Data analysis model management method, electronic device and storage medium
The implementation of a model management service module with interfaces for ML models in 5G networks addresses the lack of network-level management, improving efficiency and stability in data analysis model operations.
Patent Information
- Application Number
- JP2024522709
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2021-12-30
- Filing Date
- 2022-12-19
- Publication Date
- 2024-11-13
- Estimated Expiration
- 2042-12-19
AI Technical Summary
Current industry lacks a model management function for ML models at the network management level in 5G networks, which are essential for intelligent network operations and management.
A model management service module with interfaces for model registration, training, validation, deployment, and update is implemented at the network management side, enabling unified management of data analysis models.
This solution simplifies the management of ML models, enhances efficiency, and ensures stable data transmission by providing a unified framework for model operations such as registration, training, validation, and deployment.
Smart Images

Figure 2024541858000001_ABST
Abstract
Description
[Technical field]
[0001] The present application relates to the technical field of wireless communication, for example, to a data analysis model management method, an electronic device, and a storage medium.
[0002] This application claims priority to a Chinese patent application bearing application number 202111659735.1, filed with the China Patent Office on December 30, 2021, the entire contents of which are incorporated herein by reference. [Background technology]
[0003] At present, mobile communication technology has entered the 5th Generation Mobile Communication Technology (5G) era, and network slices are an important part of 5G. A network slice is an instantiated logical network of characteristic network characteristics mainly composed of network functions, where the network functions include network resources supporting the network functions, and the logical network is used to meet the needs of characteristic network traffic, such as ultra-low latency and ultra-high reliability. With the development of communication technology, 5G networks have higher requirements for intelligence in addition to characteristics such as high bandwidth, ultra-low latency and ultra-high reliability. At present, related technologies for automation in mobile communication can use Self-Organized Networks (SON) and Intent-driven networks, etc. Artificial Intelligence (AI) technology, big data technology, etc. will be further introduced into the 5G network system. Based on these technologies, the 5G network system will gradually have the ability of intelligent analysis and formulation, and then realize the automated operation and management of the 5G network through the automatic execution of the intelligent analysis and formulation results. Meanwhile, the Third Generation Partnership Project (3GPP) research also sees the importance of data analysis for the automated management and operation of the system, so the concepts of the Network Data Analytics Function (NWDAF) and Management Data Analytics Function (MDAF) are also put forward, and research is also being conducted on using these functions to optimize system performance, troubleshoot, predict system resources, etc. On the Radio Access Network (RAN) side, research is also being started on the RAN intelligent framework.The intelligent functions in all these RAN intelligent frameworks use AI and Machine Learning (ML) technology, and AI / ML refers to ML models. Meanwhile, in the research process of RAN intelligent frameworks, one aspect put forward is to train the ML models required for the intelligent functions on the RAN side at the network management level, and perform corresponding updates and configurations after training. To meet this need, it is necessary to provide a corresponding ML model management function at the network management level, but the current industry does not yet have such an aspect of model management function. Summary of the Invention [Means for solving the problem]
[0004] The embodiment of the present application is obtaining, at a network management side of the communication network, a model management service module including at least one model management service interface; and performing a model management operation on the data analysis model based on the model management service interface. A method for managing a data analysis model is provided.
[0005] The embodiment of the present application is one or more processors; and a memory configured to store one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement a method according to any one of the embodiments of the present application. An electronic device is also provided.
[0006] The embodiment of the present application is A computer-readable storage medium having a computer program stored thereon, The computer program, when executed by a processor, performs the method according to any one of the embodiments of the present application. A computer readable storage medium is also provided. [Brief description of the drawings]
[0007] [Figure 1] 1 is a flowchart of a data analysis model management method according to an embodiment of the present application; [Diagram 2] 1 is a flowchart of another data analysis model management method according to an embodiment of the present application; [Diagram 3] 1 is a flowchart of another data analysis model management method according to an embodiment of the present application; [Figure 4] 1 is a flowchart of another data analysis model management method according to an embodiment of the present application; [Diagram 5] 1 is a flowchart of another data analysis model management method according to an embodiment of the present application; [Figure 6] 1 is a flowchart of another data analysis model management method according to an embodiment of the present application; [Figure 7] 1 is a flowchart of another data analysis model management method according to an embodiment of the present application; [Figure 8] FIG. 2 is an exemplary diagram of a data analysis model management method according to an embodiment of the present application. [Figure 9] FIG. 1 is a structural schematic diagram of a data analysis model management device according to an embodiment of the present application; [Figure 10] 1 is a structural schematic diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0008] It should be understood that the specific examples described herein are merely for the purpose of illustrating the present application and are not intended to limit the present application.
