WO2021063254A1 - 一种信息处理方法、相关设备及计算机存储介质 - Google Patents

一种信息处理方法、相关设备及计算机存储介质 Download PDF

Info

Publication number
WO2021063254A1
WO2021063254A1 PCT/CN2020/117681 CN2020117681W WO2021063254A1 WO 2021063254 A1 WO2021063254 A1 WO 2021063254A1 CN 2020117681 W CN2020117681 W CN 2020117681W WO 2021063254 A1 WO2021063254 A1 WO 2021063254A1
Authority
WO
WIPO (PCT)
Prior art keywords
information
action
network element
target
decision
Prior art date
Application number
PCT/CN2020/117681
Other languages
English (en)
French (fr)
Inventor
王园园
王岩
苏琪
Original Assignee
华为技术有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 华为技术有限公司 filed Critical 华为技术有限公司
Publication of WO2021063254A1 publication Critical patent/WO2021063254A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/50—Network service management, e.g. ensuring proper service fulfilment according to agreements
    • H04L41/5003—Managing SLA; Interaction between SLA and QoS
    • H04L41/5009—Determining service level performance parameters or violations of service level contracts, e.g. violations of agreed response time or mean time between failures [MTBF]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
    • G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis

Definitions

  • This application relates to the field of information processing technology, and in particular to an information processing method, related equipment and computer storage media.
  • intent-based network management has also been applied in wireless networks to reduce human-computer interaction and improve network performance.
  • intent-based network management uses simple instructions to tell the network "what to do” instead of "how to do” to reduce the inaccuracy of large-scale manual decision-making.
  • users in the vertical industry have insufficient operation and maintenance knowledge, and intent-based network management can reduce users' demand for operation and maintenance knowledge.
  • the above-mentioned telling the network "what to do” is the intention to be realized, such as “improving the video service experience of users in zone x"; the process from “what to do” to "how to do” and to execute “how to do” is the translation of intent , The process of transforming the intent into a reasonable network deployment and executing the actions of the deployment.
  • intent translation mainly uses intent-driven management providers to receive intents issued by consumers of intent-driven management services, and query the knowledge base for the actions required to meet the intent and the action execution conditions. After the knowledge base matches the goals and intentions, the results are Feedback to consumers of intention-driven management services.
  • the intention-driven management service provider queries the network element for the current network status, and issues the actions that need to be performed in this status.
  • the intent of the wireless network is mainly oriented to operation and maintenance management, and most intents can be satisfied through characteristic operation and maintenance.
  • the characteristics of the wireless network are not completely orthogonal. There are scenarios where multiple solutions can satisfy the same intention at the same time.
  • the intention-driven management service provider selects the solution randomly or tries one by one, which leads to the current intention. Translation efficiency is low.
  • the embodiments of the present application disclose an information processing method, related equipment, and computer storage medium, which can quickly determine target action information to achieve an intention by adding auxiliary selection functions and processes, thereby improving the efficiency of intention translation.
  • the embodiments of the present application disclose data transmission methods, including:
  • the first network element obtains multiple action information of the intention
  • the first network element determines target action information according to the decision information of the multiple action information, and the target action information is used to realize the intention.
  • the first network element determining target action information according to the decision information of the multiple action information includes:
  • the first network element obtains the intended business request or historical experience information achieved by the intended intent, where the business request refers to a business requirement corresponding to the intent, and the historical experience information is the same as the current scenario
  • the intention to translate event information, the same scenario means that the intention to be achieved and the network status are the same;
  • the first network element determines the target action information according to the intended service request or historical experience information achieved by the intent and the decision information.
  • the first network element determining target action information according to the decision information of the multiple action information includes:
  • the first network element inputs the decision information of the multiple actions into a machine learning model to obtain the target action information.
  • the determining target action information according to the decision information of the multiple action information includes:
  • the first network element sends an action selection request to a second network element or a third network element, the action selection request carries decision information of the multiple action information, and the action selection request is for the second network element or the third network element.
  • the three network elements determine target decision information according to the decision information of the multiple action information;
  • the first network element determines the target action information according to the target decision information or the identification information.
  • the first network element determining the decision information of the multiple action information includes:
  • the optimization gains include actions and action objects
  • the negative effects include actions and action objects
  • the first network element determines the decision information of the multiple actions according to the operations of the multiple action information, and the operation information includes the action and the action object.
  • the first network element determining the decision information of the multiple actions according to the optimization gains and negative effects of the multiple action information includes:
  • the first network element determines the decision information of the multiple actions according to the optimized gains and negative effects of the multiple action information and operations.
  • an information processing method including:
  • the second network element receives an action selection request from the first network element, the action selection request includes decision information of a plurality of action information, and the plurality of action information is obtained by the first network element according to an intention, the decision The information is the information contained in the multiple action information;
  • the second network element determines target decision information according to the decision information of the multiple action information, and the target decision information is used to determine the target action;
  • the second network element sends an action selection response to the first network element, and the action selection response carries the target decision information or carries identification information of the target decision information.
  • the second network element determines the target decision information according to the decision information of the multiple action information, including:
  • the second network element In the case that the second network element cannot process the action selection request, the second network element sends the action selection request to a third network element;
  • the second network element receives the action selection response from the third network element.
  • the second network element determines the target decision information according to the decision information of the multiple action information, including:
  • the second network element obtains the intended business request or historical experience information achieved by the intended intent, where the business request refers to a business requirement corresponding to the intent, and the historical experience information is the same as the current scenario
  • the intention to translate event information, the same scenario means that the intention to be achieved and the network status are the same;
  • the second network element determines the target decision information according to the decision information of the multiple actions and the request information or the historical experience information.
  • the second network element determines the target decision information according to the decision information of the multiple action information, including:
  • the second network element inputs the decision information of the multiple actions into a machine learning model to obtain the target decision information.
  • the decision information of the multiple action information includes the operation of the multiple action information, or the decision information of the multiple action information includes the optimization gains and negatives of the multiple action information.
  • Influence, or the decision information of the multiple action information includes the operation of the multiple action information, the optimization gain and the negative impact of the multiple action information, the optimization gain includes the action and the action object, and the negative impact includes Actions and action objects, and the operation information includes actions and action objects.
  • an embodiment of the present application provides an information processing method, including:
  • the third network element receives an action selection request from the first network element or the second network element, the action selection request, the action selection request including decision information of multiple action information, and the multiple action information is the first Obtained by a network element according to an intention, the decision information is information included in the multiple action information;
  • the third network element determines target decision information according to the decision information of the multiple action information, where the target decision information is used to determine the target action;
  • the third network element sends an action selection response to the first network element or the second network element, where the action selection response carries the target decision information or carries identification information of the target decision information.
  • the third network element determines the target decision information according to the decision information of the multiple action information, including:
  • the third network element obtains the intended service request or historical experience information achieved by the intended intent, the service request refers to the service requirement corresponding to the intent, and the historical experience information is the same as the current scenario
  • the intention to translate event information, the same scenario means that the intention to be achieved and the network status are the same;
  • the third network element determines the target decision information according to the decision information of the multiple actions and the request information or the historical experience information.
  • the third network element determines the target decision information according to the decision information of the multiple action information, including:
  • the second network element inputs the decision information of the multiple actions into a machine learning model to obtain the target decision information.
  • the decision information of the multiple action information includes the operation of the multiple action information, or the decision information of the multiple action information includes the optimization gains and negatives of the multiple action information.
  • Influence, or the decision information of the multiple action information includes the operation of the multiple action information, the optimization gain and the negative impact of the multiple action information, the optimization gain includes the action and the action object, and the negative impact includes Actions and action objects, and the operation information includes actions and action objects.
  • the first network element uses the optimization gains and negative effects of multiple actions as decision information, and then Through the business demands of the intention or the historical experience information achieved by the intention of the intention, the target decision information in the decision information (that is, the acceptable optimization gain and negative impact and the unacceptable negative impact), or determined by a machine learning model
  • the target decision information in the decision information further determines the target action information in the multiple action information according to the target decision information.
  • the embodiment of the present application adds a selection process of action information to determine target action information, which can quickly determine target action information, and The accuracy of the action information selection, thereby improving the efficiency of intent translation.
  • an information processing device including:
  • the acquisition module is used to acquire multiple action information of the intent
  • the first determining module is configured to determine decision information of the multiple action information, where the decision information is the information contained in the multiple action information;
  • the second determining module is configured to determine target action information according to the decision information of the multiple action information, and the target action information is used to realize the intention.
  • the information processing device in the embodiment of the present application obtains multiple intended action information, then extracts decision information from the multiple pieces of information, and determines target action information in the multiple pieces of action information according to the obtained decision information. It can be seen that the information processing device quickly determines the target action information for realizing the intention by adding auxiliary selection functions and processes, thereby improving the efficiency of intention translation.
  • the second determination module is specifically configured to obtain the business appeal of the intention or historical experience information achieved by the intent of the intention, and the business appeal refers to the information corresponding to the intention Business requirements, the historical experience information is the same intent translation event information as the current scenario, the same scenario means that the intent to be realized and the network state are the same; the business appeal according to the intent or the history of the intent to achieve The empirical information and the decision information determine the target action information.
  • the second determining module is specifically configured to input the decision information of the multiple actions into a machine learning model to obtain the target action information.
  • the second determining module is specifically configured to send an action selection request to a second network element or a third network element, and the action selection request carries decision information of the multiple action information .
  • the action selection request is used by the second network element or the third network element to determine target decision information according to the decision information of the multiple action information; receiving the action selection from the second network element or the third network element
  • the action selection response carries the target decision information or identification information of the target decision information; the target action information is determined according to the target decision information or the identification information.
  • the first determining module is specifically configured to determine the decision information of the multiple actions according to the optimization gains and negative effects of the multiple action information, and the optimization gains include actions and actions Object, the negative influence includes an action and an action object; or the decision information of the multiple actions is determined according to the operation of the multiple action information, and the operation information includes the action and the action object.
  • the first determining module is specifically configured to determine the decision information of the multiple actions according to the optimization gains, negative effects and operations of the multiple action information.
  • an embodiment of the present application also provides an information processing device, including:
  • the receiving module is configured to receive an action selection request from a first network element, the action selection request includes decision information of multiple action information, and the multiple action information is obtained by the first network element according to the intention,
  • the decision information is the information contained in the multiple action information
  • the determining module is configured to determine target decision information according to the decision information of multiple action information, and the target decision information is used to determine the target action;
  • the sending module is configured to send an action selection response to the first network element, where the action selection response carries the target decision information or carries identification information of the target decision information.
  • the information processing device in the embodiment of the present application generally receives an action selection request from a first network element, where the action selection request carries decision information; then, the target decision information in the decision information is determined, and finally the first network element is sent to the first network element.
  • the network element sends the target decision information to generate an action selection response, so that the first network element determines target action information according to the target decision information. It can be seen that the information processing device helps the first network element to determine the target action information for achieving the intention by adding auxiliary selection functions and processes, thereby improving the efficiency of intention translation.
  • the determining module is specifically configured to send the second network element to the third network element when the second network element cannot process the action selection request Action selection request; receiving the action selection response from the third network element.
  • the determining module is specifically configured to obtain the business appeal of the intention or historical experience information achieved by the intention of the intention, and the business appeal refers to the business corresponding to the intention. It is required that the historical experience information is the same intent-to-translate event information as the current scenario, the same scenario means that the intent to be realized and the network state are the same; according to the decision information of the multiple actions and the request information or The historical experience information determines the target decision information.
  • the determining module is specifically configured to input the decision information of the multiple actions into a machine learning model by the second network element to obtain the target decision information.
  • the decision information of the multiple action information includes the operation of the multiple action information, or the decision information of the multiple action information includes the optimization gain and the optimization gain of the multiple action information.
  • Negative influence, or the decision information of the multiple action information includes the operation of the multiple action information, the optimization gain and the negative impact of the multiple action information, the optimization gain includes the action and the action object, the negative impact Contains actions and action objects, and the operation information contains actions and action objects.
  • an embodiment of the present application also provides an information processing device, including:
  • the receiving module is configured to receive an action selection request from a first network element or a second network element, the action selection request, the action selection request including decision information of multiple action information, and the multiple action information is the Obtained by the first network element according to an intention, the decision information is information included in the plurality of action information;
  • the determining module is configured to determine target decision information according to the decision information of multiple action information, and the target decision information is used to determine the target action;
  • the sending module is configured to send an action selection response to the first network element or the second network element, where the action selection response carries the target decision information or carries identification information of the target decision information.
  • the information processing device in the embodiment of the application generally receives an action selection request from the first network element or the second network element, wherein the action selection request carries decision information; then the target decision information in the decision information is determined, and finally Sending an action selection response carrying target decision information to the first network element or the second network element, so that the first network element determines target action information according to the target decision information. It can be seen that the information processing device helps the first network element to determine the target action information for achieving the intention by adding auxiliary selection functions and processes, thereby improving the efficiency of intention translation.
  • the determining module is specifically configured to obtain the business appeal of the intention or historical experience information achieved by the intention of the intention, and the business appeal refers to the business corresponding to the intention. It is required that the historical experience information is the same intent-to-translate event information as the current scenario, the same scenario means that the intent to be realized and the network state are the same; according to the decision information of the multiple actions and the request information or The historical experience information determines the target decision information.
  • the determining module is specifically configured to input the decision information of the multiple actions into a machine learning model by the second network element to obtain the target decision information.
  • the decision information of the multiple action information includes the operation of the multiple action information, or the decision information of the multiple action information includes the optimization gain and the optimization gain of the multiple action information.
  • Negative influence, or the decision information of the multiple action information includes the operation of the multiple action information, the optimization gain and the negative impact of the multiple action information, the optimization gain includes the action and the action object, the negative impact Contains actions and action objects, and the operation information contains actions and action objects.
  • an embodiment of the present application provides an information processing device, including a processor and a transceiver, wherein: the processor and the transceiver are connected to each other, the transceiver is used to communicate with a data analysis device, and the processing The device is configured to perform the method as described in the first aspect.
  • an embodiment of the present application further provides an information processing device, including a processor and a transceiver, wherein: the processor and the transceiver are connected to each other, the transceiver is used to communicate with a data analysis device, and the The processor is configured to perform the method as described in the second aspect.
  • an embodiment of the present application provides an information processing device, including a processor and a transceiver, wherein: the processor and the transceiver are connected to each other, the transceiver is used to communicate with a data analysis device, and the processing The device is configured to perform the method as described in the third aspect.
  • an embodiment of the present application provides a computer storage medium, where the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method as described in the first aspect is implemented.
  • an embodiment of the present application further provides a computer storage medium, the computer-readable storage medium stores a computer program, and the computer program implements the method described in the second aspect when the computer program is executed by a processor.
  • an embodiment of the present application further provides a computer storage medium, the computer-readable storage medium stores a computer program, and the computer program implements the method described in the third aspect when the computer program is executed by a processor.
  • Figure 1 is a schematic diagram of the interaction between an IDMS consumer and an IDMS provider provided by an embodiment of the present application
  • FIG. 2 is a schematic flowchart of an information processing method provided by an embodiment of the present application.
