interface corresponding to the textual display, and displays the target network opening scheme in the form of text.

[0267] As shown in FIG. 13, which is a schematic diagram of graphical display using an agent framework, the electronic device obtains a request (i.e., a second natural language) of the user for graphical display of the network opening scheme, inputs the request into the agent framework, and outputs a graphical display mode by AI network large model reasoning and decision-making on the request. The agent selects a preset tool to call a visual scheme API (i.e., a second interface), and outputs the graphical display of the network opening scheme.

[0268] In the embodiments of the present application, the electronic device can execute steps S121-S123 after obtaining the target network opening scheme to display the obtained target network opening scheme. The electronic device can also execute steps S121-S122 to obtain the target display mode before obtaining the target network opening scheme, and execute step S123 to display the obtained target network opening scheme after obtaining the target network opening scheme. For example, the electronic device can collect the first natural language and the second natural language at the same time, and input the first natural language and the second natural language into the AI network large model to obtain the target network opening scheme and the target display mode. The electronic device can also collect the second natural language after collecting the first natural language. The execution time of steps S121-S123 is not limited herein.

[0269] In some embodiments, the step S73 can be implemented by the following steps: adjusting the target network opening scheme based on user operations; and performing simulation evaluation on the adjusted target network opening scheme. The user operations can be operations such as dragging and editing, and the user operations are not limited herein. After the step S123 of displaying the target network opening scheme is performed, the electronic device can provide multiple human-computer interaction modes for the user, and the user can operate the displayed target network opening scheme through the multiple human-computer interaction modes. Then, the electronic device can collect the user operations based on the multiple human-computer interaction modes, and adjust the displayed target network opening scheme according to the user operations.

[0270] For example, when the target display mode is to display in the form of a graph, the electronic device can perform human-computer interaction based on a graphical interface, collect the operations such as dragging and editing performed by the user on the displayed target network opening scheme in the graphical interface, and perform corresponding operations such as dragging and editing on the displayed target network opening scheme according to the collected operations, so as to adjust the target network opening scheme. In the embodiments of the present application, the electronic device can also adjust the displayed target network opening scheme based on a human-computer interaction dialog box, and the adjustment is not limited herein.

[0271] The electronic device adjusts the displayed target network opening scheme based on the user operations, so that the adjusted target network opening scheme is more consistent with the user's business intention and meets the user's expectation. Then, the electronic device performs simulation evaluation on the adjusted target network opening scheme.

[0272] In some embodiments, the AI network large model can output multiple target network opening schemes in the step S72, and the step S73 can include the following steps, as shown in FIG. 14, which is a first flowchart of the step S73 according to an embodiment of the present application.

[0273] In the step S141, one target network opening scheme is determined from the multiple target network opening schemes.

[0274] In the step S142, simulation evaluation is performed on the determined target network opening scheme.

[0275] In the technical scheme provided in the embodiments of the present application, when the AI network large model outputs multiple target network opening schemes, the electronic device can determine one target network opening scheme that is most consistent with the user's application online business intention from the multiple target network opening schemes, and the accuracy of the network opening scheme generation and recommendation is further improved.

[0276] In step S141, the electronic device determines one target network opening scheme from the plurality of target network opening schemes output by the AI network large model. For example, the electronic device can receive information indicating the determined target network opening scheme input by the user, and determine one target network opening scheme according to the information input by the user. For another example, the electronic device can randomly determine one target network opening scheme from the plurality of target network opening schemes, or the electronic device can determine one target network opening scheme from the plurality of target network opening schemes according to a preset rule. The preset rule can be set according to actual conditions, and the preset rule can be to select the safest network opening scheme. The manner in which the electronic device determines one target network opening scheme is not limited herein.

[0277] In step S142, the electronic device performs simulation evaluation on the determined target network opening scheme. For details, refer to the related description in step S73.

[0278] In the embodiments of the present application, before step S141, the electronic device can display the plurality of target network opening schemes output by the AI network large model. For details of the display manner, refer to the related description in FIG. 12. After displaying the plurality of target network opening schemes, the electronic device can further fine-tune the displayed plurality of target network opening schemes, determine one fine-tuned target network opening scheme from the fine-tuned plurality of target network opening schemes, and perform simulation evaluation on the determined fine-tuned target network opening scheme.

[0279] In some embodiments, the electronic device can determine one target network opening scheme based on a human-computer interaction manner. Referring to FIG. 15, FIG. 15 is a detailed schematic diagram of step S141 provided by the embodiments of the present application, which can include the following steps.

[0280] In step S151, the third natural language indicating the determined target network opening scheme is collected based on the human-computer interaction manner.

[0281] In step S152, the third natural language is input into the AI network large model to obtain the determined target network opening scheme.

[0282] In the technical scheme provided by the embodiments of the present application, based on the strong natural language processing capability of the AI network large model, the AI network large model is used to analyze and process the natural language (i.e. the third natural language) expressing the user's demand and determining the target network opening scheme, to obtain the determined target network opening scheme, thereby improving the interactivity, making the determined target network opening scheme more consistent with the user's expectation, and improving the user experience.

[0283] In step S151, the third natural language is any natural language used to indicate the determined target network opening scheme, and the third natural language can be text or voice, and the form of the third natural language is not limited herein.

[0284] The electronic device can provide multiple human-computer interaction modes for the user, and the user inputs the third natural language to the electronic device through the multiple human-computer interaction modes, and then the electronic device collects the third natural language input by the user based on the multiple human-computer interaction modes.

[0285] When the electronic device provides the human-computer interaction dialog box, step S151 can be: displaying the human-computer interaction dialog box; receiving the third natural language input by the user on the human-computer interaction dialog box to indicate the determined target network opening scheme. In this case, the electronic device can also display prompt information (i.e., third prompt information) for guiding the input of the determined target network opening scheme on the human-computer interaction dialog box. For example, the third prompt information can be "please enter the determined network opening scheme" or "please enter the tendency of the overall requirement for the network opening scheme", and the content of the third prompt information is not limited herein. Through the human-computer interaction dialog box, the interaction between the user and the electronic device can be intuitively displayed, and by displaying the prompt information on the human-computer interaction dialog box, the user can be guided to input the related information of the determined target network opening scheme, which meets the user's demand and facilitates the user's operation.

[0286] In the embodiments of the present application, the electronic device collects the third natural language based on the human-computer interaction mode in a manner similar to the manner of collecting the first natural language, and the specific description can be referred to the related description of step S71.

[0287] In step S152, the electronic device inputs the collected third natural language into the AI network large model, so that the AI network large model outputs a target network opening scheme based on the third natural language, and the electronic device takes the target network opening scheme as the determined target network opening scheme.

[0288] In the embodiments of the present application, the electronic device can call the interface (i.e., the first interface) corresponding to the AI network large model to input the third natural language into the AI network large model, and the input is not limited herein.