[0009] In the following description, the suffixes for expressing elements such as "module", "member" or "unit" are merely used to contribute to the description of this application and have no specific meaning in themselves. Therefore, "module", "member" or "unit" can be used interchangeably.
[0010] FIG. 1 is a flowchart of a data analysis model management method according to an embodiment of the present application, which can be applied to the case of data analysis model management in a wireless communication scene, and the method can be performed by a data analysis module management device according to an embodiment of the present application, which can be realized in the form of software and / or hardware. Referring to FIG. 1, the method according to the embodiment of the present application specifically includes the following steps:
[0011] At step 110, at the network management side of the communications network, a model management service module is obtained that includes at least one model management service interface.
[0012] Here, the network management side of the communication network may be an equipment side that manages network communications in a wireless communication network, or may be an operation support system layer, a network management layer, a network element management layer, a slice management layer, a sub-slice management layer, etc., and the model management service module may be a software / hardware module that manages data analysis models on the network management layer side of the communication network, and the model management service module may include one or more model management service interfaces, and each model management service interface can realize different management of the data analysis models.
[0013] In an embodiment of the present application, a model management service module may be provided on the network management side of the communication network, and the model management service model may include one or more model management service interfaces, and each model management service interface may be packaged with operation rules for different management operations between data analysis models.
[0014] In step 120, a model management operation is performed on the data analysis model based on the model management service interface.
[0015] Specifically, by invoking the model management service interface, the model management operation of the data analysis model can be realized, and the difficulty of using the data analysis model in the communication network can be reduced.
[0016] In an embodiment of the present application, a model management service model including at least one model management service interface is obtained on the network management side of a communication network, and model management operations for data analysis models are realized by using the model management service interface. In the embodiment of the present application, unified management of data analysis models in a communication network is realized through the model management model, so as to reduce the difficulty of using data analysis models, improve the efficiency of model management, and ensure the stability of data transmission in the communication network.
[0017] In one embodiment, based on the embodiment of the above application, the model management service interface includes at least a model registration interface, a model training interface, a model validation interface, a model update interface, a model deployment interface, and a model registration information update interface.
[0018] Specifically, the model management service module can be composed of interfaces such as a model registration interface, a model training interface, a model verification interface, a model update interface, a model deployment interface, and a model registration information update interface, and by invoking the interfaces, operations such as model registration, model training, model verification, model update, model deployment, and registration information update can be realized, and processing rules for corresponding operations may be set in advance for each interface.
[0019] FIG. 2 is a flowchart of another data analysis model management method according to an embodiment of the present application. Referring to FIG. 2, taking model registration as an example, the method according to the embodiment of the present application specifically includes the following steps:
[0020] At step 210, at the network management side of the communications network, a model management service module is obtained that includes at least one model management service interface.
[0021] In step 220, a model registration interface is invoked to register the data analysis model and store attribute information of the data analysis model.
[0022] Here, the module registration interface may be a software program interface, and the model registration interface may be packaged with processing rules for registering a data analysis model, and when the model registration interface is invoked, it can register information of the data analysis model in a preset location, and the preset location may be a preset memory space on the network management side of the communication network.
[0023] An embodiment of the present application can realize the registration process of a data analysis model by calling a model registration interface in a model management service module and storing attribute information of the data analysis model in the model registration interface, and a user can process based on the data analysis model to be registered.
[0024] In one embodiment, based on the embodiment of the above application, the attribute information includes at least one of a model identifier, a model name, a model function description, a model algorithm type, a model parameter, a model initial hyperparameter, a training dataset, and a test dataset.