  • FIG. 3 is a schematic flowchart of another information processing method provided by an embodiment of the present application.
  • FIG. 4 is a schematic flowchart of another information processing method provided by an embodiment of the present application.
  • FIG. 5 is a schematic flowchart of yet another information processing method provided by an embodiment of the present application.
  • FIG. 6 is a schematic flowchart of yet another information processing method provided by an embodiment of the present application.
  • FIG. 7 is a schematic flowchart of yet another information processing method provided by an embodiment of the present application.
  • FIG. 8 is a schematic diagram of a layered architecture of an IDM system provided by an embodiment of the present application.
  • FIG. 9 is a schematic structural diagram of an information processing device provided by an embodiment of the present application.
  • FIG. 10 is a schematic structural diagram of another information processing device provided by an embodiment of the present application.
  • FIG. 11 is a schematic structural diagram of another information processing device provided by an embodiment of the present application.
  • FIG. 12 is a schematic structural diagram of another information processing device provided by an embodiment of the present application.
  • FIG. 13 is a schematic structural diagram of another information processing device provided by an embodiment of the present application.
  • FIG. 14 is a schematic structural diagram of another information processing device provided by an embodiment of the present application.
  • the embodiments of this application provide an information processing method and related products, which are applied to intent-driven management services (IDMS), and an auxiliary translation system or function is added in the intent-driven management service to assist in the provision of intent-driven management services.
  • IDMS intent-driven management services
  • the IDMS Provider completes the target action selection in the translation process.
  • Intent-driven management includes intent expression, intent interface, and intent lifecycle management.
  • Intent-driven management service allows IDMS consumers to express their intent to manage networks and services, and IDMS providers perform intent translation, convert their intents into reasonable network deployment information, and perform deployment operations.
  • the interaction between IDMS consumers and IDMS providers is shown in Figure 1. As shown in Figure 1, the IDMS consumer issues intent information to the IDMS provider. After receiving the intent information, the IDMS provider feeds back an intent to receive notification to the IDMS consumer, and feeds back the intent achievement status.
  • Intent expression is composed of action Action and object Object, in the form of ⁇ Intent Driven Action> ⁇ Intent Driven Object> two-tuple, in which Action is used to abstract and simplify the operation of the network, and formally includes an action name and a series of related attributes ⁇ intent driven action name>, ⁇ intent driven action properties>; Object is used to provide management object information, formally contains an object name and a series of properties used to identify the object ⁇ intent driven object name>, ⁇ intent driven object properties>.
  • condition information is introduced into the intention expression, and the intention expression carries the condition information, which is mainly divided into two categories: post-condition and pre-condition.
  • Constraints are the parameters or performance constraints that need to be met after an Action is executed on the specified Object.
  • the constraint has nothing to do with the Action; for example, under the premise of not affecting the utilization of physical resource blocks (PRB), the downlink users in area X
  • the rate is increased to 1Gbps; the effective conditions are the prerequisites that need to be met before the Action is executed on the Object, which can be time limit, network status, etc.; for example, when the edge downlink reference signal received power (RSRP) of site X is less than- At 110dbm, the coverage enhancement feature is enabled.
  • PRB physical resource blocks
  • the IDMS provider After the IDMS consumer delivers the intent to the IDMS provider, the IDMS provider translates the intent based on the intent knowledge base and the current network status. Among them, the design of the intention knowledge base is shown in Table 1.
  • Goal It mainly describes the achievable effect of the operation.
  • the goal of the intention knowledge base corresponds to the action in the intention expression.
  • the table item corresponding to the goal is in line with the intention.
  • the goal may be "to improve the quality of long-term evolution network language services (Voice over Long Term Evolution, VoLTE)".
  • Intention satisfies conditions It mainly describes the conditions that need to be met for the intention to be fulfilled.
  • the intention satisfaction conditions of the intention knowledge base are a Boolean expression.
  • the intent to meet the conditions may be: the video first broadcast delay ⁇ 10ms, the average number of video freezes per play ⁇ 2, and the video freeze duration ⁇ 10%.
  • Operation executable conditions It mainly describes the conditions under which the operation corresponding to the target can be executed.
  • the operation executable condition of the intention knowledge base is a Boolean expression in which the data can be collected from the object indicated by the intention Yes, the collected data is substituted into the Boolean expression to determine whether the condition is true.
  • the operation type of the intention knowledge base is an enumerated type with two values: if it is "script”, the corresponding operation describes a script; if it is "optimization”, the corresponding What is described in the operation is an optimization problem.
  • Operation It mainly describes the operation corresponding to the specific target and the executable condition of the operation.
  • the operation of the intention knowledge base is a string that describes the specific operation. According to the different operation types, different operations are described.
  • the operation type is "script”
  • the operation describes a series of operation instructions for the operation object indicated by the intent, such as: MML commands
  • the operation type is "optimization”
  • the operation describes the operation for the intent.
  • Remarks are a character string, which can be empty, and are the descriptive text of the entry.
  • the IDMS provider receives the intent information issued by the IDMS consumer, and then queries the knowledge base for the action information that needs to be executed to satisfy the intent.
  • the action information includes the operation, the operation execution condition, the intent achievement condition, and optimization.
  • One or more of the information such as gain and negative impact; after the knowledge base receives the query request from the IDMS provider, the request includes the target, and the knowledge base matches the target with the intent, and then feeds back the result (that is, the action information) To IDMS providers. Then, the IDMS provider queries the network element for the current network status, and issues the operations that need to be performed in this status in order to achieve the intention.
  • Table 2 An example of a knowledge base where multiple action information can achieve the intention at the same time
  • the detection index refers to the key performance indicator (KPI) involved in the intent to meet the condition and the operation executable condition, the KPI for which the intent to meet the condition is an intent to monitor KPI, and the operation The KPI that can execute the condition is the condition KPI.
  • KPI key performance indicator
  • the CCE-AvgUtilizationRate in the operation executable conditions represents the average utilization of control channel elements
  • DL-Packet-Delay represents the downlink packet delay
  • DL-User-Thp( ⁇ 5M)-SampIndex represents the downlink user throughput rate The ratio is lower than 5M.
  • the meanings of the operation of the three action messages are as follows: 1 Turn on the upstream VOIP scheduling optimization switch to ensure that voice users can be scheduled in time when SR missed detection occurs, to avoid PDCP packet loss timer timeout and packet loss while reducing
  • condition information (which may include the constraints and effective conditions) carried in the intent information is optional, and the condition information may not necessarily be included in the optimization gains and negatives.
  • Information that affects information matching therefore, when the IDMS provider searches through the knowledge base and finds that the target has multiple action information to achieve the intention, the IDMS provider has no information for reference when selecting the action information; in this case Generally, random selection or one-by-one method is used to select the action information, which will reduce the efficiency of intention translation.
  • the embodiment of the present application proposes an information processing method to assist the IDMS provider in selecting action information.
  • the following describes the embodiment of the present application in detail.
  • the first network element mentioned below refers to the party that converts the intent into a network management task or strategy during the intent translation process, that is, the IDMS provider in the intent-driven management service IDMS;
  • the second network The element refers to the party that issues or proposes the intention in the intention translation process, that is, the IDMS consumer in the intention-driven management service IDMS;
  • the third network element refers to the network element used to assist the first network element to select action information.
  • FIG. 2 is a schematic flowchart of an information processing method provided by an embodiment of the present application. As shown in the figure, the information processing method includes:
  • the first network element obtains multiple action information corresponding to the intent information.
  • the intention information is information related to the intention, and the intention information may carry information such as an intention identifier, an intention action, and an intention object.
  • Each action information includes information such as the operation, the execution conditions of the operation, and the intention to meet the conditions, and does not include the optimization gain and negative effects of the operation.
  • each action information corresponds to an operation
  • an operation can include at least one operation instruction
  • each operation instruction includes an action and an action object.
  • the process for the first network element to obtain multiple action information corresponding to the intent information may specifically include 1011-1019 as shown in FIG. 2.
  • the first network element receives the intention information from the second network element.
  • the second network element delivers the intention information that carries information such as the intention identifier, the intention action, and the intention object to the first network element, so that the first network element completes the intention translation.
  • the first network element sends an intent to receive notification to the second network element.
  • the notification of intent to receive is used to notify the second network element of the receiving status of the intent information, and the receiving status may include successful reception and/or reception failure.
  • the first network element extracts keywords of the intention target in the intention information.
  • the first network element After receiving the intention information, the first network element performs lexical, grammatical, and semantic analysis on the intention information to extract keywords of the intention target.
  • the intent information is: intent1, which improves the video service experience of users in zone x; the keywords of the intent target can be extracted as: promotion, video, and service experience.
  • the lexical, grammatical, and semantic analysis can be implemented using a lexical analyzer, a syntax analyzer, and a semantic analyzer in the computer domain.
  • the first network element sends a query request to the knowledge base.
  • the query request carries the keyword of the intended target, that is, the keyword extracted in step 1013.
  • the query request is used to obtain the action information corresponding to the target matched with the keyword.
  • the knowledge base performs target matching according to the query request.
  • the knowledge base After receiving the query request, the knowledge base performs target matching according to the keywords in the query request to obtain optional action information corresponding to the target, wherein the optional action information corresponds to the optional action information At least one action information can be included.
  • each action information includes operation, operation executable conditions, and intention to satisfy conditions.
  • the operation executable condition includes a condition that a key performance indicator (KPI) needs to be met, where the condition KPI refers to the corresponding KPI in the precondition for executing the operation.
  • KPI key performance indicator
  • the operation executable condition may be: CCE-AvgUtilizationRate ⁇ 0.7&&DL-Packet-Delay>20ms, where CCE-AvgUtilizationRate represents the average utilization of control unit channel elements, DL-Packet-Delay represents the downlink packet delay, and CCE- Both AvgUtilizationRate and DL-Packet-Delay are conditional KPIs.
  • the conditions for satisfying the intention include conditions to be met by the KPI for intention monitoring, and the KPI for intention monitoring refers to the KPI that needs to be monitored for the intention to be achieved.
  • the intent to meet the conditions may be: the first video delay ⁇ 10ms, and the average number of video jams per play ⁇ 2, and the proportion of the video jam duration ⁇ 10%; the video first play delay, the average video per play Both the number of freezes and the proportion of video freeze duration are monitoring KPIs.
  • the first network element receives query result feedback from the knowledge base.
  • the query result feedback includes the action information corresponding to the target obtained in step 1015.
  • the first network element sends an intention monitoring KPI and a conditional KPI subscription request to the network element.
  • the intention monitoring KPI and the conditional KPI subscription request are used to obtain the intention monitoring KPI and the condition KPI of the network element in the current network state.
  • the first network element receives intent monitoring KPI and conditional KPI subscription feedback from the network element.
  • the intention monitoring KPI and the condition KPI subscription feedback carry the intention monitoring KPI and the condition KPI of the network element in the current network state.
  • the first network element determines the action information that matches the current network state, and saves the intention monitoring KPI.
  • the first network element after the first network element receives the intention monitoring KPI and the conditional KPI subscription feedback, it determines from the action information carried in the query result feedback the conditional KPI (That is, the activity information matched by the intention monitoring KPI and the condition KPI in the condition KPI subscription feedback, so as to obtain the multiple action information.
  • the first network element determines the decision information of the multiple action information.
  • the decision information includes operations of the multiple actions.
  • the first network element determines target action information according to the decision information of the multiple action information.
  • determining the target action information by the first network element according to the decision information of the multiple action information may include: obtaining, by the first network element, a service request corresponding to the intention information; Then, the target action is determined according to the decision information and the appeal information.
  • the first network element determines the operation to be performed in the decision information according to the service request information; then determines the target action information according to the operation to be performed .
  • determining the target action information by the first network element according to the decision information of the multiple action information may include: the first network element acquiring the intention achievement corresponding to the intention information The historical experience information; and then the target action is determined according to the decision information and the historical experience information of the intention to achieve.
  • the historical experience information mentioned in the embodiments of the present application and the following refers to the same intent translation event information as the current scene, and the same scene means that the intention to be achieved and the network state are the same; the business appeal Refers to the business requirements corresponding to the intention information.
  • the first network element determines the operation to be performed in the decision information according to the historical experience information of the intention achievement; then the operation to be performed according to the need Determine the target action information.
  • the first network element determining target action information according to the decision information of the multiple action information may include: the intention-driven management service consumer inputs the decision information into the machine learning Model, and determine the target action information according to the output of the machine learning model.
  • the machine learning model may be an abstract modeling of historical experience, and a two-layer neural network may be selected.
  • the input is information such as operations in the decision information, and the output is an acceptable or required operation.
  • the first network element sends an operation execution request to the network element.
  • the operation execution request carries the operation of the target action information, that is, the operation that the network element needs to perform.
  • the first network element receives an operation execution response from the network element.
  • the operation execution response is used to notify the first network element whether the operation of the target action is successfully executed.
  • the first network element sends an intention monitoring KPI subscription request to the network element.
  • the intention monitoring KPI subscription request includes the intention monitoring KPI in the query result feedback described in step 1016.
  • the first network element receives an intention monitoring KPI subscription response from the network element.
  • the KPI intent to monitor subscription response carries the value or state corresponding to the KPI in intent to monitor after the network element performs the operation.
  • the first network element compares the intention monitoring KPI saved before the execution of the operation and the intention monitoring KPI after the execution of the operation.
  • the IDMS After receiving the intent monitoring KPI subscription response from the network element, the IDMS reports the value and status of the intent monitoring KPI before and after the operation is performed (that is, the IDMS saved by the IDMS before the operation is performed).
  • the value of the intention-monitoring KPI is compared with the value or status of the intention-monitoring KPI in the subscription response of the intention-monitoring KPI after the execution of the operation, and the comparison result of the intention-monitoring KPI is obtained. Whether the stated intention is achieved.
  • the first network element sends the intention achievement status to the second network element.
  • the intention fulfillment situation may include the intention fulfilled and the unfulfilled intention.
  • the first network element determines that the intention has been achieved according to the comparison result
  • the first network element sends a message indicating that the intention has been achieved to the second network element
  • the comparison result determines that the intention is not fulfilled
  • the first network element sends a message that the intention is not fulfilled to the second network element; optionally, the unfulfilled intention message may include a comparison of the intention monitoring KPIs result.
  • the first network element uses the operation of multiple actions as decision information, and then passes the service of the intention
  • the historical experience information of the request or the intention of the intention is the target decision information in the decision information (that is, the target operation), or the target decision information in the decision information is determined through a machine learning model, and the target decision information is further determined according to the target
  • the information determines target action information in the plurality of action information.
  • the embodiment of the present application adds a selection process of action information to determine the solution of target action information, so that the first network element can quickly
  • the determination of the target action information and the accuracy of the action information selection further improve the efficiency of the first network element’s intention translation.
  • FIG. 3 is a schematic flowchart of another information processing method provided by an embodiment of the present application. As shown in the figure, this information processing method includes:
  • the first network element obtains multiple action information corresponding to the intent information.