[0289] In the embodiments of the present application, according to the execution time of step S151 and step S152, the following two cases can be divided.

[0290] Case 1: After obtaining multiple target network opening schemes, the electronic device executes step S151 and step S152 to collect the third natural language and obtain the determined target network opening scheme.

[0291] In case 2, before obtaining the plurality of target network opening solutions, the electronic device performs step S151 to collect a third natural language. For example, the electronic device can collect the third natural language at the same time of collecting the first natural language, in which case, the electronic device can perform step S72 and step S152 at the same time, input the first natural language and the third natural language into the AI network large model to obtain the determined target network opening solution. That is, when the user input application online business intent has the tendency of the overall requirement of the network opening solution, the electronic device can input all the content input by the user including the first natural language and the third natural language into the AI network large model, so that the AI network large model outputs a target network opening solution to obtain the determined target network opening solution. The execution time of step S151 and step S152 is not limited herein.

[0292] The electronic device can also receive a user operation of selecting a target network opening solution on the electronic device, such as clicking, switching, etc., and determine a target network opening solution from the plurality of target network opening solutions in response to the user operation. The manner in which the electronic device determines a target network opening solution is not limited herein.

[0293] In some embodiments, the electronic device can perform simulation evaluation on the target network opening solution based on the human-computer interaction mode. Referring to FIG. 16, which is a second flowchart of step S73 provided by an embodiment of the present application, step S73 can include the following steps.

[0294] Step S161, collecting a fourth natural language indicating simulation evaluation based on the human-computer interaction mode.

[0295] Step S162, inputting the fourth natural language into the AI network large model to obtain a simulation evaluation instruction.

[0296] Step S163, calling an interface corresponding to the simulation evaluation instruction by the agent to perform simulation evaluation on the target network opening solution.

[0297] In the technical solution provided by the embodiments of the present application, based on the strong natural language processing capability of the AI network large model, the AI network large model is used to analyze and process the natural language (i.e. the fourth natural language) used to express simulation evaluation to obtain a simulation evaluation instruction, and the corresponding API is called by the agent to realize simulation evaluation on the target network opening solution.

[0298] In step S161, the fourth natural language is any natural language used to indicate simulation evaluation, and the fourth natural language can be text or voice, etc. The form of the fourth natural language is not limited herein.

[0299] The electronic device can provide a plurality of human-computer interaction modes for the user, and the user inputs a fourth natural language to the electronic device through the plurality of human-computer interaction modes, and then the electronic device collects the fourth natural language input by the user based on the plurality of human-computer interaction modes

[0300] When the electronic device provides the human-computer interaction dialog box, the above step S161 can be: displaying the human-computer interaction dialog box; and receiving the fourth natural language input by the user on the human-computer interaction dialog box to indicate the simulation evaluation. In this case, the electronic device can also display prompt information (i.e., fourth prompt information) for guiding the input of execution information of the simulation evaluation on the human-computer interaction dialog box. For example, the fourth prompt information can be "please input whether to perform simulation evaluation", and the content of the fourth prompt information is not limited herein. Through the human-computer interaction dialog box, the interaction between the user and the electronic device can be intuitively displayed, and by displaying the prompt information on the human-computer interaction dialog box, the user can be guided to input the related information of whether to perform the simulation evaluation, which meets the user's demand and facilitates the user's operation.

[0301] In the embodiments of the present application, the electronic device collects the fourth natural language based on the human-computer interaction mode in a manner similar to the above-mentioned manner of collecting the first natural language, and the specific description can be referred to the above-mentioned description of step S71.

[0302] In the above step S162, the electronic device inputs the collected fourth natural language into the AI network large model, so that the AI network large model outputs the simulation evaluation instruction based on the fourth natural language.

[0303] In the embodiments of the present application, the electronic device can call the interface (i.e., the first interface) corresponding to the AI network large model, and input the fourth natural language into the AI network large model, which is not limited herein.

[0304] In the above step S163, the electronic device uses the agent to call the interface (i.e., the third interface) corresponding to the simulation evaluation instruction, to realize the simulation evaluation on the target network opening scheme.

[0305] In some embodiments, after the simulation evaluation result meets the preset condition, the electronic device can open the network based on the human-computer interaction mode. Referring to FIG. 17, FIG. 17 is a flow diagram of step S74 provided by the embodiments of the present application, and the above step S74 can include the following steps.

[0306] Step S171: collecting a fifth natural language indicating network opening based on the human-computer interaction mode.

[0307] Step S172: inputting the fifth natural language into the AI network large model to obtain a network opening instruction.

[0308] In step S173, the intelligent agent calls an interface corresponding to the network opening instruction to open the network by using the target network opening scheme.

[0309] In the technical scheme provided by the embodiments of the present application, based on the strong natural language processing capability of the AI network large model, the AI network large model is used to analyze and process the natural language (i.e., the fifth natural language) for expressing network opening, to obtain a network opening instruction, and an intelligent agent is used to call a corresponding API to open the network by using a target network opening scheme, thereby realizing application online.

[0310] In the above step S171, the fifth natural language is any natural language for indicating network opening, and the fifth natural language can be text or voice, etc., and the form of the fifth natural language is not limited herein.

[0311] The electronic device can provide multiple human-computer interaction modes for the user, and the user inputs the fifth natural language to the electronic device through multiple human-computer interaction modes, and then the electronic device collects the fifth natural language input by the user based on multiple human-computer interaction modes.

[0312] When the electronic device provides a human-computer interaction dialog box, the above step S171 can be: displaying the human-computer interaction dialog box; receiving the fifth natural language input by the user on the human-computer interaction dialog box to indicate network opening. In this case, the electronic device can also display prompt information (i.e., the fifth prompt information) for guiding the input of execution information of network opening on the human-computer interaction dialog box. For example, the fifth prompt information can be "please input whether to open the network and deploy the application", etc., and the content of the fifth prompt information is not limited herein. Through the human-computer interaction dialog box, the interaction between the user and the electronic device can be directly displayed, and by displaying the prompt information on the human-computer interaction dialog box, the user can be guided to input the related information of whether to open the network, thereby meeting the user's demand and facilitating the user's operation.

[0313] In some embodiments, the electronic device can also determine an operation to be performed, and display prompt information associated with the operation to be performed on the man-machine interaction dialogue box. The electronic device can detect the interface displayed to the user (or the stage at which the scheme execution is located) and the content input by the user (or the operation performed) in real time, determine the operation to be performed by the electronic device, and display the prompt information associated with the operation to be performed on the man-machine interaction dialogue box according to the operation to be performed. For example, when the electronic device detects that the interface displayed to the user is an intent analysis interface (i.e., in the intent analysis stage), and detects that the content input by the user on the man-machine interaction dialogue box last time is a natural language (i.e., a second natural language) indicating the display mode of the network opening scheme, the electronic device can determine that the operation to be performed is to determine a target network opening scheme (in the case that the AI network large model outputs multiple target network opening schemes) from the target network opening scheme, and display third prompt information guiding the input of the determined target network opening scheme on the man-machine interaction dialogue box.