[0025] In an embodiment of the present application, registration of a data analysis model can be realized by storing one or more of the model identifier, model name, model function description, model algorithm type, model parameters, model initial hyperparameters, training data set, and test data set of the data analysis model, and an application in a communication network can realize the use of the data analysis model based on the above attribute information.
[0026] In the embodiment of the present application, a model registration interface can be called to register a data analysis model, and attribute information of the data analysis model can be stored in a preset storage space to realize the registration of the data analysis model.
[0027] FIG. 3 is a flowchart of another data analysis model management method according to an embodiment of the present application. Referring to FIG. 3, taking model training as an example, the method according to the embodiment of the present application specifically includes the following steps:
[0028] At step 310, at the network management side of the communications network, a model management service module is obtained that includes at least one model management service interface.
[0029] In step 320, a model training interface is invoked to train the data analysis model according to a training invocation command that includes at least a model identifier or a model name.
[0030] Here, the training initiation command may be instruction information for controlling a model to perform training, and the training initiation command may include a model identifier or a model name of the data analysis model waiting to be trained, and the model identifier or the model name may be composed of numbers, alphabets, or special characters.
[0031] Specifically, when a training start command is received, the model identifier or model name included in the training start command can be extracted, and the model training interface can be used to obtain a data analysis model corresponding to the model identifier or model name, and training can be performed.
[0032] FIG. 4 is a flowchart of another data analysis model management method according to an embodiment of the present application. Referring to FIG. 4, taking the verification of a data analysis model as an example, the method according to the embodiment of the present application may specifically include the following steps:
[0033] At step 410, at the network management side of the communications network, a model management service module is obtained that includes at least one model management service interface.
[0034] In step 420, a model validation interface is invoked to validate the data analysis model according to a validation invocation command that includes at least a model identifier or a model name.
[0035] Here, the validation start command may be instruction information for controlling a model to perform training, and the validation start command may include a model identifier or a model name of the data analysis model to be validated, and the model identifier or the model name may be composed of numbers, alphabets, or special characters.
[0036] In an embodiment of the present application, when a verification initiation command is received, a model verification interface is invoked according to a model identifier or a model name included in the verification initiation command, and a corresponding data analysis model is obtained to perform verification.
[0037] FIG. 5 is a flowchart of another data analysis model management method according to an embodiment of the present application. Referring to FIG. 5, taking the update of an application that uses a data analysis model as an example, the method according to the embodiment of the present application specifically includes the following steps:
[0038] At step 510, at a network management side of the communications network, a model management service module is obtained that includes at least one model management service interface.
[0039] In step 520, according to an update initiation instruction including at least an application to be updated and a model identifier, the model update interface is invoked to update a data analysis model corresponding to the application.
[0040] Here, the update start command may be instruction information that controls the update of the data analysis model, and the update start command may include a model identifier of the data analysis model waiting to be updated and an application waiting to be updated, the model identifier may be unique identification information of the data analysis model waiting to be updated, and the application waiting to be updated may be an application that uses the data analysis model.
[0041] In an embodiment of the present application, when an update start command is received, the application waiting to be updated and the model identifier in the update start command are extracted, and the application waiting to be updated and the model identifier are used to call a model update interface, so that the data analysis model used in the application waiting to be updated can be updated by the model update interface.
[0042] FIG. 6 is a flowchart of another data analysis model management method according to an embodiment of the present application. Referring to FIG. 6, taking the configuration of a data analysis model as an example, the method according to the embodiment of the present application specifically includes the following steps:
[0043] At step 610, at the network management side of the communications network, a model management service module is obtained that includes at least one model management service interface.
[0044] In step 620, a model deployment interface is invoked to deploy the data analysis model used in the application according to the deployment instructions, which include at least the application to be deployed and the model identifier.
[0045] Specifically, when a deployment command is received, a model deployment interface can be called to deploy a data analysis model corresponding to the model identifier in an application to be deployed, so that the application to be deployed can realize functions according to the data analysis model.
[0046] FIG. 7 is a flowchart of another data analysis model management method according to an embodiment of the present application. Referring to FIG. 7, taking updating model registration information as an example, the method according to the embodiment of the present application specifically includes the following steps:
[0047] At step 710, at the network management side of the communications network, a model management service module is obtained that includes at least one model management service interface.