  • the intention information is information related to the intention, and the intention information may carry information such as an intention identifier, an intention action, and an intention object.
  • the action information includes information such as operation, operation executable conditions, intent to satisfy conditions, optimization gains, and negative effects.
  • each action information corresponds to an operation
  • an operation can include at least one operation instruction
  • each operation instruction includes an action and an action object.
  • Each operation can correspond to at least one optimized gain and/or at least one negative influence.
  • Each optimization gain contains actions and action objects, and each negative influence contains actions and action objects.
  • the process for the first network element to obtain multiple action information corresponding to the intent information may specifically include 2011-2019 as shown in FIG. 3.
  • the first network element receives the intent information from the second network element.
  • the second network element delivers the intention information that carries information such as the intention identifier, the intention action, and the intention object to the first network element, so that the first network element completes the intention translation.
  • the first network element sends a notification of intent to receive to the second network element.
  • the notification of intent to receive is used to notify the second network element of the receiving status of the intent information, and the receiving status may include successful reception and/or reception failure.
  • the first network element extracts the keywords of the intention target in the intention information.
  • the first network element After receiving the intention information, the first network element performs lexical, grammatical, and semantic analysis on the intention information to extract keywords of the intention target.
  • the intent information is: intent1, which improves the video service experience of users in zone x; the keywords of the intent target can be extracted as: promotion, video, and service experience.
  • the lexical, grammatical, and semantic analysis can be implemented using a lexical analyzer, a syntax analyzer, and a semantic analyzer in the computer domain.
  • the first network element sends a query request to the knowledge base.
  • the query request carries the keywords of the intended target, that is, the keywords extracted in the 2013 step.
  • the query request is used to obtain the action information corresponding to the target matched with the keyword.
  • the knowledge base performs target matching according to the query request.
  • each action information includes information such as operation, operation executable conditions, intent to meet conditions, and optimization gains and negative effects.
  • the operating conditions include conditions that the conditional KPI needs to meet, where the conditional KPI refers to the corresponding KPI in the preconditions for performing the operation.
  • the operation executable condition may be: CCE-AvgUtilizationRate ⁇ 0.7&&DL-Packet-Delay>20ms, where CCE-AvgUtilizationRate represents the average utilization of control unit channel elements, DL-Packet-Delay represents the downlink packet delay, and CCE- Both AvgUtilizationRate and DL-Packet-Delay are conditional KPIs.
  • the conditions for satisfying the intention include conditions to be met by the KPI for intention monitoring, and the KPI for intention monitoring refers to the KPI that needs to be monitored for the intention to be achieved.
  • the intent to meet the conditions may be: the first video delay ⁇ 10ms, and the average number of video jams per play ⁇ 2, and the proportion of the video jam duration ⁇ 10%; the video first play delay, the average video per play Both the number of freezes and the proportion of video freeze duration are monitoring KPIs.
  • the first network element receives the query result feedback from the knowledge base.
  • the query result feedback includes the action information corresponding to the target obtained in step 2015.
  • the first network element sends an intention monitoring KPI and conditional KPI subscription request to the network element.
  • the intention monitoring KPI and the conditional KPI subscription request are used to obtain the intention monitoring KPI and the condition KPI of the network element in the current network state.
  • the first network element receives intent monitoring KPI and conditional KPI subscription feedback from the network element.
  • the intention monitoring KPI and the condition KPI subscription feedback carry the intention monitoring KPI and the condition KPI of the network element in the current network state.
  • the first network element determines the action information that matches the current network state, and saves the intention monitoring KPI.
  • the first network element after the first network element receives the intention monitoring KPI and the conditional KPI subscription feedback, it determines from the action information carried in the query result feedback the conditional KPI (That is, the activity information matched by the intention monitoring KPI and the condition KPI in the condition KPI subscription feedback, so as to obtain the multiple action information.
  • the first network element determines the decision information of the multiple action information.
  • the decision information includes optimization gains and negative effects of the multiple actions.
  • the decision information may also include operations of the multiple actions.
  • the first network element determines target action information according to the decision information of the multiple action information.
  • determining the target action information by the first network element according to the decision information of the multiple action information may include: obtaining, by the first network element, a service request corresponding to the intention information; Then, the target action is determined according to the decision information and the appeal information.
  • the first network element determines the decision according to the service request information Acceptable optimization gains and negative effects, and unacceptable negative effects in the information; and then determine the target action information based on the acceptable optimization gains, negative effects, and unacceptable negative effects.
  • the business demand is to protect a concert in a venue.
  • the unacceptable negative impact of the business demand is the reduction of cell capacity.
  • Other negative impacts and optimization gains are acceptable.
  • When determining the target action information only multiple action information is required.
  • the negative effects in the medium include the elimination of the action information of the cell capacity reduction, and the remaining action information is the target action information.
  • the first network element determines the acceptable optimization gain and negative impact in the decision information according to the service demand information, and Before the unacceptable negative impact, the first network element may determine an acceptable operation according to the service request, and determine the action information corresponding to the acceptable operation, if the acceptable operation corresponds to the action information If it is one, it is determined that the action information corresponding to the acceptable operation is the target action information; if there are multiple action information corresponding to the acceptable operation, the first network element continues to execute
  • the business demand information determines acceptable optimization gains and negative effects in the decision information, as well as operations for unacceptable negative effects.
  • determining the target action information by the first network element according to the decision information of the multiple action information may include: the first network element acquiring the intention achievement corresponding to the intention information The historical experience information; and then the target action is determined according to the decision information and the historical experience information of the intention to achieve.
  • the decision information includes the optimization gains and negative impacts of the multiple actions
  • the first network element is based on the history of the intention achievement
  • the empirical information determines acceptable optimization gains and negative effects, and unacceptable negative effects in the decision information; then, the target action information is determined according to the acceptable optimization gains, negative effects, and unacceptable negative effects.
  • the first network element determines the acceptable optimization gain and negative effects in the decision information according to the historical experience information achieved by the intention. Before the impact, and the unacceptable negative impact, the first network element may determine an acceptable operation according to the service requirement, and determine the action information corresponding to the acceptable operation, if the acceptable operation corresponds to When there is one action information, it is determined that the action information corresponding to the acceptable operation is the target action information; if there are multiple action information corresponding to the acceptable operation, the execution of the first network is continued.
  • the meta determines acceptable optimization gains and negative effects in the decision information, and operations for unacceptable negative effects in the decision information based on the historical empirical information achieved by the intention.
  • the first network element determining target action information according to the decision information of the multiple action information may include: the intention-driven management service consumer inputs the decision information into the machine learning Model, and determine the target action information according to the output of the machine learning model.
  • the machine learning model can be an abstract modeling of historical experience, and a two-layer neural network can be selected.
  • the input is the optimization gain, negative influence and operation information in the decision information, and the output is the acceptable optimization gain, acceptable The negative impact or action that needs to be performed.
  • the first network element sends an operation execution request to the network element.
  • steps 305-309 shown in FIG. 3 is similar to that of steps 105-109 shown in FIG. 2, and therefore will not be described again.
  • the first network element uses the optimization gains and negative effects of multiple actions as decision information, and then Through the business demands of the intention or the historical experience information achieved by the intention of the intention, the target decision information in the decision information (that is, the acceptable optimization gain and negative impact and the unacceptable negative impact), or determined by a machine learning model
  • the target decision information in the decision information further determines the target action information in the multiple action information according to the target decision information.
  • the embodiment of the present application adds a selection process of action information to determine the solution of target action information, so that the first network element can quickly
  • the determination of the target action information and the accuracy of the action information selection further improve the efficiency of the first network element’s intention translation.
  • FIG. 4 is a schematic flowchart of another information processing method provided by an embodiment of the present application. As shown in the figure, this information processing method includes:
  • the first network element acquires multiple action information corresponding to the intent information.
  • the process for the first network element to obtain multiple action information corresponding to the intent information is the same as step 101 shown in FIG. 2 and will not be repeated here.
  • the first network element determines the decision information of the multiple action information.
  • the decision information includes the operations of the multiple actions and the operation identifiers corresponding to the operations.
  • the operation identifier is the association information between the operation and the action information, that is, the operation information corresponding to the operation can be determined by the operation identifier of the operation.
  • the first network element determines target action information according to the decision information of the multiple action information.
  • the process of determining the target action information by the first network element according to the decision information of the multiple action information may specifically include 3031-3033 as shown in FIG. 4.
  • the first network element sends an action selection request to the second network element.
  • the action selection request carries decision information of the multiple actions.
  • the second network element determines target decision information according to the action selection request.
  • determining the target decision information by the second network element according to the action selection request may include: acquiring, by the second network element, a service request corresponding to the intention information; and then according to the The decision information and the appeal information determine the target decision information.
  • the second network element determines the operation to be performed in the decision information according to the service request information; then determines the target decision information according to the operation to be performed .
  • determining the target decision information by the second network element according to the action selection request may include: obtaining, by the second network element, historical empirical information of the intention achievement corresponding to the intention information ; Then determine the target decision information according to the decision information and the historical experience information of the intention to achieve.
  • the second network element determines the operation to be performed in the decision information according to the historical experience information of the intention achievement; then the operation to be performed according to the need Determine the target decision information.
  • the second network element determining target decision information according to the action selection request may include: the intention-driven management service consumer inputs the decision information into a machine learning model, and according to The output of the machine learning model determines the target decision information.
  • the machine learning model may be an abstract modeling of historical experience, and a two-layer neural network may be selected.
  • the input is information such as operations in the decision information, and the output is an acceptable or required operation, or The identifier of the operation that is acceptable or needs to be performed.
  • step 3032a the second network element can obtain the service request or the historical experience information of the intention fulfillment, or the second network element can call the machine learning model. If the prerequisites are not met, step 3032b is executed.
  • the second network element determines target decision information according to the action selection request.
  • the process for the second network element to determine the target decision information according to the action selection request may specifically include 3032b1-3032b3 as shown in FIG. 4.
  • the second network element sends the action selection request to the third network element.
  • the third network element determines target decision information according to the action selection request.
  • the determination of the target decision information by the third network element according to the action selection request may include: the third network element obtains the service request corresponding to the intent information; and then according to the The decision information and the appeal information determine the target decision information.
  • the third network element determines the operation to be performed in the decision information according to the service request information; then determines the target decision information according to the operation to be performed .
  • determining the target decision information by the third network element according to the action selection request may include: obtaining, by the third network element, historical experience information of intention achievement corresponding to the intention information ; Then determine the target decision information according to the decision information and the historical experience information of the intention to achieve.
  • the third network element determines the operation that needs to be performed in the decision information according to the historical experience information of the intention achievement; then the operation that needs to be performed Determine the target decision information.
  • the third network element determining target decision information according to the action selection request may include: the intention-driven management service consumer inputs the decision information into a machine learning model, and according to The output of the machine learning model determines the target decision information.
  • the third network element sends an action selection response to the second network element.
  • the third network element generates the action selection response according to the target decision information, and sends the action selection response to the second network element.
  • the action selection response is used by the first network element to determine a target action according to the information carried in the action selection response.
  • the action selection response carries the target decision information or carries identification information of the target decision information, and the identification information includes the operation identification in the target decision information.
  • the second network element sends an action selection response to the first network element.
  • the second network element receives the action selection response from the third network element, and then sends the action selection response to the first network element.
  • the second network element generates the action selection response according to the target decision information, and sends the action selection response to the first network element.
  • the action selection response carries the target decision information or carries identification information of the target decision information, and the identification information includes the operation identification in the target decision information.
  • the action selection response is used by the first network element to determine a target action according to the information carried in the action selection response.
  • the first network element sends an operation execution request to the network element.
  • the first network element After receiving the action selection response, the first network element determines a target action according to the information carried in the action selection response.
  • steps 305-309 shown in FIG. 4 is similar to that of steps 105-109 shown in FIG. 2, and therefore will not be described again.
  • the first network element when there are multiple solutions (that is, multiple action information) in the intention translation process that can satisfy the same intention at the same time, the first network element sends operations carrying multiple actions to the second network element (that is, decision-making). Information) action selection request.
  • the second network element determines the target decision information in the decision information (that is, the target operation) in the decision information through the intended business appeal or the historical experience information achieved by the intention of the intention, or determines the target decision in the decision information through a machine learning model information.
  • the action selection request is sent to the third network element so that the third network element can determine the target decision information (the process is the same as that of the second network element determining the target decision information ), and receive an action selection response carrying target decision information or identification information of the target decision information. Then the second network element sends an action selection response carrying target decision information or identification information of the target decision information to the first network element, so that the first network element determines the multiple actions according to the target decision information or the information identification Target action information in the message.
  • the embodiment of the present application adds the selection process of action information to determine the solution of the target action information, so that the first network element can quickly
  • the determination of the target action information and the accuracy of the action information selection further improve the efficiency of the first network element’s intention translation.
  • FIG. 5 is a schematic flowchart of an information processing method provided by an embodiment of the present application. As shown in the figure, the information processing method includes:
  • the first network element obtains multiple action information corresponding to the intention information.
  • the process for the first network element to obtain multiple action information corresponding to the intent information is the same as step 201 shown in FIG. 3, and will not be repeated here.
  • the first network element determines the decision information of the multiple action information.
  • the decision information includes the optimization gains of the multiple actions, the optimization gain identification corresponding to the optimization gain, the negative influence, and the negative influence identification corresponding to the negative influence.
  • the optimized gain identifier is the correlation information between the optimized gain and the action information, that is, the action information corresponding to the optimized gain can be determined through the optimized gain identifier of the optimized gain.
  • the negative impact identification is the associated information between the negative impact and the action information, that is, the action information corresponding to the negative impact can be determined through the negative impact identification of the negative impact.
  • the optional decision information includes the operations of the multiple actions and the operation identifiers corresponding to the operations.
  • the operation identifier is the association information between the operation and the action information, that is, the operation information corresponding to the operation can be determined by the operation identifier of the operation.
  • the first network element determines target action information according to the decision information of the multiple action information.
  • the process of determining the target action information by the first network element according to the decision information of the multiple action information may specifically include 4031-4033 as shown in FIG. 5.
  • the first network element sends an action selection request to the second network element.
  • the action selection request carries decision information of the multiple actions.
  • the second network element determines target decision information according to the action selection request.
  • determining the target decision information by the second network element according to the action selection request may include: acquiring, by the second network element, a service request corresponding to the intention information; and then according to the The decision information and the appeal information determine the target decision information.
  • the second network element determines the acceptable optimization gain and negative effects in the decision information according to the service request information Impact, and unacceptable negative impact; then the target decision information is determined according to the acceptable optimization gain and negative impact, and the unacceptable negative impact.
  • the business demand is to protect a concert in a venue.
  • the unacceptable negative impact of the business demand is the reduction of cell capacity.
  • Other negative impacts and optimization gains are acceptable.
  • When determining the target action information only multiple action information is required.
  • the negative effects in the medium include the elimination of the action information of the cell capacity reduction, and the remaining action information is the target action information.