[0314] In the embodiments of the present application, the electronic device collects the fifth natural language based on the man-machine interaction mode in a manner similar to the manner of collecting the first natural language described above, and specific details can be referred to the related description of step S71.

[0315] In step S172 described above, the electronic device inputs the collected fifth natural language into the AI network large model, so that the AI network large model outputs a network opening instruction based on the fifth natural language.

[0316] In the embodiments of the present application, the electronic device can call an interface (i.e., a first interface) corresponding to the AI network large model, and input the fifth natural language into the AI network large model, which is not limited.

[0317] In step S173 described above, the electronic device uses the agent, calls an interface (i.e., a fourth interface) corresponding to the network opening instruction, and opens the network using the target network opening scheme to realize the application online.

[0318] In the embodiments of the present application, the electronic device can display a dialogue box icon on the interface, and after the user clicks the dialogue box icon, the electronic device responds to the operation of the user clicking the dialogue box icon, calls a front-end general interface (i.e., a fifth interface) corresponding to the operation of clicking the dialogue box icon, and displays the man-machine interaction dialogue box.

[0319] In the embodiments of the present application, the electronic device can also continuously update the tool set for man-machine interaction, i.e., continuously update the interface, so that the electronic device can call more abundant interfaces and improve the ability of using the agent to open the network in the application online scenario.

[0320] The application online network opening method provided by the embodiments of the present application is described in detail below based on the network opening method framework for application online shown in FIG. 18 and the SDN controller interfaces shown in FIGS. 19a-19c, 20a-20e, 21 and 22a-22b. The electronic device is taken as the SDN controller, and the SDN controller is taken as an example of human-computer interaction based on a human-computer interaction dialog box, which does not have a limiting effect.

[0321] FIG. 18 is a structural schematic diagram of a network opening method framework for application online provided by the embodiments of the present application. As shown in FIG. 18, the application online intent management system (i.e., the SDN controller) of the control plane of the self-intelligent network in the data center receives the application online service intent (i.e., the first natural language) input by the customer (i.e., the user), and performs application online intent perception, intent analysis, intent decision and intent execution based on the AI network large model, the scheme knowledge base and the intelligent agent. After determining the network opening scheme (i.e., the target network opening scheme), the SDN controller performs configuration and delivery to the forwarding plane of the self-intelligent network, and collects information from the forwarding plane of the self-intelligent network. The forwarding plane of the self-intelligent network is a leaf-spine architecture, including multiple spine nodes, leaf nodes and servers. After the network opening for application online, the forwarding path between the communicating servers is opened, and the servers complete interaction through the spine nodes and the leaf nodes, thereby realizing the application online.

[0322] The four stages of the network opening for application online provided by the present application are described in detail below in combination with the SDN controller interfaces shown in FIGS. 19a-19c, 20a-20e, 21 and 22a-22b.

[0323] 1) Intent perception stage

[0324] FIGS. 19a-19c are a first schematic diagram of the SDN controller interface provided by the embodiments of the present application. FIGS. 19a-19c are all the SDN controller interfaces in the intent perception stage. As shown in FIGS. 19a-19c, the SDN controller interface mainly includes a stage column 191, a model selection icon 196, a display column 197 and an interaction icon 198. The stage column 191 includes an intent perception stage tab 192, an intent analysis stage tab 193, an intent decision stage tab 194 and an intent execution stage tab 195, which respectively correspond to the four stages of the network opening. The intent perception stage tab 192 is in a bold state, indicating that the current stage is the intent perception stage.

[0325] When the SDN controller interface shown in FIG. 19a is displayed, the user can click the model selection icon 196. The SDN controller displays a plurality of models selectable by the user for intent analysis (i.e., recommending a network opening scheme) in response to the user's operation of clicking the model selection icon 196. After the user clicks to select a model, the SDN controller switches the model for intent analysis in response to the user's operation of clicking to select the model. In FIG. 19a, the currently used model is the AI network large model.

[0326] The user can click the interaction icon 198 (i.e., the dialog box icon) on the SDN controller interface. The SDN controller invokes the front-end general API (i.e., the fifth interface) corresponding to the user's operation of clicking the interaction icon 198 and displays the man-machine natural language interaction interface (i.e., the man-machine interaction dialog box) in the display bar 197 in response to the user's operation of clicking the interaction icon 198. That is, the man-machine natural language interaction interface 199 shown in FIG. 19b. The man-machine natural language interaction interface 199 can include the confirmation button 1910.

[0327] The SDN controller can detect the to-be-executed operation in real time or event-triggered. The event triggering the detection of the to-be-executed operation can be the operation event of the user clicking the interface, the operation event (such as the intent analysis event) of the SDN controller or the AI network large model completing the operation, and the like.

[0328] When the SDN controller detects that the current stage is the intent perception stage and determines that the user has not performed any operation, the SDN controller can determine that the to-be-executed operation is to collect the business intent of the application going online. Then, the SDN controller displays the prompt information (i.e., the first prompt information) associated with the to-be-executed operation on the man-machine natural language interaction interface 199, such as the prompt information displayed in the man-machine natural language interaction interface 199 shown in FIG. 19b.

[0329] The user can input the natural language (i.e., the first natural language) expressing the business intent of the application going online in the man-machine natural language interaction interface 199 according to the prompt information in the man-machine natural language interaction interface 199, such as the natural language displayed in the man-machine natural language interaction interface 199 shown in FIG. 19c. The user clicks the confirmation button 1910. The SDN controller collects the natural language in the man-machine natural language interaction interface 199 as the business intent of the application going online in response to the user's operation of clicking the confirmation button 1910. The prompt information and the business intent shown in FIGS. 19b and 19c are only examples and are not limiting in this regard.

[0330] In the embodiments of the present application, the interactive icon 198 and the man-machine natural language interaction interface 199 can be suspended on the display bar 197, that is, the user can drag the interactive icon 198 and the man-machine natural language interaction interface 199 on the SDN controller interface, and the SDN controller responds to the user's dragging operation, calls the front-end general API corresponding to the dragging operation, and moves the interactive icon 198 and the man-machine natural language interaction interface 199.

[0331] The user can click the interactive icon 198 again or click a blank position of the display bar 197, and the SDN controller responds to the user's operation of clicking the interactive icon 198 again or the operation of the user clicking the blank position of the display bar 197, calls the front-end general API corresponding to the clicking operation, and closes the man-machine natural language interaction interface 199.

[0332] The man-machine natural language interaction interface 199 on the SDN controller is not limited to the SDN controller interface in the intent perception stage, that is, the man-machine natural language interaction interface 199 can also be displayed on the SDN controller interface in the intent analysis stage, the intent decision stage and the intent execution stage.