[0048] In step 720, a model registration information update interface is invoked to update the registration information of the data analysis model according to a registration update instruction including at least a model identifier and registration information to be updated.
[0049] In an embodiment of the present application, when a registration update command is received, a model registration information update interface is called according to the model identifier and the registration information waiting to be updated in the registration update command, and the original registration information corresponding to the model identifier is replaced with the registration information waiting to be updated, thereby realizing the update of the registration information.
[0050] In one embodiment, based on the embodiment of the above application, the method further includes generating notification information according to an execution result of the model management operation, and transmitting the notification information to an application corresponding to the data analysis model.
[0051] In an embodiment of the present application, after the execution of the model management operation is completed, corresponding notification information, such as information on successful model registration or successful model update, can be generated according to the execution result, and the notification information can be sent to the application in which the data analysis model is deployed, to remind the application to use the data analysis model of the information.
[0052] In one embodiment, based on the embodiment of the above application, the model management service module is located in at least one of an Operation-Support System (OSS) layer, a network management layer, a network element management layer, a slice management layer, and a sub-slice management layer.
[0053] FIG. 8 is an exemplary diagram of a data analysis model management method according to an embodiment of the present application. Referring to FIG. 8, the data analysis model may specifically be an ML model, and the management method according to an embodiment of the present application may include the following steps:
[0054] 1) Register (or distribute) the model to the model database (or model market) using the ML model registration (or distribution) interface.
[0055] 2) Train the registered model through the ML model training interface.
[0056] 3) Validate the trained model through the ML model validation interface.
[0057] 4) The ML model deployment interface deploys the validated model to a specific application.
[0058] 5) Periodically training, verifying, and updating the ML model, including periodically launching an ML model training interface to train the model, invoking an ML model verification interface to verify the trained model, and updating the trained model to the characteristic application in which the ML model is deployed through an ML model update interface.
[0059] 6) Call the ML model registration information update model according to needs, and update the model registration information, which may include changes in the model algorithm, changes in the training dataset, and changes in the test dataset, etc.
[0060] The management of the above ML model can be realized by an ML management service module, which may include functional interfaces such as, but not limited to, an ML model registration interface, an ML model training interface, an ML model verification interface, an ML model update interface, an ML model configuration interface, and an ML model registration information update interface.
[0061] Here, the ML model registration interface provides an ML model registration function. Any authorized consumer of the RAN intelligent framework or its underlying RAN intelligent function or ML model can register an ML model by invoking the ML model registration interface. The registration information includes, but is not limited to, one or more of the following attributes: model identifier, model name, model function description, model algorithm type (or specific definition of model algorithm), model parameter, model initial (or default) hyperparameter, training data set acquisition path, training data set, model test data set acquisition path, and model test data set. After the model registration is completed, corresponding model information is generated, and the model information is stored in a model database (or model market), and the model information further includes an attribute for storing identification information of an application using the ML model, and when the model is registered, the value of the attribute is empty. The above training data set acquisition path and model test data set acquisition path can also be replaced with a specific model training data set and model test data set.
[0062] The ML model training interface provides an ML model training function. The RAN intelligent framework or the RAN intelligent function thereunder, the upper network management system, the local network management function, or any other authorized consumer can all invoke the ML model training interface to initiate training of an ML model. When initiating training of a model, the initiation command includes at least one of a model identifier and a model name. First, based on the model identifier or the model name, the corresponding model and parameters such as model initial (or default) hyperparameters, training data set acquisition path or training data set itself are obtained from the model database, and then model training is initiated based on the obtained parameters. After the training is completed, the corresponding information in the model database is updated, and a model training completion notification can also be sent to the consumer who initiated the training, and the same notification can also be sent to other applications that use the model at the same time.
[0063] The ML model validation interface provides an ML model validation function. The RAN intelligence framework or the RAN intelligence function thereunder, the upper network management system, the lower network management function, or any other authorized consumer can invoke the ML model validation interface to initiate validation of the ML model. When initiating model validation, the initiation command includes at least one of a model identifier and a model name. First, according to the model identifier or the model name, parameters such as the corresponding model test data set acquisition path or the model test data set itself are obtained from the model database, and then model validation is initiated according to the acquired parameters or data. After the validation is completed, the corresponding information in the model database is updated, and a model validation result notification can also be sent to the consumer who initiated the model validation, and the same notification can also be sent to other applications that use the model.