  • the second network element determines the acceptable optimization gain and negative impact in the decision information according to the service demand information, and Before the unacceptable negative impact, the second network element may determine an acceptable operation according to the service requirement, and if the acceptable operation is one, then determine the acceptable operation as the target decision Information; if there are multiple acceptable operations, continue to perform operations in which the second network element determines acceptable optimization gains and negative effects in the decision information according to the service demand information.
  • determining the target decision information by the second network element according to the action selection request may include: obtaining, by the second network element, historical empirical information of the intention achievement corresponding to the intention information ; Then determine the target decision information according to the decision information and the historical experience information of the intention to achieve.
  • the second network element determines the acceptable optimization gain and negative influence in the decision information according to the historical experience information of the intention achievement, and the unacceptable Accepted negative impact; and then determine the target decision information based on the acceptable optimization gain and negative impact, as well as the unacceptable negative impact.
  • the second network element determines the acceptable optimization gain and negative effects in the decision information according to the historical experience information achieved by the intention. Before the impact and unacceptable negative impact, the second network element may determine an acceptable operation according to the service requirement, and if the acceptable operation is one, then determine that the acceptable operation is all The target decision information; if there are multiple acceptable operations, continue to execute the second network element to determine acceptable optimization gains and negative effects in the decision information according to the historical experience information achieved by the intention, And operations that have unacceptable negative effects.
  • the second network element determining target decision information according to the action selection request may include: the intention-driven management service consumer inputs the decision information into a machine learning model, and according to The output of the machine learning model determines the target decision information.
  • the machine learning model can be an abstract modeling of historical experience, and a two-layer neural network can be selected.
  • the input is the optimization gain in the decision information, the negative impact and operation information, and the output is the acceptable optimization gain. , Acceptable negative effects or actions that need to be performed.
  • the prerequisite for performing step 4032a is: the second network element can obtain the service request or the historical experience information of the intention fulfillment, or the second network element can call the machine learning model. If the prerequisites are not met, step 4032b is executed.
  • the second network element determines target decision information according to the action selection request.
  • the process for the second network element to determine the target decision information according to the action selection request may specifically include 3032b1-3032b3 as shown in FIG. 4.
  • the second network element sends the action selection request to the third network element.
  • the third network element determines target decision information according to the action selection request.
  • the determination of the target decision information by the third network element according to the action selection request may include: the third network element obtains the service request corresponding to the intent information; and then according to the The decision information and the appeal information determine the target decision information.
  • the third network element determines the acceptable optimization gain and negative impact, and unacceptable negative impact in the decision information according to the service demand information; then The target decision information is determined according to acceptable optimization gains and negative effects, as well as unacceptable negative effects.
  • the business demand is to protect a concert in a venue.
  • the unacceptable negative impact of the business demand is the reduction of cell capacity.
  • Other negative impacts and optimization gains are acceptable.
  • When determining the target action information only multiple action information is required.
  • the negative effects in the medium include the elimination of the action information of the cell capacity reduction, and the remaining action information is the target action information.
  • the third network element determines the acceptable optimization gain and negative impact in the decision information according to the service demand information, and Before the unacceptable negative impact, the third network element may determine an acceptable operation according to the service requirement, and if the acceptable operation is one, determine the acceptable operation as the target decision Information; if there are multiple acceptable operations, continue to perform operations in which the third network element determines acceptable optimization gains and negative effects in the decision information according to the service demand information.
  • determining the target decision information by the third network element according to the action selection request may include: obtaining, by the third network element, historical experience information of intention achievement corresponding to the intention information ; Then determine the target decision information according to the decision information and the historical experience information of the intention to achieve.
  • the third network element determines the acceptable optimization gain and negative influence in the decision information according to the historical experience information of the intention achievement, and the unacceptable Accepted negative impact; and then determine the target decision information based on the acceptable optimization gain and negative impact, as well as the unacceptable negative impact.
  • the third network element determines the acceptable optimization gain and negative effects in the decision information according to the historical experience information achieved by the intention. Before the impact and unacceptable negative impact, the third network element may determine an acceptable operation according to the service requirement, and if the acceptable operation is one, then determine that the acceptable operation is all The target decision information; if there are multiple acceptable operations, continue to execute the third network element to determine acceptable optimization gains and negative effects in the decision information according to the historical experience information achieved by the intention, And operations that have unacceptable negative effects.
  • the third network element determining target decision information according to the action selection request may include: the intention-driven management service consumer inputs the decision information into a machine learning model, and according to The output of the machine learning model determines the target decision information.
  • the machine learning model can be an abstract modeling of historical experience, and a two-layer neural network can be selected.
  • the input is the optimization gain in the decision information, the negative impact and operation information, and the output is the acceptable optimization gain. , Acceptable negative effects or actions that need to be performed.
  • the third network element sends an action selection response to the second network element.
  • the third network element generates the action selection response according to the decision information, and sends an action selection response to the second network element, and the action selection response carries the target decision information or carries the Identification information of decision information.
  • the identification information includes the operation identifier of the operation; if the decision information includes an acceptable optimization gain and/or negative impact, and includes an unacceptable optimization gain and/or negative impact, Then the indication information includes the acceptable optimization gain indicator and/or negative impact indicator, and unacceptable optimization gain indicator and/or negative impact indicator.
  • the second network element sends an action selection response to the first network element.
  • the second network element receives the action selection response from the third network element, and then sends the action selection response to the first network element.
  • the second network element generates the action selection response according to the target decision information, and sends the action selection response to the first network element.
  • the second network element generating the action selection response according to the target decision information may include: the second network element generating the action selection response according to the decision information, the action selection response Carry the target decision information, or carry the identification information of the decision information.
  • the identification information includes the operation identifier of the operation; if the decision information includes an acceptable optimization gain and/or negative impact, and includes an unacceptable optimization gain and/or negative impact, Then the indication information includes the acceptable optimization gain indicator and/or negative impact indicator, and unacceptable optimization gain indicator and/or negative impact indicator.
  • the first network element sends an operation execution request to the network element.
  • the first network element After receiving the action selection response, the first network element determines a target action according to the information carried in the action selection response.
  • steps 405-409 shown in FIG. 5 is similar to steps 105-109 shown in FIG. 2, and therefore will not be described again.
  • the first network element when there are multiple solutions (that is, multiple action information) in the intention translation process that can satisfy the same intention at the same time, the first network element sends to the second network element optimized gains and negatives that carry multiple actions.
  • Action selection request that affects (that is, decision information).
  • the second network element achieves the goal decision information (that is, acceptable optimization gain and negative impact, and unacceptable negative impact) in the decision information through the intended business appeal or the historical experience information of the intended intention, or through the machine
  • the learning model is used to determine the target decision information in the decision information.
  • the action selection request is sent to the third network element so that the third network element can determine the target decision information (the process is the same as that of the second network element determining the target decision information ), and receive an action selection response carrying target decision information or identification information of the target decision information. Then the second network element sends an action selection response carrying the target decision information or the identification information of the target decision information to the first network element, so that the first network element determines the multiple actions according to the target decision information or the information identification Target action information in the message.
  • the embodiment of the present application adds a selection process of action information to determine the solution of target action information, so that the first network element can quickly
  • the determination of the target action information and the accuracy of the action information selection further improve the efficiency of the first network element’s intention translation.
  • FIG. 6 is a schematic flowchart of an information processing method provided by an embodiment of the present application. As shown in the figure, the information processing method includes:
  • the first network element obtains multiple action information corresponding to the intent information.
  • the process for the first network element to obtain multiple action information corresponding to the intent information is the same as step 101 shown in FIG. 2 and will not be repeated here.
  • the first network element determines decision information of the multiple action information.
  • the decision information includes the operations of the multiple actions and the operation identifiers corresponding to the operations.
  • the operation identifier is the association information between the operation and the action information, that is, the operation information corresponding to the operation can be determined by the operation identifier of the operation.
  • the first network element determines target action information according to the decision information of the multiple action information.
  • the process of determining the target action information by the first network element according to the decision information of the multiple action information may specifically include 3031-3033 as shown in FIG. 6.
  • the first network element sends an action selection request to the third network element.
  • the action selection request carries decision information of the multiple actions.
  • the third network element determines target decision information according to the action selection request.
  • the determination of the target decision information by the third network element according to the action selection request may include: the third network element obtains the service request corresponding to the intent information; and then according to the The decision information and the appeal information determine the target decision information.
  • the third network element determines the operation to be performed in the decision information according to the service request information; then determines the target decision information according to the operation to be performed .
  • determining the target decision information by the third network element according to the action selection request may include: obtaining, by the third network element, historical experience information of intention achievement corresponding to the intention information ; Then determine the target decision information according to the decision information and the historical experience information of the intention to achieve.
  • the third network element determines the operation that needs to be performed in the decision information according to the historical experience information of the intention achievement; then the operation that needs to be performed Determine the target decision information.
  • the third network element determining target decision information according to the action selection request may include: the intention-driven management service consumer inputs the decision information into a machine learning model, and according to The output of the machine learning model determines the target decision information.
  • the third network element sends an action selection response to the first network element.
  • the third network element generates the action selection response according to the target decision information, and sends the action selection response to the first network element.
  • the action selection response is used by the first network element to determine a target action according to the information carried in the action selection response.
  • the action selection response carries the target decision information or carries identification information of the target decision information, and the identification information includes the operation identification in the target decision information.
  • the first network element sends an operation execution request to the network element.
  • the first network element After receiving the action selection response, the first network element determines the target action according to the information carried in the action selection response.
  • steps 605-609 shown in FIG. 6 is similar to steps 105-109 shown in FIG. 2, and therefore will not be described again.
  • the first network element when there are multiple solutions (that is, multiple action information) in the intention translation process that can satisfy the same intention at the same time, the first network element sends operations carrying multiple actions to the third network element (that is, decision-making). Information) action selection request.
  • the third network element determines the target decision information in the decision information (that is, the target operation) through the intended business demands or the historical experience information achieved by the intention of the intention, or determines the target decision in the decision information through a machine learning model information. Then the third network element sends an action selection response carrying target decision information or identification information of the target decision information to the first network element, so that the first network element determines the multiple actions according to the target decision information or the information identification Target action information in the message.
  • the embodiment of the present application adds a selection process of action information to determine the solution of target action information, so that the first network element can quickly
  • the determination of the target action information and the accuracy of the action information selection further improve the efficiency of the first network element’s intention translation.
  • FIG. 7 is a schematic flowchart of an information processing method provided by an embodiment of the present application. As shown in the figure, the information processing method includes:
  • the first network element obtains multiple action information corresponding to the intent information.
  • the process for the first network element to obtain multiple action information corresponding to the intent information is the same as step 201 shown in FIG. 3, and will not be repeated here.
  • the first network element determines decision information of the multiple pieces of action information.
  • the decision information includes the optimization gains of the multiple actions, the optimization gain identification corresponding to the optimization gain, the negative influence, and the negative influence identification corresponding to the negative influence.
  • the optimized gain identifier is the correlation information between the optimized gain and the action information, that is, the action information corresponding to the optimized gain can be determined through the optimized gain identifier of the optimized gain.
  • the negative impact identification is the associated information between the negative impact and the action information, that is, the action information corresponding to the negative impact can be determined through the negative impact identification of the negative impact.
  • the optional decision information includes the operations of the multiple actions and the operation identifiers corresponding to the operations.
  • the operation identifier is the association information between the operation and the action information, that is, the operation information corresponding to the operation can be determined by the operation identifier of the operation.
  • the first network element determines target action information according to the decision information of the multiple action information.
  • the process of determining the target action information by the first network element according to the decision information of the multiple action information may specifically include 6031-6033 as shown in FIG. 7.
  • the first network element sends an action selection request to the third network element.
  • the action selection request carries decision information of the multiple actions.
  • the third network element determines target decision information according to the action selection request.
  • the determination of the target decision information by the third network element according to the action selection request may include: the third network element obtains the service request corresponding to the intent information; and then according to the The decision information and the appeal information determine the target decision information.
  • the third network element determines the acceptable optimization gain and negative effects in the decision information according to the service request information Impact, and unacceptable negative impact; then the target decision information is determined according to the acceptable optimization gain and negative impact, and the unacceptable negative impact.
  • the business demand is to protect a concert in a venue.
  • the unacceptable negative impact of the business demand is the reduction of cell capacity.
  • Other negative impacts and optimization gains are acceptable.
  • When determining the target action information only multiple action information is required.
  • the negative effects in the medium include the elimination of the action information of the cell capacity reduction, and the remaining action information is the target action information.
  • the third network element determines the acceptable optimization gain and negative impact in the decision information according to the service demand information, and Before the unacceptable negative impact, the third network element may determine an acceptable operation according to the service requirement, and if the acceptable operation is one, determine the acceptable operation as the target decision Information; if there are multiple acceptable operations, continue to perform operations in which the third network element determines acceptable optimization gains and negative effects in the decision information according to the service demand information.
  • determining the target decision information by the third network element according to the action selection request may include: obtaining, by the third network element, historical experience information of intention achievement corresponding to the intention information ; Then determine the target decision information according to the decision information and the historical experience information of the intention to achieve.
  • the third network element determines the acceptable optimization gain and negative influence in the decision information according to the historical experience information of the intention achievement, and the unacceptable Accepted negative impact; and then determine the target decision information based on the acceptable optimization gain and negative impact, as well as the unacceptable negative impact.
  • the third network element determines the acceptable optimization gain and negative effects in the decision information according to the historical experience information achieved by the intention. Before the impact and unacceptable negative impact, the third network element may determine an acceptable operation according to the service requirement, and if the acceptable operation is one, then determine that the acceptable operation is all The target decision information; if there are multiple acceptable operations, continue to execute the third network element to determine acceptable optimization gains and negative effects in the decision information according to the historical experience information achieved by the intention, And operations that have unacceptable negative effects.
  • the third network element determining target decision information according to the action selection request may include: the intention-driven management service consumer inputs the decision information into a machine learning model, and according to The output of the machine learning model determines the target decision information.
  • the machine learning model can be an abstract modeling of historical experience, and a two-layer neural network can be selected.
  • the input is the optimization gain in the decision information, the negative impact and operation information, and the output is the acceptable optimization gain. , Acceptable negative effects or actions that need to be performed.
  • the third network element sends an action selection response to the first network element.
  • the third network element generates the action selection response according to the decision information, and sends an action selection response to the first network element, and the action selection response carries the target Decision information, or identification information that carries the target decision information.
  • the identification information includes the operation identifier of the operation; if the decision information includes an acceptable optimization gain and/or negative impact, and includes an unacceptable optimization gain and/or negative impact, Then the indication information includes the acceptable optimization gain indicator and/or negative impact indicator, and unacceptable optimization gain indicator and/or negative impact indicator.
  • the first network element sends an operation execution request to the network element.
  • the first network element After receiving the action selection response, the first network element determines a target action according to the information carried in the action selection response.
  • steps 605-609 shown in FIG. 7 is similar to steps 105-109 shown in FIG. 2, and therefore will not be described again.
  • the first network element when there are multiple solutions (that is, multiple action information) in the intention translation process that can satisfy the same intention at the same time, the first network element sends to the third network element optimized gains and negatives that carry multiple actions.