[0333] The SDN controller provides the man-machine natural language interaction interface on the SDN controller interface of the data center self-intelligent network, integrates the ability of the AI network large model through API calling, realizes the collection of the application online business intent input by the user in the form of natural language, and predefines the prompt information of the input of the application online business intent in the man-machine natural language interaction interface, so as to guide the user to make efficient application online business intent input.

[0334] 2) Intent analysis stage

[0335] In the embodiments of the present application, the following several ways can be used to enter the intent analysis stage.

[0336] Method one, after the user clicks the confirmation button 1910 in the intent perception stage, the application online network opening scheme jumps to the intent analysis stage.

[0337] Specifically, after the user inputs the application online business intent in the man-machine natural language interaction interface 199 and clicks the confirmation button 1910, the SDN controller responds to the user's operation of clicking the confirmation button 1910, calls the front-end general API corresponding to the jump to the corresponding interface, jumps to the SDN controller interface in the intent analysis stage, and enters the intent analysis stage. At this time, the application online business intent is the natural language in the man-machine natural language interaction interface 199 when the user clicks the confirmation button 1910.

[0338] Method two, the user clicks the intent analysis stage tab 193, and the application online network opening scheme jumps to the intent analysis stage.

[0339] Specifically, after the user inputs the business intent of application online in the man-machine natural language interaction interface 199 and clicks the confirmation button 1910, the user can click the intent analysis stage tab 193. The SDN controller responds to the user's click operation, calls the front-end general API for jumping to the corresponding interface, jumps to the SDN controller interface of the intent analysis stage, and enters the intent analysis stage. At this time, the business intent of application online is the natural language in the man-machine natural language interaction interface 199 when the user clicks the intent analysis stage tab 193.

[0340] After entering the intent analysis stage, the SDN controller calls the API (i.e., the first interface) corresponding to the AI network large model, inputs the natural language displayed in the man-machine natural language interaction interface 199 into the AI network large model, performs intent analysis, and obtains a network opening scheme.

[0341] FIGS. 20a-20e are a second kind of schematic diagram of the SDN controller interface provided by the embodiments of the present application. FIGS. 20a-20e are all the SDN controller interfaces of the intent analysis stage, and the intent analysis stage tab 193 is in a bold state.

[0342] The man-machine natural language interaction interface 199 can also be displayed on the SDN controller interface in the intent analysis stage. The opening and closing mode of the man-machine natural language interaction interface 199 can be referred to the related description in the intent perception stage.

[0343] In the case of opening the man-machine natural language interaction interface 199, the man-machine natural language interaction interface 199 can display historical question and answer information, such as the historical problem information 201 (i.e., the business intent) input by the user into the AI network large model as shown in FIG. 20a; and the historical answer information 202 (i.e., the scheme description of the target network opening scheme recommended by the AI network large model) output by the AI network large model as shown in FIG. 20a.

[0344] In the embodiments of the present application, the business intent of application online input by the user into the AI network large model can include the tendency of the overall requirement of the network opening scheme, or can not include the tendency of the overall requirement of the network opening scheme; the content of the historical answer information 202 is determined according to whether the historical problem information 201 has the tendency requirement.

[0345] For example, in the case that the application online business intention includes the tendency of the overall requirement of the network opening scheme, taking the tendency of the overall requirement of the network opening scheme as the highest security as an example, the AI network large model can generate a network opening scheme based on the application online business intention with the highest security requirement, which is a network opening scheme with the highest security. The historical answer information output by the AI network large model shows the scheme description of the network opening scheme, such as the scheme description in the historical answer information 202 shown in FIG. 20a, that is, 5 services are divided into 5 security partitions, maintaining isolation, with the highest security; each service in the security partition exclusively occupies a virtual router, with relatively high resource consumption; services visit each other through virtual router connection configuration; and services implement external network services through external network binding of virtual router configuration.

[0346] For another example, in the case that the application online business intention does not include the tendency of the overall requirement of the network opening scheme, the AI network large model can generate multiple network opening schemes based on the application online business intention. At this time, the historical answer information output by the AI network large model can include the scheme description of the multiple network opening schemes.

[0347] In the embodiment of the application, the SDN controller can detect that the current stage is the intention analysis stage, and determine that the application online business intention is collected, and then determine that the to-be-executed operation is to display the network opening scheme. Correspondingly, the SDN controller displays the prompt information (i.e., the second prompt information) associated with the to-be-executed operation on the man-machine natural language interaction interface 199, as shown in the prompt information displayed in the man-machine natural language interaction interface 199 in FIG. 20a.

[0348] The user can input the natural language (i.e., the second natural language) expressing the display method in the man-machine natural language interaction interface 199 according to the prompt information in the man-machine natural language interaction interface 199, and click the confirmation button 1910. Taking the case that the user inputs the natural language expressing the graphical display in the man-machine natural language interaction interface 199 as an example. After the user inputs the natural language expressing the graphical display, the historical question and answer information of the man-machine natural language interaction interface 199 can increase the corresponding historical information, such as the historical question information 203 in FIG. 20b.

[0349] In addition, the user inputs the natural language expressing the graphical display mode in the man-machine natural language interaction interface 199 and clicks the confirmation button 1910. The SDN controller, in response to the operation of the user clicking the confirmation button 1910, calls the API (i.e., the first interface) corresponding to the AI network large model, inputs the natural language input by the user into the AI network large model, so that the AI network large model outputs the corresponding display mode; and then, after obtaining the network opening solution output by the AI network large model, the SDN controller calls the SDN controller API (i.e., the second interface / visual solution API) corresponding to the display mode by using the agent, and graphically displays the network opening solution, as shown in FIGS. 20c and 20d.

[0350] FIG. 20c is an interface schematic diagram of the intent analysis stage when the business intent of the application online includes the tendency of the overall requirement of the network opening solution, and FIG. 20d is an interface schematic diagram of the intent analysis stage when the business intent of the application online does not include the tendency of the overall requirement of the network opening solution. In FIGS. 20c and 20d, the SDN controller closes the man-machine natural language interaction interface 199. In addition, the display bar 197 can also include the solution description 204 of the network opening solution in the form of text, which is not limited.

[0351] The difference between FIGS. 20c and 20d is that when the business intent of the application online does not include the tendency of the overall requirement of the network opening solution, the display bar 197 also includes the tabs 205 corresponding to a plurality of network opening solutions, and different tabs correspond to different network opening solutions. As shown in FIG. 20d, the display bar 197 includes the tabs 205 corresponding to the four network opening solutions of the highest security, the lightest resource occupation, the highest application service instance deployment reliability, and the lowest inter-service access delay. Under different tabs, the display bar 197 graphically displays different network opening solutions. At present, the display bar 197 is located under the tab 205 corresponding to the network opening solution with the highest security, and the SDN controller graphically displays the network opening solution with the highest security.