[0064] The ML model update interface provides an ML model update function. The RAN intelligent framework or the RAN intelligent function thereunder, the upper network management system, the lower network management function, or any other authorized consumer can initiate an ML model update by calling the ML model update interface. When initiating a model update, the initiation command includes at least one of a model identifier and a model name, and an application identifier or application list that needs to update the model. According to the model identifier or the model name, the corresponding model information can be obtained from the model database and the model can be updated. After the update is completed, a model update completion notification can be sent to the consumer that initiated the model verification, and the same notification can also be sent to the application that updated the model at the same time.
[0065] The ML model management server module in an embodiment of the present application may be located in an operation support system layer, a network management layer, a network element management layer, a slice management layer, and a sub-slice management layer.
[0066] In another exemplary embodiment, an ML model information model, or an ML model information object class (IOC), or an ML model managed object class (MOC) can be defined, and the model can be named MLModel. The ML model information model includes: 1) an ML model identifier (MLModelID) for uniquely identifying a model; 2) an ML model algorithm definition for defining a model algorithm; 3) an ML model parameters for recording model parameters (immediately after a model is registered, initial or default model parameters are recorded, and after a model is trained, post-training model parameters are recorded); 4) model initial (or default) hyperparameters for recording model hyperparameters; 5) a training data set acquisition path or the training data set itself; 6) a model testing data set; The ML model information model may include, but is not limited to, one of the following attributes: 1) acquisition path or model test dataset itself; 2) model state for recording the state of the model (available states include, but are not limited to, untrained, training, trained, unverified, verified, undeployed, and deployed); and 3) model deployment application list for recording the application to which the model belongs. The above ML model information model may further include attributes of model name and model function description. A model can be registered by creating an instance of MLModel by a createMOI operation, and a model can be deleted by deleting the instance of MLModel by a deleteMOI operation. Model training is performed by first training the model by a model training operation, and then writing the trained model parameters to the instance of MLModel by a modifyMOIAttributes operation to modify the model state.Model deployment is performed by first deploying the model through a model deployment operation, then modifying the model state in the MLModel instance through the modifyMOIAttributes operation, and recording the application information to which the model belongs. ML model registration information is updated by modifying the registration information in the MLModel instance through the modifyMOIAttributes operation, including the algorithm definition, model initial (or default) hyperparameters, training data set acquisition path, model testing data set acquisition path, etc.
[0067] FIG. 9 is a structural schematic diagram of a data analysis model management device according to an embodiment of the present application, which can execute the data analysis model management method described in any of the embodiments of the present application, and specifically includes functional modules corresponding to the execution of the method and beneficial effects, and the device can be realized in software and / or hardware, and specifically includes a management service module 801 and an operation execution module 802.
[0068] The management service module 801 is configured to obtain, on a network management side of a communication network, a model management service module including at least one model management service interface.
[0069] The operation execution module 802 is configured to execute a model management operation on a data analysis model based on the model management service interface.
[0070] In an embodiment of the present application, a management service module obtains a model management service model including at least one model management service interface on the network management side of a communication network, and an operation execution module realizes model management operations for data analysis models by using the model management service interface. In an embodiment of the present application, the model management model realizes unified management of data analysis models in a communication network, reduces the difficulty of using data analysis models, improves the efficiency of model management, and ensures the stability of data transmission in the communication network.
[0071] In one embodiment, based on the embodiment of the above application, the model management service interface in the apparatus includes at least a model registration interface, a model training interface, a model validation interface, a model update interface, a model configuration interface, and a model registration information update interface.
[0072] In one embodiment, based on the embodiment of the above application, the operation execution module 802 in the device includes a registration unit configured to call the model registration interface to register the data analysis model, and store attribute information of the data analysis model.
[0073] In one embodiment, based on the embodiment of the above application, the attribute information includes at least one of a model identifier, a model name, a model function description, a model algorithm type, a model parameter, a model initial hyperparameter, a training dataset, and a test dataset.