  • Action selection request that affects (that is, decision information).
  • the third network element achieves the goal decision information in the decision information (that is, the acceptable optimization gain and negative impact and unacceptable negative impact) through the intended business appeal or the historical experience information achieved by the intended intention, or through the machine
  • the learning model is used to determine the target decision information in the decision information.
  • the third network element sends an action selection response carrying target decision information or identification information of the target decision information to the first network element, so that the first network element determines the multiple actions according to the target decision information or the information identification Target action information in the message.
  • the embodiment of the present application adds a selection process of action information to determine the solution of target action information, so that the first network element can quickly
  • the determination of the target action information and the accuracy of the action information selection further improve the efficiency of the first network element’s intention translation.
  • communication service consumers may include end users, small and medium enterprises, large enterprises, vertical industries, and other communication service providers.
  • IDMS consumers mainly issue intents
  • IDMS providers mainly complete intent translation and convert intents into network management tasks or strategies.
  • the intentions can be divided into three categories: communication service consumer intentions, communication service provider intentions, and network operator intentions.
  • each upper and lower layers can mutually act as a pair of IDMS consumers and IDMS providers.
  • the upper layer is the IDMS consumer and the lower layer is the IDMS provider.
  • IDMS consumers mainly issue intents and add auxiliary translation functions, which can be deployed in business support systems (BSS), network management systems (NMS), and network management systems.
  • BSS business support systems
  • NMS network management systems
  • Element management system Element Management System
  • EMS RAN cluster controller
  • CLRC Cluster RAN Controller
  • the IDMS provider mainly completes intent translation and intent execution and location, which can be deployed on NMS, EMS, CLRC, and network elements (NE).
  • the knowledge base mainly stores network operation and maintenance knowledge, which can be deployed in BSS, NMS, EMS, and independently deployed in servers or operator data centers;
  • the network element can be the network element eNodeB under the 4G LTE standard, and the network element CU/DU/gNodeB of 5G.
  • the logical network element auxiliary translation system mainly completes auxiliary translation, and can be deployed in BSS, NMS, EMS, CLRC, and independently deployed in servers or operator data centers.
  • each network element such as the aforementioned first network element, second network element, and third network element, includes hardware structures and/or software modules corresponding to each function in order to realize the aforementioned functions.
  • the present application can be implemented in the form of hardware or a combination of hardware and computer software. Whether a certain function is executed by hardware or computer software-driven hardware depends on the specific application and design constraint conditions of the technical solution. Professionals and technicians can use different methods for each specific application to implement the described functions, but such implementation should not be considered beyond the scope of this application.
  • FIG. 9 is a schematic structural diagram of an information processing device provided by an embodiment of the application.
  • the information processing device may be used to perform the operations of the first network element in FIG. 2 to FIG. 7.
  • the information processing device 900 includes at least: an acquisition module 910, a first determination module 920, and a second determination module 930; wherein,
  • the obtaining module 910 is used to obtain multiple action information of the intent
  • the first determining module 920 is configured to determine decision information of the multiple action information, where the decision information is information included in the multiple action information;
  • the second determining module 930 is configured to determine target action information according to the decision information of the multiple action information, where the target action information is used to realize the intention.
  • the information processing device in the embodiment of the present application obtains multiple intended action information, then extracts decision information from the multiple pieces of information, and determines target action information in the multiple pieces of action information according to the obtained decision information. It can be seen that the information processing device quickly determines the target action information for realizing the intention by adding auxiliary selection functions and processes, thereby improving the efficiency of intention translation.
  • the second determining module 930 is specifically configured to obtain the business appeal of the intention or historical experience information of the intention achievement of the intention, and the business appeal refers to the corresponding to the intention Business requirements, the historical experience information is the same intent translation event information as the current scenario, the same scenario means that the intent to be realized and the network state are the same; the business appeal according to the intent or the history of the intent to achieve The empirical information and the decision information determine the target action information.
  • the second determining module 930 is specifically configured to input the decision information of the multiple actions into a machine learning model to obtain the target action information.
  • the second determining module 930 is specifically configured to send an action selection request to a second network element or a third network element, and the action selection request carries decision information of the multiple action information ,
  • the action selection request is used by the second network element or the third network element to determine target decision information according to the decision information of the multiple action information; receiving the action selection from the second network element or the third network element
  • the action selection response carries the target decision information or identification information of the target decision information; the target action information is determined according to the target decision information or the identification information.
  • the first determining module 920 is specifically configured to determine the decision information of the multiple actions according to the optimization gains and negative effects of the multiple action information, and the optimization gains include actions and actions.
  • the negative influence includes an action and an action object; or the decision information of the multiple actions is determined according to the operation of the multiple action information, and the operation information includes the action and the action object.
  • the first determining module 920 is specifically configured to determine the decision information of the multiple actions according to the optimized gains, negative effects and operations of the multiple action information.
  • Fig. 10 is a schematic structural diagram of an information processing device according to an embodiment of the application.
  • the information processing device can be used to perform the operations of the second network element in Figs. 2-7.
  • the information processing device 1000 includes at least: a receiving module 1010, a determining module 1020, and a sending module 1030, where:
  • the receiving module 1010 is configured to receive an action selection request from a first network element, where the action selection request includes decision information of multiple action information, and the multiple action information is obtained by the first network element according to an intention ,
  • the decision information is information included in the multiple action information
  • the determining module 1020 is configured to determine target decision information according to decision information of multiple action information, and the target decision information is used to determine the target action;
  • the sending module 1030 is configured to send an action selection response to the first network element, where the action selection response carries the target decision information or carries identification information of the target decision information.
  • the information processing device in the embodiment of the present application generally receives an action selection request from a first network element, where the action selection request carries decision information; then, the target decision information in the decision information is determined, and finally the first network element is sent to the first network element.
  • the network element sends the target decision information to generate an action selection response, so that the first network element determines target action information according to the target decision information. It can be seen that the information processing device helps the first network element to determine the target action information for achieving the intention by adding auxiliary selection functions and processes, thereby improving the efficiency of intention translation.
  • the determining module 1020 is specifically configured to send the second network element to the third network element when the second network element cannot process the action selection request Action selection request; receiving the action selection response from the third network element.
  • the determining module 1020 is specifically configured to obtain the business appeal of the intention or the historical experience information of the intention fulfillment of the intention, and the business appeal refers to the business corresponding to the intention. It is required that the historical experience information is the same intent-to-translate event information as the current scenario, the same scenario means that the intent to be realized and the network state are the same; according to the decision information of the multiple actions and the request information or The historical experience information determines the target decision information.
  • the determining module 1020 is specifically configured to input the decision information of the multiple actions into a machine learning model by the second network element to obtain the target decision information.
  • the decision information of the multiple action information includes the operation of the multiple action information, or the decision information of the multiple action information includes the optimization gains and negatives of the multiple action information.
  • Influence, or the decision information of the multiple action information includes the operation of the multiple action information, the optimization gain and the negative impact of the multiple action information, the optimization gain includes the action and the action object, and the negative impact includes Actions and action objects, and the operation information includes actions and action objects.
  • FIG. 11 is a schematic structural diagram of an information processing device according to an embodiment of the present application.
  • the information processing device may be used to perform the operations of the third network element in FIGS. 2-7.
  • the information processing device 1100 includes at least: a receiving module 1110, a determining module 1120, and a sending module 1130, where:
  • the receiving module 1110 is configured to receive an action selection request from a first network element or a second network element, the action selection request, the action selection request including decision information of multiple action information, and the multiple action information Obtained by the first network element according to the intention, the decision information is information included in the multiple action information;
  • the determining module 1120 is configured to determine target decision information according to decision information of multiple action information, and the target decision information is used to determine the target action;
  • the sending module 1130 is configured to send an action selection response to the first network element or the second network element, where the action selection response carries the target decision information or carries identification information of the target decision information.
  • the information processing device in the embodiment of the application generally receives an action selection request from the first network element or the second network element, wherein the action selection request carries decision information; then the target decision information in the decision information is determined, and finally Sending an action selection response carrying target decision information to the first network element or the second network element, so that the first network element determines target action information according to the target decision information. It can be seen that the information processing device helps the first network element to determine the target action information for achieving the intention by adding auxiliary selection functions and processes, thereby improving the efficiency of intention translation.
  • the determining module 1120 is specifically configured to obtain the business request of the intention or the historical experience information of the intention achievement of the intention, and the business request refers to the business corresponding to the intention. It is required that the historical experience information is the same intent-to-translate event information as the current scenario, the same scenario means that the intent to be realized and the network state are the same; according to the decision information of the multiple actions and the request information or The historical experience information determines the target decision information.
  • the determining module 1120 is specifically configured to input the decision information of the multiple actions into a machine learning model by the second network element to obtain the target decision information.
  • the decision information of the multiple action information includes the operation of the multiple action information, or the decision information of the multiple action information includes the optimization gains and negatives of the multiple action information.
  • Influence, or the decision information of the multiple action information includes the operation of the multiple action information, the optimization gain and the negative impact of the multiple action information, the optimization gain includes the action and the action object, and the negative impact includes Actions and action objects, and the operation information includes actions and action objects.
  • FIG. 12 is a schematic structural diagram of an information processing device provided by an embodiment of the present application.
  • the information processing device 1200 includes at least a processor 1210, a transceiver 1220, and a memory 1230.
  • the processor 1210 and the transceiver 1220 and memory 1230 are connected to each other through a bus 1240, where,
  • the processor 1210 may be a central processing unit (CPU), or a combination of a CPU and a hardware chip.
  • the aforementioned hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof.
  • ASIC application-specific integrated circuit
  • PLD programmable logic device
  • the above-mentioned PLD may be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL) or any combination thereof.
  • CPLD complex programmable logic device
  • FPGA field-programmable gate array
  • GAL generic array logic
  • the transceiver 1220 may include a receiver and a transmitter, for example, a radio frequency module.
  • the processor 1210 described below receives or sends a message, which can be specifically understood as the processor 1210 receiving or sending through the transceiver. .
  • the memory 1230 includes, but is not limited to, random access memory (Random Access Memory, RAM), read-only memory (Read-Only Memory, ROM), or Erasable Programmable Read-Only Memory (EPROM). Or flash memory), the memory 1230 is used to store related instructions and data, and can transmit the stored data to the processor 1210.
  • RAM Random Access Memory
  • ROM read-only memory
  • EPROM Erasable Programmable Read-Only Memory
  • flash memory the memory 1230 is used to store related instructions and data, and can transmit the stored data to the processor 1210.
  • the processor 1210 in the information processing device 1200 is configured to read related instructions in the memory 1230 to perform the following operations:
  • the processor 1210 controls the receiver in the transceiver 1220 to receive a plurality of intended action information
  • the processor 1210 determines decision information of the multiple pieces of action information, where the decision information is information included in the multiple pieces of action information;
  • the processor 1210 determines target action information according to the decision information of the multiple action information, where the target action information is used to realize the intention.
  • FIG. 13 is a schematic structural diagram of an information processing device according to an embodiment of the present application.
  • the information processing device 1300 includes at least a processor 1310, a transceiver 1320, and a memory 1330.
  • the processor 1310 and the transceiver 1320 and memory 1330 are connected to each other through a bus 1340, where
  • the processor 1310 may be a central processing unit (CPU), or a combination of a CPU and a hardware chip.
  • the aforementioned hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof.
  • ASIC application-specific integrated circuit
  • PLD programmable logic device
  • the above-mentioned PLD may be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL) or any combination thereof.
  • the transceiver 1320 may include a receiver and a transmitter, for example, a radio frequency module.
  • the processor 1310 described below receives or sends a certain message. Specifically, it can be understood that the processor 1310 receives or sends a message through the transceiver. .
  • the memory 1330 includes, but is not limited to, Random Access Memory (RAM), Read-Only Memory (Read-Only Memory, ROM), or Erasable Programmable Read-Only Memory (EPROM). Or flash memory), the memory 1330 is used to store related instructions and data, and can transmit the stored data to the processor 1310.
  • RAM Random Access Memory
  • ROM Read-Only Memory
  • EPROM Erasable Programmable Read-Only Memory
  • flash memory the memory 1330 is used to store related instructions and data, and can transmit the stored data to the processor 1310.
  • the processor 1310 in the information processing device 1300 controls the receiver in the transceiver 1320 to receive an action selection request from the first network element.
  • the action selection request includes decision information of multiple action information, and the multiple actions The information is obtained by the first network element according to the intention, and the decision information is the information included in the plurality of action information;
  • the processor 1310 determines target decision information according to decision information of multiple action information, where the target decision information is used to determine the target action;
  • the processor 1310 sends an action selection response to the first network element through a transmitter in the transceiver 1320, where the action selection response carries the target decision information or carries identification information of the target decision information.
  • FIG. 14 is a schematic structural diagram of an information processing device provided by an embodiment of the present application.
  • the information processing device 1400 includes at least a processor 1410, a transceiver 1420, and a memory 1430.
  • the processor 1410 and the transceiver 1420 and memory 1430 are connected to each other through a bus 1440, where,
  • the processor 1410 may be a central processing unit (CPU), or a combination of a CPU and a hardware chip.
  • the aforementioned hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof.
  • ASIC application-specific integrated circuit
  • PLD programmable logic device
  • the above-mentioned PLD may be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL) or any combination thereof.
  • the transceiver 1420 may include a receiver and a transmitter, for example, a radio frequency module.
  • the processor 1410 described below receives or sends a certain message. Specifically, it can be understood that the processor 1410 receives or sends a message through the transceiver. .
  • the memory 1430 includes, but is not limited to, random access memory (Random Access Memory, RAM), read-only memory (Read-Only Memory, ROM), or Erasable Programmable Read-Only Memory (EPROM). Or flash memory), the memory 1430 is used to store related instructions and data, and can transmit the stored data to the processor 1410.
  • RAM Random Access Memory
  • ROM read-only memory
  • EPROM Erasable Programmable Read-Only Memory
  • flash memory the memory 1430 is used to store related instructions and data, and can transmit the stored data to the processor 1410.
  • the processor 1410 in the information processing device 1400 is configured to read related instructions in the memory 1430 to perform the following operations:
  • the processor 1410 receives an action selection request from the first network element or the second network element through the receiver in the transceiver 1420, the action selection request, the action selection request includes decision information of multiple action information, the The multiple pieces of action information are obtained by the first network element according to intentions, and the decision information is the information included in the multiple pieces of action information;
  • the processor 1410 determines target decision information according to decision information of multiple action information, and the target decision information is used to determine the target action;
  • the processor 1410 sends an action selection response to the first network element or the second network element through the transmitter in the transceiver 1420, and the action selection response carries the target decision information or the target decision information. Identification information.
  • the foregoing information processing apparatus 1400 reference may be made to the specific operations of the third network element in the foregoing method embodiment, which will not be repeated here.
  • the computer may be implemented in whole or in part by software, hardware, firmware, or any combination thereof.
  • software it can be implemented in the form of a computer program product in whole or in part.