[0352] In FIG. 20d, the user can click the tabs 205 corresponding to different network opening solutions. The SDN controller, in response to the operation of the user clicking the corresponding tab 205, calls the corresponding front-end general API, and displays the corresponding network opening solution to the user in the display bar 197, so as to facilitate the user to view by himself / herself.

[0353] In the embodiment of the present application, in the case where the user does not input the display mode, the SDN controller can display the network opening solution by using a preset display mode, for example, display the network opening solution in the form of text, as shown in FIG. 20e, the display bar 197 displays the solution description 204 of the network opening solution in the form of text.

[0354] In addition, the user can perform drag, edit, and the like on the graphically displayed network opening scheme in the display bar 197, and the SDN controller fine-tunes the network opening scheme in response to the user's drag, edit, and the like.

[0355] After displaying the network opening scheme, the SDN controller can determine that the operation to be performed is to determine the network opening scheme, and then the prompt information (i.e., third prompt information) guiding the input of determining the network opening scheme can be displayed on the human-computer natural language interaction interface 199.

[0356] The user can input the natural language (i.e., third natural language) indicating the determination of the network opening scheme in the human-computer natural language interaction interface 199 according to the prompt information (i.e., third prompt information) guiding the input of determining the network opening scheme displayed on the human-computer natural language interaction interface 199, and click the confirmation button 1910. The SDN controller, in response to the user's operation of clicking the confirmation button 1910, calls the API (i.e., first interface) corresponding to the AI network large model, inputs the natural language input by the user into the AI network large model, so that the AI network large model outputs the determined network opening scheme. The SDN controller can call the corresponding API to display the determined network opening scheme in the display bar 197 by using the intelligent agent.

[0357] If the displayed network opening scheme does not meet the expectation, the user can continue to input the third natural language until the displayed network opening scheme meets the expectation. The operation here is the same as the effect of the user clicking the tab 205 corresponding to a different network opening scheme to switch the display of different network opening schemes.

[0358] After determining the network opening scheme, the SDN controller can determine that the operation to be performed is simulation evaluation, and then the prompt information (i.e., fourth prompt information) guiding the execution of simulation evaluation can be displayed on the human-computer natural language interaction interface 199, as shown in the prompt information in the human-computer natural language interaction interface 199 in FIG. 20b.

[0359] The user can input the natural language (i.e., fourth natural language) indicating the execution of simulation evaluation in the human-computer natural language interaction interface 199 according to the prompt information in the human-computer natural language interaction interface 199.

[0360] In the embodiment of the present application, in the case where the application online business intention includes the tendency of the overall requirement of the network opening scheme, the AI network large model outputs one network opening scheme. At this time, the SDN controller can omit the display process of the third prompt information and directly display the fourth prompt information.

[0361] In the embodiment of the present application, the default processing flow of the SDN controller in the intent analysis stage is: first, the network opening scheme output by the AI network large model is displayed, then the network opening scheme for simulation evaluation is determined, and then simulation evaluation is performed. The SDN controller can also change the processing flow according to the input of the user. For example, when the prompt information (i.e., the second prompt information) guiding the display mode is displayed on the man-machine natural language interaction interface 199, the user can input the natural language (i.e., the third natural language) for determining the network opening scheme to determine the network opening scheme. Then, when the prompt information (i.e., the fourth prompt information) guiding the input of the execution information of the simulation evaluation is displayed on the man-machine natural language interaction interface 199, the user can input the display mode (i.e., the second natural language) to display the network opening scheme. This is not limited.

[0362] In actual execution, the default processing flow of the SDN controller in the intent analysis stage can be set according to actual needs.

[0363] 3) Intent decision stage

[0364] In the embodiment of the present application, the following several ways can be used to enter the intent decision stage.

[0365] Method one, the user inputs the fourth natural language in the man-machine natural language interaction interface 199 and clicks the confirmation button 1910, and the application online network opening scheme jumps to the intent decision stage.

[0366] Specifically, the user can input the natural language (i.e., the fourth natural language) indicating simulation evaluation in the man-machine natural language interaction interface 199 and click the confirmation button 1910. The SDN controller responds to the operation of the user clicking the confirmation button 1910, calls the API (i.e., the first interface) corresponding to the AI network large model, inputs the natural language input by the user into the AI network large model, so that the AI network large model outputs the corresponding simulation evaluation instruction, and then the intelligent agent is used. The SDN controller calls the corresponding API, jumps to the SDN controller interface in the intent decision stage, enters the intent decision stage, and performs simulation evaluation on the determined network opening scheme.

[0367] Method two, the user clicks the intent decision stage tab 194, and the application online network opening scheme jumps to the intent decision stage.

[0368] Specifically, in the intent analysis stage, if the user does not input the fourth natural language, the user can click the intent decision stage tab 194. The SDN controller responds to the operation of the user clicking the intent decision stage tab 194, calls the front-end general API for jumping to the corresponding interface, jumps to the SDN controller interface in the intent decision stage, enters the intent decision stage, and performs simulation evaluation on the determined network opening scheme.

[0369] In the embodiment, the SDN controller can regard the operation of clicking the intention decision stage tab 194 as an instruction of performing simulation evaluation. After jumping to the SDN controller interface of the intention decision stage, the SDN controller can directly perform simulation evaluation on the determined network opening scheme. Alternatively, the SDN controller can additionally input an instruction of performing simulation evaluation. After obtaining the instruction of performing simulation evaluation, the SDN controller performs simulation evaluation on the determined network opening scheme.

[0370] For example, after jumping to the SDN controller interface of the intention decision stage, the fourth natural language is input in the man-machine natural language interaction interface 199 in the first mode to obtain the instruction of performing simulation evaluation.

[0371] For another example, a start simulation button can be set on the SDN controller interface of the intention decision stage. After jumping to the SDN controller interface of the intention decision stage, the user clicks the start simulation button. The SDN controller generates the instruction of performing simulation evaluation in response to the operation of clicking the start simulation button.

[0372] In the embodiment, the determined network opening scheme is the network opening scheme displayed on the display bar 197 when the user clicks the confirmation button 1910 or the intention decision stage tab 194. The network opening scheme can be determined by the user by clicking the tabs 205 corresponding to different network opening schemes in FIG. 20d, or can be determined by the user by inputting the natural language (i.e., the third natural language) indicating the determination of the network opening scheme in the man-machine natural language interaction interface 199.

[0373] FIG. 21 is a third schematic diagram of the SDN controller interface provided in the embodiment. FIG. 21 is the SDN controller interface of the intention decision stage, and the intention decision stage tab 194 is in the bold state.

[0374] In FIG. 21, the display bar 197 includes a simulation interface 211 and a simulation report interface 212. The simulation interface 211 is used to display the topology structure of the network opening scheme after configuration generation, and the simulation report interface 212 is used to display the simulation evaluation result. The user can view the simulation evaluation report in the simulation report interface 212.