[0074] In one embodiment, based on the embodiment of the above application, the operation execution module 802 in the apparatus includes a training unit configured to call the model training interface to train the data analysis model according to a training initiation instruction including at least a model identifier or a model name.
[0075] In one embodiment, based on the embodiment of the above application, the operation execution module 802 in the apparatus includes a validation unit configured to call the model validation interface to validate the data analysis model according to a validation invocation instruction including at least a model identifier or a model name.
[0076] In one embodiment, based on the embodiment of the above-mentioned application, the operation execution module 802 in the device includes an update unit configured to call the model update interface to update the data analysis model corresponding to the application according to an update activation instruction including at least an application to be updated and a model identifier.
[0077] In one embodiment, based on the embodiment of the above-mentioned application, the operation execution module 802 in the apparatus includes a deployment unit configured to call the model deployment interface to deploy the data analysis model used in the application according to a deployment instruction including at least an application to be deployed and a model identifier.
[0078] In one embodiment, based on the embodiment of the above-mentioned application, the operation execution module 802 in the apparatus includes a registration information unit configured to call the model registration information update interface to update the registration information of the data analysis model according to a registration update instruction including at least a model identifier and registration information to be updated.
[0079] In one embodiment, based on the embodiment of the above application, the apparatus further includes an information notification module configured to generate notification information according to the execution result of the model management operation, and transmit the notification information to an application corresponding to the data analysis model.
[0080] In one embodiment, based on the embodiment of the above application, the model management service module in the device is located in at least one of an operation support system layer, a network management layer, a network element management layer, a slice management layer, and a sub-slice management layer.
[0081] FIG. 10 is a structural schematic diagram of an electronic device according to an embodiment of the present application, which includes a processor 90, a memory 91, an input device 92 and an output device 93. The number of processors 90 in the electronic device may be one or more. In FIG. 10, one processor 90 is taken as an example. The processor 90, memory 91, input device 92 and output device 93 in the electronic device may be connected via a bus or in other manners. In FIG. 10, connection via a bus is taken as an example.
[0082] The memory 91 can be used as a computer-readable storage medium to store software programs, computer-executable programs and modules, such as modules corresponding to the data transmission device in the embodiment of the present application (e.g., the management service module 801 and the operation execution module 802). The processor 90 executes the software programs, instructions and modules stored in the memory 91 to execute various functional applications and data processing of the electronic device, that is, to realize the above-mentioned data analysis model management method.
[0083] The memory 91 may include a program storage area and a data storage area, where the program storage area may store an operating system, an application program required for at least one function, and the data storage area may store data generated based on the use of the electronic device, etc. The memory 91 may also include a high-speed random access memory, and may further include a non-volatile memory, such as at least one magnetic disk storage device, flash memory, or other non-volatile solid-state storage device. In some embodiments, the memory 91 may include memories located remotely to the processor 90, and these remote memories may be connected to the device via a network. Instances of the network may include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.
[0084] The input device 92 can be used to receive input numeric or textual information and generate key signal inputs associated with user settings and function control of the electronic device. The output device 93 can include a display device such as a display.
[0085] An embodiment of the present application further provides a storage medium including computer-executable instructions, which, when executed by a computer processor, perform a data analysis model management method, the method including: obtaining, at a network management side of the communication network, a model management service module including at least one model management service interface; and performing a model management operation on the data analysis model based on the model management service interface.
[0086] From the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be realized by adding a general-purpose hardware platform required for software, and of course, can also be realized by hardware, and in many cases, the former is a more preferred embodiment. Based on this understanding, the technical solution of the present application can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium such as a computer floppy disk, a read-only memory (ROM), a random access memory (RAM), a flash memory (FLASH), a hard disk, or an optical disk, and includes a plurality of instructions for making a computer device (which may be a personal computer, a server, or a network device, etc.) execute the methods of the embodiments of the present application.
[0087] In the above embodiment of the device, each unit and module is merely divided according to a functional logic, and is not limited to the above division, as long as it can realize the corresponding function. Furthermore, the specific names of each functional unit are merely for easy distinction from each other, and are not intended to limit the scope of protection of the present application.