  • the computer program product includes one or more computer instructions.
  • the computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.
  • the computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium.
  • the computer instructions may be transmitted from a website, computer, server, or data center.
  • the computer-readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server or data center integrated with one or more available media.
  • the usable medium may be a magnetic medium (for example, a floppy disk, a hard disk, and a magnetic tape), an optical medium (for example, a DVD), or a semiconductor medium (for example, a solid state disk (SSD)).

Landscapes

  • Business, Economics & Management (AREA)
  • Human Resources & Organizations (AREA)
  • Engineering & Computer Science (AREA)
  • Economics (AREA)
  • Strategic Management (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Development Economics (AREA)
  • Educational Administration (AREA)
  • General Business, Economics & Management (AREA)
  • Game Theory and Decision Science (AREA)
  • Marketing (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Machine Translation (AREA)
  • Information Transfer Between Computers (AREA)

Abstract

本申请实施例公开了一种信息处理方法、相关设备及计算机存储介质，其中方法包括：获取意图的多个动作信息；确定所述多个动作信息的决策信息，所述决策信息为所述多个动作信息中包含的信息；根据所述多个动作信息的决策信息确定目标动作信息，所述目标动作信息用于实现所述意图。本申请实施例通过增加辅助选择功能以及流程来快速确定实现意图的目标动作信息，从而提升意图转译的效率。

Description

一种信息处理方法、相关设备及计算机存储介质
本申请要求于2019年09月30日提交中国国家知识产权局、申请号为201910945933.0、申请名称为“一种信息处理方法、相关设备及计算机存储介质”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本申请涉及信息处理技术领域，尤其涉及一种信息处理方法、相关设备及计算机存储介质。
背景技术
随着机器学习和大数据分析等技术的发展，基于意图的网络管理在无线网络中也得到应用，旨在减少人机交互并提高网络性能。与现有的基于人的网络管理相比，基于意图的网络管理使用简单的指令，告诉网络“做什么”而不是“怎么做”，以减少大规模的手动决策时的不准确性。此外，垂直行业的无线网络运维场景下，垂直行业用户的运维知识不足，基于意图的网络管理可降低用户对运维知识的需求。其中，上述告诉网络“做什么”，即要实现的意图，例如“提升x区用户视频业务体验”；而从“做什么”到“怎么做”以及执行“怎么做”的过程则为意图转译，即将意图转换为合理的网络部署并执行部署的动作的过程。
目前意图转译主要通过意图驱动管理提供者接收意图驱动管理服务消费者下发的意图，向知识库查询为满足意图所需执行的动作及动作执行条件，知识库进行目标与意图匹配后，将结果反馈给意图驱动管理服务消费者。意图驱动管理服务提供者向网元查询当前网络的状态，并下发该状态下需要执行的动作。
但是无线网络的意图主要面向运维管理，且多数意图可通过特性运维得到满足。无线网络各特性不是完全正交，存在多个方案可同时满足同一意图的场景，意图驱动管理服务提供者在方案选择时，通过随机选择，或逐个尝试的方法进行方案选择，从而导致目前的意图转译的转译效率低。
发明内容
本申请实施例公开了一种信息处理方法、相关设备及计算机存储介质，通过增加辅助选择功能以及流程来快速确定实现意图的目标动作信息，从而提升意图转译的效率。
第一方面，本申请实施例公开提供了数据传输方法，包括：
第一网元获取意图的多个动作信息；
所述第一网元确定所述多个动作信息的决策信息，所述决策信息为所述多个动作信息中包含的信息；
所述第一网元根据所述多个动作信息的决策信息确定目标动作信息，所述目标动作信息用于实现所述意图。
在一种可能的实施方式中，所述第一网元根据所述多个动作信息的决策信息确定目标动作信息，包括：
所述第一网元获取所述意图的业务诉求或所述意图的意图达成的历史经验信息，所述 业务诉求是指与所述意图对应的业务要求，所述历史经验信息为与当前场景相同的意图转译事件的信息，所述场景相同是指要实现的意图以及网络状态相同；
所述第一网元根据所述意图的业务诉求或所述意图达成的历史经验信息以及所述决策信息确定所述目标动作信息。
在一种可能的实施方式中，所述第一网元根据所述多个动作信息的决策信息确定目标动作信息，包括：
所述第一网元将所述多个动作的决策信息输入机器学习模型，得到所述目标动作信息。
在一种可能的实施方式中，所述根据所述多个动作信息的决策信息确定目标动作信息，包括：
所述第一网元向第二网元或第三网元发送动作选择请求，所述动作选择请求携带所述多个动作信息的决策信息，所述动作选择请求用于第二网元或第三网元根据所述多个动作信息的决策信息确定目标决策信息；
所述第一网元接收来自所述第二网元或所述第三网元的动作选择响应，所述动作选择响应携带所述目标决策信息或所述目标决策信息的标识信息；
所述第一网元根据所述目标决策信息或所述标识信息确定所述目标动作信息。
在一种可能的实施方式中，所述第一网元确定所述多个动作信息的决策信息，包括：
所述第一网元根据所述多个动作信息的优化增益和负面影响确定所多个动作的决策信息，所述优化增益包含动作和动作对象，所述负面影响包含动作和动作对象；
或者所述第一网元根据所述多个动作信息的操作确定所多个动作的决策信息，所述操作信息包含动作和动作对象。
在一种可能的实施方式中，所述第一网元根据所述多个动作信息的优化增益和负面影响确定所多个动作的决策信息，包括：
所述第一网元根据所述多个动作信息的优化增益和负面影响以及操作确定所多个动作的决策信息。
第二方面，本申请实施例提供一种信息处理方法，包括：
第二网元接收来自第一网元的动作选择请求，所述动作选择请求包括多个动作信息的决策信息，所述多个动作信息为所述第一网元根据意图获取的，所述决策信息为所述多个动作信息中包含的信息；
所述第二网元根据多个动作信息的决策信息确定目标决策信息，所述目标决策信息用于确定所述目标动作；
所述第二网元向所述第一网元发送动作选择响应，所述动作选择响应携带所述目标决策信息，或者携带所述目标决策信息的标识信息。
在一种可能的实施方式中，所述第二网元根据多个动作信息的决策信息确定目标决策信息，包括：
在所述第二网元不能处理所述动作选择请求的情况下，所述第二网元向第三网元发送所述动作选择请求；
所述第二网元接收来自所述第三网元的所述动作选择响应。
在一种可能的实施方式中，所述第二网元根据多个动作信息的决策信息确定目标决策 信息，包括：
所述第二网元获取所述意图的业务诉求或所述意图的意图达成的历史经验信息，所述业务诉求是指与所述意图对应的业务要求，所述历史经验信息为与当前场景相同的意图转译事件的信息，所述场景相同是指要实现的意图以及网络状态相同；
所述第二网元根据所述多个动作的决策信息以及所述诉求信息或所述历史经验信息确定所述目标决策信息。
在一种可能的实施方式中，所述第二网元根据多个动作信息的决策信息确定目标决策信息，包括：
所述第二网元将所述多个动作的决策信息输入机器学习模型，得到所述目标决策信息。
在一种可能的实施方式中，所述多个动作信息的决策信息包括所述多个动作信息的操作，或者所述多个动作信息的决策信息包括所述多个动作信息的优化增益和负面影响，或者所述多个动作信息的决策信息包括所述多个动作信息的操作、所述多个动作信息的优化增益和负面影响，所述优化增益包含动作和动作对象，所述负面影响包含动作和动作对象，所述操作信息包含动作和动作对象。
第三方面，本申请实施例提供一种信息处理方法，包括：
第三网元接收来自第一网元或第二网元的动作选择请求，所述动作选择请求，所述动作选择请求包括多个动作信息的决策信息，所述多个动作信息为所述第一网元根据意图获取的，所述决策信息为所述多个动作信息中包含的信息；
第三网元根据多个动作信息的决策信息确定目标决策信息，所述目标决策信息用于确定所述目标动作；
第三网元向所述第一网元或第二网元发送动作选择响应，所述动作选择响应携带所述目标决策信息，或者携带所述目标决策信息的标识信息。
在一种可能的实施方式中，第三网元根据多个动作信息的决策信息确定目标决策信息，包括：
所述第三网元获取所述意图的业务诉求或所述意图的意图达成的历史经验信息，所述业务诉求是指与所述意图对应的业务要求，所述历史经验信息为与当前场景相同的意图转译事件的信息，所述场景相同是指要实现的意图以及网络状态相同；
所述第三网元根据所述多个动作的决策信息以及所述诉求信息或所述历史经验信息确定所述目标决策信息。
在一种可能的实施方式中，第三网元根据多个动作信息的决策信息确定目标决策信息，包括：
所述第二网元将所述多个动作的决策信息输入机器学习模型，得到所述目标决策信息。
在一种可能的实施方式中，所述多个动作信息的决策信息包括所述多个动作信息的操作，或者所述多个动作信息的决策信息包括所述多个动作信息的优化增益和负面影响，或者所述多个动作信息的决策信息包括所述多个动作信息的操作、所述多个动作信息的优化增益和负面影响，所述优化增益包含动作和动作对象，所述负面影响包含动作和动作对象，所述操作信息包含动作和动作对象。
在本申请实施例中，当意图转译过程中存在多个方案(即多个动作信息)可以同时满 足同一意图时，第一网元通过将多个动作的优化增益和负面影响作为决策信息，然后通过意图的业务诉求或所述意图的意图达成的历史经验信息所述决策信息中的目标决策信息(即可接受的优化增益和负面影响以及不可接受的负面影响)，或者通过机器学习模型来确定所述决策信息中的目标决策信息，进一步根据所述目标决策信息确定所述多个动作信息中的目标动作信息。可以看出，相比于已有的随机选择动作信息或逐个尝试动作信息的方案，本申请实施例通过增加动作信息的选择流程来确定目标动作信息的方案，可以快速的确定目标动作信息，以及动作信息选择的准确性，从而提升意图转译的效率。
第四方面，本申请实施例提供一种信息处理装置，包括：
获取模块，用于获取意图的多个动作信息；
第一确定模块，用于确定所述多个动作信息的决策信息，所述决策信息为所述多个动作信息中包含的信息；
第二确定模块，用于根据所述多个动作信息的决策信息确定目标动作信息，所述目标动作信息用于实现所述意图。
本申请实施例中所述信息处理装置通过获取意图的多个动作信息，然后从多个信息中提取出决策信息，并根据得到的决策信息来确定所述多个动作信息中的目标动作信息。可见，所述信息处理装置通过增加辅助选择功能以及流程来快速确定实现意图的目标动作信息，从而提升意图转译的效率。
在一种可选的实施方式中，所述第二确定模块，具体用于获取所述意图的业务诉求或所述意图的意图达成的历史经验信息，所述业务诉求是指与所述意图对应的业务要求，所述历史经验信息为与当前场景相同的意图转译事件的信息，所述场景相同是指要实现的意图以及网络状态相同；根据所述意图的业务诉求或所述意图达成的历史经验信息以及所述决策信息确定所述目标动作信息。
在一种可选的实施方式中，所述第二确定模块，具体用于将所述多个动作的决策信息输入机器学习模型，得到所述目标动作信息。
在一种可选的实施方式中，所述第二确定模块，具体用于向第二网元或第三网元发送动作选择请求，所述动作选择请求携带所述多个动作信息的决策信息，所述动作选择请求用于第二网元或第三网元根据所述多个动作信息的决策信息确定目标决策信息；接收来自所述第二网元或所述第三网元的动作选择响应，所述动作选择响应携带所述目标决策信息或所述目标决策信息的标识信息；根据所述目标决策信息或所述标识信息确定所述目标动作信息。
在一种可选的实施方式中，所述第一确定模块，具体用于根据所述多个动作信息的优化增益和负面影响确定所多个动作的决策信息，所述优化增益包含动作和动作对象，所述负面影响包含动作和动作对象；或者根据所述多个动作信息的操作确定所多个动作的决策信息，所述操作信息包含动作和动作对象。
在一种可选的实施方式中，所述第一确定模块，具体用于根据所述多个动作信息的优化增益和负面影响以及操作确定所多个动作的决策信息。
第五方面，本申请实施例还提供一种信息处理装置，包括：
接收模块，用于接收来自第一网元的动作选择请求，所述动作选择请求包括多个动作 信息的决策信息，所述多个动作信息为所述第一网元根据意图获取的，所述决策信息为所述多个动作信息中包含的信息；
确定模块，用于根据多个动作信息的决策信息确定目标决策信息，所述目标决策信息用于确定所述目标动作；
发送模块，用于向所述第一网元发送动作选择响应，所述动作选择响应携带所述目标决策信息，或者携带所述目标决策信息的标识信息。
本申请实施例中所述信息处理装置通接收来自第一网元的动作选择请求，其中所动作选择请求携带决策信息；然后确定出所述决策信息中的目标决策信息，最后向所述第一网元发送携带目标决策信息生成动作选择响应，以使所述第一网元根据所述目标决策信息确定目标动作信息。可见，所述信息处理装置通过增加辅助选择功能以及流程来帮助所述第一网元确定实现意图的目标动作信息，从而提升意图转译的效率。
在一种可选的实施方式中，所述确定模块，具体用于在所述第二网元不能处理所述动作选择请求的情况下，所述第二网元向第三网元发送所述动作选择请求；接收来自所述第三网元的所述动作选择响应。
在一种可选的实施方式中，所述确定模块，具体用于获取所述意图的业务诉求或所述意图的意图达成的历史经验信息，所述业务诉求是指与所述意图对应的业务要求，所述历史经验信息为与当前场景相同的意图转译事件的信息，所述场景相同是指要实现的意图以及网络状态相同；根据所述多个动作的决策信息以及所述诉求信息或所述历史经验信息确定所述目标决策信息。
在一种可选的实施方式中，所述确定模块，具体用于所述第二网元将所述多个动作的决策信息输入机器学习模型，得到所述目标决策信息。