[0375] When the simulation evaluation report shows that the simulation evaluation is successful (i.e., 100%), it indicates that the network opening scheme passes the simulation evaluation verification (i.e., meets the preset condition). In the embodiment, the user can click the report viewing button 213 in the simulation report interface 212 to view the specific simulation evaluation result, and fine-tune the current network opening scheme according to the specific simulation evaluation result, or enter the intention execution stage.

[0376] The SDN controller performs comprehensive simulation evaluation on the satisfaction degree of the online application to the existing network resources, the expected inter-service connectivity, and the impact on the existing network based on the simulation platform in the intent decision stage. The simulation evaluation result can be confirmed by checking the simulation report. After the simulation evaluation result meets the expectation, the intent execution stage can be entered.

[0377] 4) Intent execution stage

[0378] In the embodiments of the present application, the following methods can be used to enter the intent execution stage.

[0379] Method one, the user clicks the start execution button 214, and the application online network opening solution jumps to the intent execution stage.

[0380] For example, in FIG. 21, the simulation interface 211 can include a start execution button 214. The user can click the start execution button 214. The SDN controller responds to the user's operation of clicking the start execution button 214, calls the corresponding front-end general API, jumps to the SDN controller interface of the intent execution stage, enters the intent execution stage, and then calls the API (i.e., the fourth interface) corresponding to the network opening instruction to open the network and realize the application online.

[0381] Method two, the user inputs the fifth natural language in the man-machine natural language interaction interface 199 and clicks the confirmation button 1910, and the application online network opening solution jumps to the intent execution stage.

[0382] Specifically, the user can input the natural language (i.e., the fifth natural language) indicating opening the network in the man-machine natural language interaction interface 199 and click the confirmation button 1910. The SDN controller responds to the user's operation of clicking the confirmation button 1910, calls the API (i.e., the first interface) corresponding to the AI network large model, inputs the natural language input by the user into the AI network large model, so that the AI network large model outputs the corresponding network opening instruction, and then uses the agent. The SDN controller calls the corresponding front-end general API, jumps to the SDN controller interface of the intent execution stage, enters the intent execution stage, calls the API (i.e., the fourth interface) corresponding to the network opening instruction, opens the network, and realizes the application online.

[0383] Method three, the user clicks the intent execution stage tab 195, and the application online network opening solution jumps to the intent execution stage.

[0384] Specifically, the user can click the intent execution stage tab 195, and the SDN controller, in response to the operation of the user clicking the intent execution stage tab 195, calls a front-end general API for jumping to a corresponding interface, jumps to the SDN controller interface of the intent execution stage, enters the intent execution stage, and calls an API (i.e., the fourth interface) corresponding to the network opening instruction to open the network and realize the application going online.

[0385] In the embodiment of the application, the SDN controller can take the operation of the user clicking the intent execution stage tab 195 as a network opening instruction. After jumping to the SDN controller interface of the intent execution stage, the SDN controller can directly call the API corresponding to the network opening instruction to open the network, or can additionally input the network opening instruction. After obtaining the network opening instruction, the SDN controller calls the API corresponding to the network opening instruction to open the network.

[0386] For example, after jumping to the SDN controller interface of the intent execution stage, the fifth natural language is input in the man-machine natural language interaction interface 199 in mode two to achieve the purpose of obtaining the network opening instruction.

[0387] For another example, a start execution button can be arranged on the SDN controller interface of the intent execution stage, such as the start execution button 221 in the display bar 197 shown in FIG. 22a. After jumping to the SDN controller interface of the intent execution stage, the user clicks the start execution button 221. The SDN controller, in response to the operation of the user clicking the start execution button 221, generates a network opening instruction, calls the API (i.e., the fourth interface) corresponding to the network opening instruction to open the network, and realizes the application going online.

[0388] FIGS. 22a-22b are a fourth kind of schematic diagram of the SDN controller interface provided in the embodiment of the application. FIGS. 22a-22b are both the SDN controller interface of the intent execution stage, and the intent execution stage tab 195 is in a bold state.

[0389] In the process of intent execution (i.e., opening the network), the SDN controller can display the execution progress 222 on the display bar 197. As shown in FIG. 22b, the current execution progress is 70%, which has not reached 100%, indicating that the network opening has not been completed. At this time, the SDN controller can display a prompt information of please wait on the man-machine natural language interaction interface 199. When the intent execution is completed (i.e., the execution progress reaches 100%), the SDN controller confirms that the network opening is completed, that is, the application going online has been completed.

[0390] In the technical scheme provided in the embodiment of the application, the AI network large model is introduced, and the ability of man-machine natural language interaction based on the AI network large model solves the problem that the data center self-intelligent network lacks the understanding ability of the user business-level intent when collecting the application going online intent.

[0391] Through training the AI network large model by a large amount of DCN corpus, combining with the labeled data set in the application online scene for model fine-tuning, and deploying the AI network large model in the production environment to improve the ability of application online intent understanding, the problem that the data center self-intelligent network cannot understand the new network intent and the changed network intent in the application online scene is solved.

[0392] Through the AI network large model + scheme knowledge base mode, the private domain knowledge of the AI network large model in the application online scene is expanded, and the accuracy of the AI network large model in scheme recommendation and generation for the application online scene of the self-intelligent network is improved.

[0393] Through the agent framework, the rich API of the SDN controller is called based on the tool set, so as to realize the graphical presentation ability of the application online network opening scheme.

[0394] Through continuously enriching the network corpus and labeled data set of the application online scene of the data center self-intelligent network, the training and fine-tuning of the AI network large model are realized, the deployment mode of the AI network large model is continuously updated in the background, the self-evolution and evolution of the network opening scheme recommendation and generation ability of the data center self-intelligent network in the application online scene are realized; through cutting and vectorization processing of new knowledge and storing in the vector database, the scheme knowledge base is continuously supplemented and improved, through the agent framework, the tool set for interaction between the application online scene and the SDN controller is continuously enriched, the intent execution ability of the application online scene is improved, and finally the self-evolution and evolution of the network scheme recommendation and generation ability of the application online scene are realized.

[0395] Corresponding to the above network opening method of application online, an application online network opening device is provided in the embodiments of the present application, as shown in FIG. 23, which is a structural schematic diagram of the application online network opening device provided by the embodiments of the present application. The device comprises:

[0396] The collection module 231 is configured to collect a first natural language indicating an application online business intent based on a human-computer interaction mode;

[0397] The determination module 232 is configured to input the first natural language into an AI network large model to obtain a target network opening scheme;

[0398] The simulation module 233 is configured to simulate and evaluate the target network opening scheme;

[0399] The opening module 234 is configured to open the network according to the target network opening scheme after the simulation evaluation result meets a preset condition.