[0088] Those skilled in the art will understand that all or part of the steps in the above disclosed methods, systems, and functional modules / units in the devices can be implemented as software, firmware, hardware, or a suitable combination thereof.
[0089] In hardware embodiments, the division among the functional modules / units mentioned in the above description does not necessarily correspond to the division of physical components. For example, one physical component may have multiple functions, or one function or step may be performed cooperatively by multiple physical components. Some or all of the physical components may be implemented as software executed by a processor, such as a central processor, digital signal processor, or microprocessor, or may be implemented as hardware, or may be implemented as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on computer readable media, which may include computer storage media (or non-transitory media) and communication media (or transitory media). As known to those skilled in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (e.g., computer readable instructions, data structures, program modules, or other data). Computer storage media may include, but is not limited to, RAM, ROM, Electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, Compact Disc Read-Only Memory (CD-ROM), Digital Versatile Disc (DVD) or other optical disk memory, magnetic cartridges, magnetic tapes, magnetic disk memory or other magnetic storage devices, or any other medium used to store the desired information and accessible by a computer. Additionally, as known to those skilled in the art, communication media typically include computer readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier or other transport mechanism, and may include any information delivery media.
[0090] Although some embodiments of the present application have been described above with reference to the drawings, they are not intended to limit the scope of the present application. Any modifications, equivalent replacements, improvements, etc. made by those skilled in the art without departing from the scope and substance of the present application should be included in the scope of protection of the present application.
Claims
1. obtaining, at a network management side of the communication network, a model management service module including at least one model management service interface; performing a model management operation on the data analysis model based on the at least one model management service interface. Data analysis model management methods.
2. The at least one model management service interface includes at least: Including a model registration interface, a model training interface, a model validation interface, a model update interface, a model configuration interface, and a model registration information update interface; The method of claim 1.
3. performing a model management operation on a data analysis model based on the at least one model management service interface, calling the model registration interface to register the data analysis model and storing attribute information of the data analysis model; The method according to claim 1 or 2.
4. The attribute information is At least one of a model identifier, a model name, a model function description, a model algorithm type, model parameters, model initial hyperparameters, a training dataset, and a test dataset; The method according to claim 3.
5. performing a model management operation on a data analysis model based on the at least one model management service interface, invoking the model training interface to train the data analysis model according to a training invocation command including at least a model identifier or a model name; The method according to claim 1 or 2.
6. performing a model management operation on a data analysis model based on the at least one model management service interface, invoking the model validation interface to validate the data analysis model according to a validation invocation command including at least a model identifier or a model name; The method according to claim 1 or 2.
7. performing a model management operation on a data analysis model based on the at least one model management service interface, invoking the model update interface to update the data analysis model corresponding to the application according to an update initiation command including at least an application to be updated and a model identifier; The method according to claim 1 or 2.
8. performing a model management operation on a data analysis model based on the at least one model management service interface, deploying the data analysis model used in the application by invoking the model deployment interface in accordance with a deployment instruction including at least a to-be-deployed application and a model identifier. The method according to claim 1 or 2.
9. performing a model management operation on a data analysis model based on the at least one model management service interface, In accordance with a registration update command including at least a model identifier and registration information to be updated, calling the model registration information update interface to update the registration information of the data analysis model. The method according to claim 1 or 2.
10. generating notification information according to a result of the execution of the model management operation; and transmitting the notification information to an application corresponding to the data analysis model. The method according to claim 1 or 2.
11. The model management service module includes: Located in at least one of an operation support system layer, a network management layer, a network element management layer, a slice management layer, and a sub-slice management layer; The method of claim 1.
12. one or more processors; and a memory configured to store one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 11. electronic equipment.
13. A computer-readable storage medium having a computer program stored thereon, The computer program, when executed by a processor, implements the method according to any one of claims 1 to 11. A computer-readable storage medium.
Citation Information
Patent Citations
Control device, control method, and program
JP2021083058A
Configuring network analytics
WO2021023388A1
Data analysis method and device, apparatus, and storage medium
WO2021047332A1
Network data analysis method, functional entity and electronic device
WO2021088520A1