在一种可选的实施方式中，所述多个动作信息的决策信息包括所述多个动作信息的操作，或者所述多个动作信息的决策信息包括所述多个动作信息的优化增益和负面影响，或者所述多个动作信息的决策信息包括所述多个动作信息的操作、所述多个动作信息的优化增益和负面影响，所述优化增益包含动作和动作对象，所述负面影响包含动作和动作对象，所述操作信息包含动作和动作对象。
第六方面，本申请实施例还提供一种信息处理装置，包括：
接收模块，用于接收来自第一网元或第二网元的动作选择请求，所述动作选择请求，所述动作选择请求包括多个动作信息的决策信息，所述多个动作信息为所述第一网元根据意图获取的，所述决策信息为所述多个动作信息中包含的信息；
确定模块，用于根据多个动作信息的决策信息确定目标决策信息，所述目标决策信息用于确定所述目标动作；
发送模块，用于向所述第一网元或第二网元发送动作选择响应，所述动作选择响应携带所述目标决策信息，或者携带所述目标决策信息的标识信息。
本申请实施例中所述信息处理装置通接收来自第一网元或第二网元的动作选择请求，其中所动作选择请求携带决策信息；然后确定出所述决策信息中的目标决策信息，最后向所述第一网元或第二网元发送携带目标决策信息生成动作选择响应，以使所述第一网元根据所述目标决策信息确定目标动作信息。可见，所述信息处理装置通过增加辅助选择功能 以及流程来帮助所述第一网元确定实现意图的目标动作信息，从而提升意图转译的效率。
在一种可选的实施方式中，所述确定模块，具体用于获取所述意图的业务诉求或所述意图的意图达成的历史经验信息，所述业务诉求是指与所述意图对应的业务要求，所述历史经验信息为与当前场景相同的意图转译事件的信息，所述场景相同是指要实现的意图以及网络状态相同；根据所述多个动作的决策信息以及所述诉求信息或所述历史经验信息确定所述目标决策信息。
在一种可选的实施方式中，所述确定模块，具体用于所述第二网元将所述多个动作的决策信息输入机器学习模型，得到所述目标决策信息。
在一种可选的实施方式中，所述多个动作信息的决策信息包括所述多个动作信息的操作，或者所述多个动作信息的决策信息包括所述多个动作信息的优化增益和负面影响，或者所述多个动作信息的决策信息包括所述多个动作信息的操作、所述多个动作信息的优化增益和负面影响，所述优化增益包含动作和动作对象，所述负面影响包含动作和动作对象，所述操作信息包含动作和动作对象。
第七方面，本申请实施例提供一种信息处理装置，包括处理器和收发器，其中：所述处理器和所述收发器相互连接，所述收发器用于与数据分析设备通信，所述处理器被配置用于执行如第一方面所述的方法。
第八方面，本申请实施例还提供一种信息处理装置，包括处理器和收发器，其中：所述处理器和所述收发器相互连接，所述收发器用于与数据分析设备通信，所述处理器被配置用于执行如第二方面所述的方法。
第九方面，本申请实施例提供一种信息处理装置，包括处理器和收发器，其中：所述处理器和所述收发器相互连接，所述收发器用于与数据分析设备通信，所述处理器被配置用于执行如第三方面所述的方法。
第十方面，本申请实施例提供一种计算机存储介质，所述计算机可读存储介质存储有计算机程序，所述计算机程序被处理器执行时实现如第一方面所述的方法。
第十一方面，本申请实施例还提供一种计算机存储介质，所述计算机可读存储介质存储有计算机程序，所述计算机程序被处理器执行时实现如第二方面所述的方法。
第十二方面，本申请实施例还提供一种计算机存储介质，所述计算机可读存储介质存储有计算机程序，所述计算机程序被处理器执行时实现如第三方面所述的方法。
附图说明
为了更清楚地说明本申请实施例或背景技术中的技术方案，下面将对本申请实施例或背景技术中所需要使用的附图进行说明。
图1是本申请实施例提供的IDMS消费者和IDMS提供者之间的交互示意图；
图2是本申请实施例提供的一种信息处理方法的流程示意图；
图3是本申请实施例提供的另一种信息处理方法的流程示意图；
图4是本申请实施例提供的又一种信息处理方法的流程示意图；
图5是本申请实施例提供的又一种信息处理方法的流程示意图；
图6是本申请实施例提供的又一种信息处理方法的流程示意图；
图7是本申请实施例提供的又一种信息处理方法的流程示意图；
图8是本申请实施例提供的IDM系统分层架构示意图；
图9是本申请实施例提供的一种信息处理装置的结构示意图；
图10是本申请实施例提供的另一种信息处理装置的结构示意图；
图11是本申请实施例提供的另一种信息处理装置的结构示意图；
图12是本申请实施例提供的另一种信息处理装置的结构示意图；
图13是本申请实施例提供的另一种信息处理装置的结构示意图；
图14是本申请实施例提供的另一种信息处理装置的结构示意图。
具体实施方式
下面结合本申请实施例中的附图对本申请实施例进行描述。
本申请实施例提供一种信息处理方法及相关产品，应用于意图驱动管理服务(intent driven management service,IDMS)，在意图转译过程中新增了辅助转译系统或功能，以辅助意图驱动管理服务提供者IDMS Provider完成转译过程中的目标动作选择。
意图驱动管理(intent driven management,IDM)包括意图表达、意图接口和意图的生命周期管理等。意图驱动管理服务(intent driven management service,IDMS)允许IDMS消费者表达管理网络和服务的意图，IDMS提供者进行意图转译，将意图转换为合理的网络部署信息并执行部署操作。IDMS消费者和IDMS提供者之间的交互如图1所示。如图1所示，由IDMS消费者向IDMS提供者下发意图信息，IDMS提供者接收到意图信息后向IDMS消费者反馈意图接收通知，并反馈意图达成情况。
意图表达由动作Action和对象Object构成，形式上为<Intent Driven Action><Intent Driven Object>二元组，其中，Action用来抽象和简化对网络的操作，形式上包含一个动作名和一系列相关属性<intent driven action name>，<intent driven action properties>；Object用来提供管理对象信息，形式上包含一个对象名和一系列用来标识对象的属性<intent driven object name>,<intent driven object properties>。
为提高运维自动化能力和运维效率，在意图表达中引入条件信息，意图表达携带条件信息，主要分为两类：约束条件(post-condition)和生效条件(pre-condition)。约束条件为对指定Object执行Action后需要满足的参数或性能约束，该约束与Action无关；例如，在不影响物理资源块(physical resource block,PRB)利用率的前提下，将区域X的下行用户速率提高至1Gbps；生效条件为对Object执行Action之前需要满足的前提条件，可以是时间限定、网络状态等；例如，当站点X的边缘下行参考信号接收功率(reference signal received power,RSRP)小于-110dbm时，开通覆盖提升特性。
当IDMS消费者将意图下发到IDMS提供者后，IDMS提供者基于意图知识库和当前网络状态进行意图转译。其中，意图知识库设计如表1所示。
表1.意图知识库设计
Figure PCTCN2020117681-appb-000001
Figure PCTCN2020117681-appb-000002
在如表1所示的意图知识库中，各项列名的具体描述如下：
(1)目标：主要描述的是操作可达成的效果，意图知识库的目标对应的是意图表达中的action，当目标与action匹配后，目标对应的表项才是符合意图的。例如目标可以是“提升长期演进网络语言业务(Voice over Long Term Evolution,VoLTE)质量”。
(2)意图满足条件：主要描述的是意图达成所需要满足的条件，意图知识库的意图满足条件是一个布尔表达式。例如，意图满足条件可以是：视频首播时延<10ms，视频平均每次播放卡顿次数<2，视频卡顿时长占比<10％。
(3)操作可执行条件：主要描述的是目标对应的操作可被执行的条件，意图知识库的操作可执行条件是一个布尔表达式，其中的数据是可以从意图所指示的对象中采集到的，采集到的数据代入布尔表达式，即可判断条件是否为真。例如，一个意图知识库的操作可执行条件为：(小区平均用户数>200或PRB利用率>30％)且CCE利用率>50％，当意图模型中的object所指示的小区的平均用户数＝220、PRB利用率＝25％、CCE利用率＝60％时，该条件为真；而当小区的平均用户数＝180、PRB利用率＝25％、CCE利用率＝60％时，该条件为假。
(4)操作类型：意图知识库的操作类型是一个枚举类型，有两个取值：如果为“脚本”，则对应的操作中描述的是一个脚本；如果为“优化”，则对应的操作中描述的是一个优化问题。
(5)操作：主要描述的是具体的目标和操作可执行条件对应的操作，意图知识库的操作是一个字符串，描述了具体的操作，根据操作类型的不同描述的是不同的操作。当操作类型为“脚本”时，操作中描述的是对意图所指示的操作对象的一系列操作指令，如：MML命令；而当操作类型为“优化”时，操作中描述的是对意图所指示的操作对象的一系列操作指令以及指令中需要优化的参数。
(6)备注：备注是一个字符串，可以为空，为该条表项的说明性文字。
在本申请实施例中，IDMS提供者接收IDMS消费者下发的意图信息，然后向知识库查询为满足意图所需执行的动作信息，该动作信息包括操作、操作执行条件、意图达成条件以及优化增益和负面影响等信息中的一个或多个；知识库接收到IDMS提供者的查询请求后，其请求中包括目标，知识库进行目标与意图匹配后，将结果(即所述动作信息)反馈给IDMS提供者。接着，IDMS提供者向网元查询当前网络的状态，并下发该状态下需要执行的操作，以便达成意图。
由于无线网络各特性不是完全正交，因此存在多个动作信息可达成意图，即存在多个方案可同时满足同一意图的场景，如表2所示，为本申请实施例提供的一种多个动作信息可同时达成意图的知识库示例。
表2 多个动作信息可同时达成意图的知识库示例
Figure PCTCN2020117681-appb-000003
Figure PCTCN2020117681-appb-000004
如表2所示，对于目标“提升用户视频业务体验”存在三个动作信息，所示三个动作信息的检测指标、意图满足条件、操作可执行条件以及操作类型均相同，具体的操作以及备注(优化增益和负面影响)不同。
其中，所述检测指标是指所述意图满足条件和所述操作可执行条件中涉及到的关键性能指标(key performance indicator,KPI)，所述意图满足条件的KPI为意图监测KPI，所述操作可执行条件的KPI为条件KPI。
另外，表2中，操作可执行条件中的CCE-AvgUtilizationRate表示控制信道元素平均利用率，DL-Packet-Delay表示下行包延时，DL-User-Thp(<5M)-SampIndex表示下行用户吞吐率低于5M比例。三个动作信息的操作的含义分别为：①、打开上行VOIP调度优化开关，保证在发生SR漏检时，也能及时对语音用户进行上行调度，避免PDCP丢包定时器超时丢包，同时降低VOLTE SINR修正算法的IBLER目标值到5，使MCS选阶更保守，降低丢包率；②、配置PDCCH初始符号数相关参数，以及固定CFI＝3，保证可用CCE资源最大化；③、打开上行连续调度开关，控制VoLTE用户在上行通话期进行连续调度，以减小弱覆盖场景下上行调度时延，从而减小VoLTE业务的上行包时延和包抖动，改善语音质量，以及打开下行重传降阶开关，仅对最后两次重传进行降TBS索引调度，其他重传根据CQI调整的结果确定重传TBS索引，降低下行RBLER，减少语音下行丢包。
另外，在IDMS消费者向IDMS提供者下发意图信息时，意图信息中携带的条件信息(可以包括所述约束条件和生效条件)是可选的，条件信息中不一定包含于优化增益及负面影响信息匹配的信息；因此，当IDMS提供者通过知识库查询到目标存在多个动作信息可达成意图的情况下，IDMS提供者在选择动作信息时没有可供参考的信息；在这种情况下，一般采用随机选择，或者逐个尝试的方法进行动作信息的选择，这将导致意图转译效率降低。
为了解决上述问题，本申请实施例提出了一种信息处理方法，用于辅助所述IDMS提供者选择动作信息，下面对本申请实施例进行详细介绍。需要说明说的是，下文中提到的第一网元是指在意图转译过程中，将意图转换成网络管理任务或策略的一方，即意图驱动 管理服务IDMS中的IDMS提供者；第二网元是指意图转译过程中下发或提出意图的一方，即意图驱动管理服务IDMS中的IDMS消费者；第三网元是指用于辅助所述第一网元进行动作信息选择的网元。
请参阅图2，图2是本申请实施例提供的一种信息处理方法的流程示意图，如图所示，本信息处理方法包括：
101：第一网元获取意图信息对应的多个动作信息。
在本申请实施例中，所述意图信息为与意图相关的信息，所述意图信息可以携带意图标识、意图action和意图object等信息。每个动作信息包括操作、操作可执行条件以及意图满足条件等信息，不包括操作的优化增益和负面影响。
其中，各个动作信息之间的操作不同，各个动作信息之间的操作可执行条件以及意图满足条件可以相同也可以不同。每个动作信息对应一个操作，一个操作可以包括至少一条操作指令，每条操作指令包含动作和动作对象。
例如，表2中存在三个动作信息，相应的对应三个操作，其中操作①中包含了两条操作指令，每条操作指令均以MOD开头。
第一网元获取意图信息对应的多个动作信息的过程，具体可以包括如图2所示的1011-1019。
1011：第一网元接收来自第二网元的意图信息。
第二网元将携带意图标识、意图action和意图object等信息的意图信息下发给第一网元，以使所述第一网元完成意图转译。
1012：第一网元向第二网元发送意图接收通知。
所述意图接收通知用于通知所述第二网元所述意图信息的接收情况，所述接收情况可以包括接收成功和/或接收失败等。
1013：第一网元提取所述意图信息中意图目标的关键词。
在接收到所述意图信息后，所述第一网元对所述意图信息进行词法、语法及语义分析来提取意图目标的关键词。例如，意图信息为：intent1，提升x区用户视频业务体验；则可以提取出意图目标的关键词为：提升、视频、业务体验。其中，所述词法、语法及语义分析可使用计算机领域词法分析器、语法分析器以及语义分析器来实现。
1014：第一网元向知识库发送查询请求。
其中，所述查询请求携带有所述意图目标的关键词，即1013步骤中提取的关键词。所述查询请求用于获取与所述关键词匹配的目标对应的动作信息。
1015：知识库根据所述查询请求进行目标匹配。
所述知识库接收到所述查询请求后，则根据所述查询请求中的关键词进行目标匹配，以得到与所述目标对应的可选动作信息，其中，所述与所述可选动作信息可以包括至少一个动作信息。其中，每个动作信息均包括操作、操作可执行条件、意图满足条件。
所述操作可执行条件包括条件关键性能指标(key performance indicator,KPI)需要满足的条件，其中所述条件KPI是指执行所述操作的前提条件中对应的KPI。例如，所述操作可执行条件可以是：CCE-AvgUtilizationRate<0.7&&DL-Packet-Delay>20ms，其中CCE-AvgUtilizationRate表示控制单元信道元素平均利用率，DL-Packet-Delay表示下行包延 时，CCE-AvgUtilizationRate和DL-Packet-Delay均为条件KPI。