[0400] The technical scheme provided in the embodiments of the present application uses natural language to express the user's intention, that is, the application online service intention. The natural language is a language that is easy for users to understand and does not require a professional knowledge background. Therefore, the use of natural language to express the user's intention realizes the network opening of the application online, and increases the understanding ability of the user's service level intention. In addition, in the technical scheme provided in the embodiments of the present application, the AI network large model is used to analyze and process the natural language (that is, the first natural language) expressing the user's intention, and obtain the target network opening scheme meeting the application online service intention. Since the AI network large model has strong natural language processing capability, it is not necessary to pre-configure the network opening scheme and intention, that is, even if new intentions and changing intentions appear, or the scene and networking requirements change, the AI network large model can analyze and obtain the required target network opening scheme, improve the expansion capability of intention collection, and improve the adaptation capability to the scene and networking.

[0401] In some embodiments, the collection module 231 is specifically configured to:

[0402] display a man-machine interactive dialogue box;

[0403] receive the first natural language indicating the application online service intention input by the user on the man-machine interactive dialogue box.

[0404] In some embodiments, the collection module 231 is further configured to:

[0405] display first prompt information on the man-machine interactive dialogue box, the first prompt information being used to guide the input of the application online service intention.

[0406] In some embodiments, the determination module 232 is specifically configured to:

[0407] match the first natural language with the pre-stored knowledge of multiple network opening schemes to obtain multiple candidate knowledge;

[0408] input the first natural language and the multiple candidate knowledge into the AI network large model to obtain the target network opening scheme.

[0409] In some embodiments, the apparatus further includes a display module configured to:

[0410] collect a second natural language indicating a display mode of the network opening scheme based on a man-machine interactive mode;

[0411] input the second natural language into the AI network large model to obtain a target display mode;

[0412] call an interface corresponding to the target display mode by using an intelligent agent to display the target network opening scheme.

[0413] In some embodiments, the display module is specifically configured to:

[0414] display the human-computer interaction dialogue box;

[0415] receive a second natural language input by a user on the human-computer interaction dialogue box, the second natural language indicating a display manner of the network opening solution.

[0416] In some embodiments, the display module is further configured to:

[0417] display second prompt information on the human-computer interaction dialogue box, the second prompt information guiding input of the display manner of the network opening solution.

[0418] In some embodiments, the simulation module 233 is specifically configured to:

[0419] adjust the displayed target network opening solution based on a user operation;

[0420] simulate and evaluate the adjusted target network opening solution.

[0421] In some embodiments, the AI network large model outputs a plurality of target network opening solutions; and the simulation module 233 is specifically configured to:

[0422] determine a target network opening solution from the plurality of target network opening solutions;

[0423] simulate and evaluate the determined target network opening solution.

[0424] In some embodiments, the simulation module 233 is specifically configured to:

[0425] collect a third natural language indicating the determined target network opening solution based on a human-computer interaction manner;

[0426] input the third natural language into the AI network large model to obtain the determined target network opening solution.

[0427] In some embodiments, the simulation module 233 is specifically configured to:

[0428] display the human-computer interaction dialogue box;

[0429] receive a third natural language input by a user on the human-computer interaction dialogue box, the third natural language indicating the determined target network opening solution.

[0430] In some embodiments, the simulation module 233 is further configured to:

[0431] display third prompt information on the human-computer interaction dialogue box, the third prompt information guiding input of the determined target network opening solution.

[0432] In some embodiments, the simulation module 233 is specifically configured to:

[0433] collect a fourth natural language indicating simulation evaluation based on the human-computer interaction mode;

[0434] input the fourth natural language into an AI network large model to obtain a simulation evaluation instruction;

[0435] call an interface corresponding to the simulation evaluation instruction by using an agent to perform simulation evaluation on the target network opening scheme.

[0436] In some embodiments, the simulation module 233 is specifically configured to:

[0437] display a human-computer interaction dialogue box;

[0438] receive a fourth natural language indicating simulation evaluation input by a user on the human-computer interaction dialogue box.

[0439] In some embodiments, the simulation module 233 is further configured to:

[0440] display fourth prompt information on the human-computer interaction dialogue box, the fourth prompt information being used to guide input of execution information of simulation evaluation.

[0441] In some embodiments, the opening module 234 is specifically configured to:

[0442] collect a fifth natural language indicating network opening based on the human-computer interaction mode;

[0443] input the fifth natural language into an AI network large model to obtain a network opening instruction;

[0444] call an interface corresponding to the network opening instruction by using an agent to open the network by using the target network opening scheme.

[0445] In some embodiments, the opening module 234 is specifically configured to:

[0446] display a human-computer interaction dialogue box;

[0447] receive a fifth natural language indicating network opening input by a user on the human-computer interaction dialogue box.

[0448] In some embodiments, the opening module 234 is further configured to:

[0449] display fifth prompt information on the human-computer interaction dialogue box, the fifth prompt information being used to guide input of execution information of network opening.

[0450] In some embodiments, the apparatus further includes a display module configured to:

[0451] determine an operation to be performed;

[0452] Display prompt information associated with the to-be-executed operation on the man-machine interaction dialog box.

[0453] In some embodiments, the apparatus further includes a fine-tuning module configured to:

[0454] Obtain a sample natural language and a sample network opening solution corresponding to the sample natural language;

[0455] Fine-tune the AI network large model using the sample natural language and the sample network opening solution.

[0456] Embodiments of the present application also provide an electronic device, as shown in FIG. 24, which includes a processor 241, a communication interface 242, a memory 243 and a communication bus 244, wherein the processor 241, the communication interface 242 and the memory 243 complete communication with each other through the communication bus 244;

[0457] The memory 243 is configured to store a computer program.

[0458] The processor 241 is configured to execute the program stored in the memory 243 to implement the network opening method for any of the applications.

[0459] The communication bus mentioned in the electronic device can be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The communication bus can be divided into an address bus, a data bus, a control bus, etc. For the convenience of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or only one type of bus.

[0460] The communication interface 242 is configured to complete communication between the electronic device and other devices.

[0461] The memory 243 can include a random access memory (RAM) and can also include a non-volatile memory, for example, at least one disk memory. Optionally, the memory 243 can also be at least one storage device located away from the aforementioned processor 241.

[0462] The processor 241 described above can be a general processor, including a central processing unit (CPU), a network processor (NP), etc.; can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component.

[0463] In a further embodiment provided by the present application, a computer readable storage medium is provided, and the computer readable storage medium stores a computer program. The computer program is executed by a processor to implement the network opening method for application online of any of the above embodiments.

[0464] In a further embodiment provided by the present application, a computer program product containing instructions is provided, which, when executed on a computer, causes the computer to perform the network opening method for application online of any of the above embodiments.

[0465] In the above embodiments, all or part of the embodiments can be implemented by software, hardware, firmware or any combination thereof. When implemented by software, all or part of the embodiments can be implemented in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of the present application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another computer readable storage medium, for example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center through wired (such as coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (such as infrared, wireless, microwave, etc.) mode. The computer readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. integrated with one or more available media. The available media can be a magnetic medium (for example, floppy disk, hard disk, magnetic tape), an optical medium (for example, DVD), or a semiconductor medium (for example, solid state disk (SSD)) and the like.

[0466] It is to be noted that, in the present text, the terms such as first and second, and the like, are used merely to differentiate one entity or operation from another entity or operation, and do not necessarily require or imply any actual such relationship or order between such entities or operations. Also, the terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can also include other elements not expressly listed or inherent to such process, method, article, or apparatus. An element proceeded by "comprises... a" does not, without more constraints, exclude the existence of additional identical elements in the process, method, article, or apparatus that comprises the element.

[0467] Each of the embodiments in the present specification is described in a related manner, and the same or similar parts between the embodiments can be referred to each other. Each of the embodiments focuses on the difference from other embodiments. In particular, for the device, electronic device, storage medium, and program product embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the part of the description of the method embodiments.

[0468] The above only describes the preferred embodiments of the present application, and is not intended to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for network activation of an application, characterized in that, The method includes: Based on human-computer interaction, collect the first natural language of the business intent to indicate the application's online launch; Input the first natural language into the AI ​​network big model to obtain the target network activation scheme; The target network activation scheme was simulated and evaluated. After the simulation evaluation results meet the preset conditions, the network is activated according to the target network activation plan.

2. The method according to claim 1, characterized in that, The step of collecting the first natural language indicating the intent to launch the application based on human-computer interaction includes: Display the human-computer interaction dialog box; Receive the user's first natural language input on the human-computer interaction dialog box indicating the intent to launch the application's online service.

3. The method according to claim 2, characterized in that, The method further includes: The first prompt message is displayed on the human-computer interaction dialog box. The first prompt message is used to guide the input of the application's intention to launch the business.

4. The method according to claim 1, characterized in that, The step of inputting the first natural language into the AI ​​network big model to obtain the target network activation scheme includes: The first natural language is matched with pre-stored knowledge of various network access schemes to obtain multiple candidate knowledge; The first natural language and the various candidate knowledge are input into the AI ​​network big model to obtain the target network opening scheme.

5. The method according to claim 1, characterized in that, The method further includes: Based on human-computer interaction, the second natural language of the display method of the network activation plan is collected; The second natural language is input into the AI ​​network model to obtain the target display method; An intelligent agent is used to call the interface corresponding to the target display method to display the target network activation scheme.

6. The method according to claim 5, characterized in that, The step of collecting the second natural language of the display method of the network activation scheme based on human-computer interaction includes: Display the human-computer interaction dialog box; The system receives a second natural language input from the user on the human-computer interaction dialog box, indicating how the network activation plan will be displayed.

7. The method according to claim 6, characterized in that, The method further includes: A second prompt message is displayed on the human-computer interaction dialog box. The second prompt message is used to guide the input of the network activation scheme display method.

8. The method according to any one of claims 5-7, characterized in that, The steps for simulating and evaluating the target network activation scheme include: Adjust the displayed target network activation plan based on user actions; The adjusted target network activation plan was simulated and evaluated.

9. The method according to claim 1, characterized in that, The AI ​​network model outputs multiple target network activation schemes; The steps for simulating and evaluating the target network activation scheme include: From the multiple target network activation schemes, one target network activation scheme is determined; The determined target network activation plan is simulated and evaluated.

10. The method according to claim 9, characterized in that, The step of determining a target network activation scheme from the plurality of target network activation schemes includes: Based on human-computer interaction, the third natural language of the target network opening scheme determined by the instruction is collected; The third natural language is input into the AI ​​network model to obtain the determined target network opening scheme.

11. The method according to claim 10, characterized in that, The step of collecting the third natural language of the target network access scheme determined by the instruction based on human-computer interaction includes: Display the human-computer interaction dialog box; The system receives the target network activation scheme determined by the user's input on the human-computer interaction dialog box in third natural language.

12. The method according to claim 11, characterized in that, The method further includes: A third prompt message is displayed on the human-computer interaction dialog box. The third prompt message is used to guide the input of the determined target network activation scheme.

13. The method according to claim 1, characterized in that, The steps for simulating and evaluating the target network activation scheme include: Based on human-computer interaction, the fourth natural language is collected to guide the simulation evaluation; The fourth natural language is input into the AI ​​network model to obtain simulation evaluation instructions; An intelligent agent invokes the interface corresponding to the simulation evaluation command to perform a simulation evaluation of the target network activation scheme.

14. The method according to claim 13, characterized in that, The step of collecting the fourth natural language of the simulation evaluation based on human-computer interaction includes: Display the human-computer interaction dialog box; The system receives instructions for simulation evaluation input by the user on the human-computer interaction dialog box in a fourth natural language.

15. The method according to claim 14, characterized in that, The method further includes: A fourth prompt message is displayed on the human-computer interaction dialog box. This fourth prompt message is used to guide the input of execution information for simulation evaluation.

16. The method according to claim 1, characterized in that, The steps for activating the network according to the target network activation scheme include: Based on human-computer interaction, the fifth natural language of the instruction network is collected; The fifth natural language is input into the AI ​​network model to obtain the network activation command; An intelligent agent invokes the interface corresponding to the network activation command to activate the network using the target network activation scheme.

17. The method according to claim 16, characterized in that, The step of collecting the fifth natural language signal from the network based on human-computer interaction includes: Display the human-computer interaction dialog box; Receive the fifth natural language input by the user on the human-computer interaction dialog box, indicating that the network should be activated.

18. The method according to claim 17, characterized in that, The method further includes: A fifth prompt message is displayed on the human-computer interaction dialog box. The fifth prompt message is used to guide the input of execution information for network activation.

19. The method according to any one of claims 3, 7, 12, 15 and 18, characterized in that, The method further includes: Determine the operation to be performed; The prompt information associated with the operation to be performed is displayed on the human-computer interaction dialog box.

20. The method according to claim 1, characterized in that, The method further includes: Obtain the sample natural language and the corresponding sample network access scheme for the sample natural language; The AI ​​network model is fine-tuned using the sample natural language and the sample network access scheme.

21. A network activation device for application deployment, characterized in that, The device includes: The data collection module is used to collect the first natural language of the business intent to indicate the application's launch, based on human-computer interaction. The determination module is used to input the first natural language into the AI ​​network big model to obtain the target network opening scheme; The simulation module is used to simulate and evaluate the target network activation scheme. The activation module is used to activate the network according to the target network activation scheme after the simulation evaluation results meet the preset conditions.

22. An electronic device, characterized in that, It includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; The memory is used to store computer programs; When the processor executes the program stored in the memory, it implements the method described in any one of claims 1-20.

23. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method described in any one of claims 1-20.

Images