所述意图满足条件包括意图监测KPI要满足的条件，所述意图监测KPI是指意图达成需要监测的KPI。例如，所述意图满足条件可以是：视频首播时延<10ms，且视频平均每次播放卡顿次数<2，且视频卡顿时长占比<10％；视频首播时延、视频平均每次播放卡顿次数和视频卡顿时长占比均为监测KPI。
1016：第一网元接收来自知识库的查询结果反馈。
其中，所述查询结果反馈包括步骤1015中得到的与所述目标对应的动作信息。
1017：第一网元向网元发送意图监测KPI、条件KPI订阅请求。
其中，所述意图监测KPI、条件KPI订阅请求用于获取当前网络状态下网元的意图监测KPI和条件KPI。
1018：第一网元接收来自所述网元的意图监测KPI、条件KPI订阅反馈。
其中，所述意图监测KPI、条件KPI订阅反馈携带所述当前网络状态下网元的意图监测KPI和条件KPI。
1019：第一网元确定与当前网络状态匹配的动作信息，并保存所述意图监测KPI。
在具体实现中，在所述第一网元接收到所述意图监测KPI、条件KPI订阅反馈后，则从所述查询结果反馈中携带的动作信息中，确定与当前网络状态下的条件KPI(即所述述意图监测KPI、条件KPI订阅反馈中的条件KPI)匹配的活动信息，从而得到所述多个动作信息。
102：第一网元确定所述多个动作信息的决策信息。
所述决策信息中包括所述多个动作的操作。
103：第一网元根据所述多个动作信息的决策信息确定目标动作信息。
在一种可选的实施方式中，所述第一网元根据所述多个动作信息的决策信息确定目标动作信息可以包括：所述第一网元获取与所述意图信息对应的业务诉求；然后根据所述决策信息以及所述诉求信息确定所述目标动作。
在具体实现中，在获取到所述业务诉求信息后，所述第一网元根据所述业务诉求信息确定所述决策信息中需要执行的操作；然后根据需要执行的操作确定所述目标动作信息。
在另外一种可选的实施方式中，所述第一网元根据所述多个动作信息的决策信息确定目标动作信息可以包括：所述第一网元获取与所述意图信息对应的意图达成的历史经验信息；然后根据所述决策信息以及所述意图达成的历史经验信息确定所述目标动作。其中，本申请实施例以及下文中提到的所述历史经验信息是指与当前的场景相同的意图转译事件的信息，所述场景相同是指要实现的意图以及网络状态相同；所述业务诉求是指与所述意图信息对应的业务要求。
在具体实现中，在获取到所述意图达成的历史经验信息后，所述第一网元根据所述意图达成的历史经验信息确定所述决策信息中需要执行的操作；然后根据需要执行的操作确定所述目标动作信息。
在又一种可选的实施方式中，所述第一网元根据所述多个动作信息的决策信息确定目标动作信息可以包括：所述意图驱动管理服务消费者将所述决策信息输入机器学习模型，并根据所述机器学习模型的输出确定所述目标动作信息。
所述机器学习模型可以是对历史经验的抽象建模，可以选用两层神经网络，输入为所述决策信息中的操作等信息，输出为可接受的或需要执行的操作。
104：第一网元向网元发送操作执行请求。
其中，所述操作执行请求，携带所述目标动作信息的操作，即所述网元需要执行的操作，例如，该操作可以是MODCELLULSCHALGO.UlEnhencedVoipSchSw:LocalCellId＝％cell_id；UlVoipSchOptSwitch＝ON。
105：第一网元接收来自所述网元的操作执行响应。
所述操作执行响应用于通知所述第一网元是否成功执行所述目标动作的操作。
106：第一网元向网元发送意图监测KPI订阅请求。
所述意图监测KPI订阅请求包括步骤1016中所述查询结果反馈中的意图监测KPI。
107：第一网元接收来自所述网元的意图监测KPI订阅响应。
所述意图监测KPI订阅响应携带所述网元执行所述操作后，所述意图监测KPI对应的值或状态。
108：第一网元对所述操作执行前保存的意图监测KPI和所述操作执行后的意图监测KPI进行比对。
在接收到来自所述网元的意图监测KPI订阅响应后，所述IDMS对所述意图监测KPI在所述操作被执行前后的值和状态(即所述操作执行前所述IDMS保存的所述意图监测KPI的值，和所述操作执行后所述意图监测KPI订阅响应中所述意图监测KPI的值或状态)进行比较，并得到意图监测KPI的比较结果，所述比较结果用于判断所述意图是否达成。
109：第一网元向第二网元发送意图达成情况。
其中，所述意图达成情况可以包括意图已达成和意图未达成。
具体的，当所述第一网元根据所述比较结果判断所述意图已达成时，则所述第一网元向第二网元发送意图已达成的消息；当所述第一网元根据所述比较结果判断所述意图未达成时，则所述第一网元向第二网元发送意图未达成的消息；可选的，所述意图未达成消息可以包括所述意图监测KPI的比较结果。
在本申请实施例中，当意图转译过程中存在多个方案(即多个动作信息)可以同时满足同一意图时，第一网元通过将多个动作的操作作为决策信息，然后通过意图的业务诉求或所述意图的意图达成的历史经验信息所述决策信息中的目标决策信息(即目标操作)，或者通过机器学习模型来确定所述决策信息中的目标决策信息，进一步根据所述目标决策信息确定所述多个动作信息中的目标动作信息。可以看出，相比于已有的随机选择动作信息或逐个尝试动作信息的方案，本申请实施例通过增加动作信息的选择流程来确定目标动作信息的方案，可以使的第一网元可以快速的确定目标动作信息，以及动作信息选择的准确性，进一步提升了第一网元意图转译的效率。
请参阅图3，图3是本申请实施例提供的另一种信息处理方法的流程示意图，如图所示，本信息处理方法包括：
201：第一网元获取意图信息对应的多个动作信息。
在本申请实施例中，所述意图信息为与意图相关的信息，所述意图信息可以携带意图 标识、意图action和意图object等信息。所述动作信息包括操作、操作可执行条件、意图满足条件以及优化增益和负面影响等信息。
其中，各个动作信息之间的操作不同，各个动作信息之间的操作可执行条件以及意图满足条件可以相同也可以不同。每个动作信息对应一个操作，一个操作可以包括至少一条操作指令，每条操作指令包含动作和动作对象。每个操作可以对应至少一个优化增益和/或至少一个负面影响。每个优化增益包含动作和动作对象，每个负面影响包含动作和动作对象。
第一网元获取意图信息对应的多个动作信息的过程，具体可以包括如图3所示的2011-2019。
2011：第一网元接收来自第二网元的意图信息。
第二网元将携带意图标识、意图action和意图object等信息的意图信息下发给第一网元，以使所述第一网元完成意图转译。
2012：第一网元向第二网元发送意图接收通知。
所述意图接收通知用于通知所述第二网元所述意图信息的接收情况，所述接收情况可以包括接收成功和/或接收失败等。
2013：第一网元提取所述意图信息中意图目标的关键词。
在接收到所述意图信息后，所述第一网元对所述意图信息进行词法、语法及语义分析来提取意图目标的关键词。例如，意图信息为：intent1，提升x区用户视频业务体验；则可以提取出意图目标的关键词为：提升、视频、业务体验。其中，所述词法、语法及语义分析可使用计算机领域词法分析器、语法分析器以及语义分析器来实现。
2014：第一网元向知识库发送查询请求。
其中，所述查询请求携带有所述意图目标的关键词，即2013步骤中提取的关键词。所述查询请求用于获取与所述关键词匹配的目标对应的动作信息。
2015：知识库根据所述查询请求进行目标匹配。
所述知识库接收到所述查询请求后，则根据所述查询请求中的关键词进行目标匹配，以得到与所述目标对应的可选动作信息，其中，所述与所述可选动作信息可以包括至少一个动作信息。其中，每个动作信息均包括操作、操作可执行条件、意图满足条件以及优化增益和负面影响等信息。
所述操作条件包括条件KPI需要满足的条件，其中所述条件KPI是指执行所述操作的前提条件中对应的KPI。例如，所述操作可执行条件可以是：CCE-AvgUtilizationRate<0.7&&DL-Packet-Delay>20ms，其中CCE-AvgUtilizationRate表示控制单元信道元素平均利用率，DL-Packet-Delay表示下行包延时，CCE-AvgUtilizationRate和DL-Packet-Delay均为条件KPI。
所述意图满足条件包括意图监测KPI要满足的条件，所述意图监测KPI是指意图达成需要监测的KPI。例如，所述意图满足条件可以是：视频首播时延<10ms，且视频平均每次播放卡顿次数<2，且视频卡顿时长占比<10％；视频首播时延、视频平均每次播放卡顿次数和视频卡顿时长占比均为监测KPI。
2016：第一网元接收来自知识库的查询结果反馈。
其中，所述查询结果反馈包括步骤2015中得到的与所述目标对应的动作信息。
2017：第一网元向网元发送意图监测KPI、条件KPI订阅请求。
其中，所述意图监测KPI、条件KPI订阅请求用于获取当前网络状态下网元的意图监测KPI和条件KPI。
2018：第一网元接收来自所述网元的意图监测KPI、条件KPI订阅反馈。
其中，所述意图监测KPI、条件KPI订阅反馈携带所述当前网络状态下网元的意图监测KPI和条件KPI。
2019：第一网元确定与当前网络状态匹配的动作信息，并保存所述意图监测KPI。
在具体实现中，在所述第一网元接收到所述意图监测KPI、条件KPI订阅反馈后，则从所述查询结果反馈中携带的动作信息中，确定与当前网络状态下的条件KPI(即所述述意图监测KPI、条件KPI订阅反馈中的条件KPI)匹配的活动信息，从而得到所述多个动作信息。
202：第一网元确定所述多个动作信息的决策信息。
所述决策信息中包括所述多个动作的优化增益和负面影响。
可选的所述决策信息中还可以包括所述多个动作的操作。
203：第一网元根据所述多个动作信息的决策信息确定目标动作信息。
在一种可选的实施方式中，所述第一网元根据所述多个动作信息的决策信息确定目标动作信息可以包括：所述第一网元获取与所述意图信息对应的业务诉求；然后根据所述决策信息以及所述诉求信息确定所述目标动作。
在具体实现中，当所述决策信息包括所述多个动作的优化增益和负面影响时，在获取到所述业务诉求信息后，所述第一网元根据所述业务诉求信息确定所述决策信息中可接受的优化增益和负面影响，以及不可接受的负面影响；然后根据可接受的优化增益和负面影响，以及不可接受的负面影响确定所述目标动作信息。
例如，业务诉求为保障某场馆演唱会，该业务诉求不能接受的负面影响为小区容量降低，其他的负面影响和优化增益都可接受，在确定目标动作信息时，则只需要将多个动作信息中负面影响包括小区容量降低的动作信息排除，剩下的动作信息则为所述目标动作信息。
可选的，当所述决策信息还包括所述多个动作的操作时，则在所述第一网元根据所述业务诉求信息确定所述决策信息中可接受的优化增益和负面影响，以及不可接受的负面影响之前，所述第一网元可以根据所述述业务诉求确定可接受的操作，并确定所述可接受的操作对应的动作信息，若所述可接受的操作对应的动作信息为一个时，则确定所述可接受的操作对应的动作信息为所述目标动作信息；若所述可接受的操作对应的动作信息为多个时，则继续执行所述第一网元根据所述业务诉求信息确定所述决策信息中可接受的优化增益和负面影响，以及不可接受的负面影响的操作。
在另外一种可选的实施方式中，所述第一网元根据所述多个动作信息的决策信息确定目标动作信息可以包括：所述第一网元获取与所述意图信息对应的意图达成的历史经验信息；然后根据所述决策信息以及所述意图达成的历史经验信息确定所述目标动作。
在具体实现中，当所述决策信息包括所述多个动作的优化增益和负面影响时，在获取到所述意图达成的历史经验信息后，所述第一网元根据所述意图达成的历史经验信息确定 所述决策信息中可接受的优化增益和负面影响，以及不可接受的负面影响；然后根据可接受的优化增益和负面影响，以及不可接受的负面影响确定所述目标动作信息。
可选的，当所述决策信息还包括所述多个动作的操作时，则在所述第一网元根据所述意图达成的历史经验信息确定所述决策信息中可接受的优化增益和负面影响，以及不可接受的负面影响之前，所述第一网元可以根据所述述业务诉求确定可接受的操作，并确定所述可接受的操作对应的动作信息，若所述可接受的操作对应的动作信息为一个时，则确定所述可接受的操作对应的动作信息为所述目标动作信息；若所述可接受的操作对应的动作信息为多个时，则继续执行所述第一网元根据所述意图达成的历史经验信息确定所述决策信息中可接受的优化增益和负面影响，以及不可接受的负面影响的操作。
在又一种可选的实施方式中，所述第一网元根据所述多个动作信息的决策信息确定目标动作信息可以包括：所述意图驱动管理服务消费者将所述决策信息输入机器学习模型，并根据所述机器学习模型的输出确定所述目标动作信息。
所述机器学习模型可以是对历史经验的抽象建模，可以选用两层神经网络，输入为所述决策信息中的优化增益，负面影响和操作等信息，输出为可接受的优化增益、可接受的负面影响或需要执行的操作。
204：第一网元向网元发送操作执行请求。
其中，所述操作执行请求，携带所述目标动作信息的操作，即所述网元需要执行的操作，例如，该操作可以是MOD CELLULSCHALGO.UlEnhencedVoipSchSw:LocalCellId＝％cell_id；UlVoipSchOptSwitch＝ON。
在本申请实施例中，图3所示的步骤305-309的实现过程和图2所示的步骤105-109类似，因此不再赘述。
在本申请实施例中，当意图转译过程中存在多个方案(即多个动作信息)可以同时满足同一意图时，第一网元通过将多个动作的优化增益和负面影响作为决策信息，然后通过意图的业务诉求或所述意图的意图达成的历史经验信息所述决策信息中的目标决策信息(即可接受的优化增益和负面影响以及不可接受的负面影响)，或者通过机器学习模型来确定所述决策信息中的目标决策信息，进一步根据所述目标决策信息确定所述多个动作信息中的目标动作信息。可以看出，相比于已有的随机选择动作信息或逐个尝试动作信息的方案，本申请实施例通过增加动作信息的选择流程来确定目标动作信息的方案，可以使的第一网元可以快速的确定目标动作信息，以及动作信息选择的准确性，进一步提升了第一网元意图转译的效率。
请参阅图4，图4是本申请实施例提供的又一种信息处理方法的流程示意图，如图所示，本信息处理方法包括：
301：第一网元获取意图信息对应的多个动作信息。
在本申请实施例中，所述第一网元获取意图信息对应的多个动作信息的过程和图2所示的步骤101相同，此处不再赘述。
302：第一网元确定所述多个动作信息的决策信息。
所述决策信息中包括所述多个动作的操作以及所述与所述操作对应的操作标识。
其中，所述操作标识为操作和动作信息之间的关联信息，即可以通过操作的操作标识确定与该操作对应的动作信息。
303：第一网元根据所述多个动作信息的决策信息确定目标动作信息。
在具体实现中，所述第一网元根据所述多个动作信息的决策信息确定目标动作信息过程，具体可以包括如图4所示的3031-3033。
3031：第一网元向第二网元发送动作选择请求。
其中，所述动作选择请求携带所述多个动作的决策信息。
3032a：第二网元根据所述动作选择请求确定目标决策信息。
在一种可选的实施方式中，所述第二网元根据所述动作选择请求确定目标决策信息可以包括：所述第二网元获取与所述意图信息对应的业务诉求；然后根据所述决策信息以及所述诉求信息确定所述目标决策信息。
在具体实现中，在获取到所述业务诉求信息后，所述第二网元根据所述业务诉求信息确定所述决策信息中需要执行的操作；然后根据需要执行的操作确定所述目标决策信息。
在另外一种可选的实施方式中，所述第二网元根据所述动作选择请求确定目标决策信息可以包括：所述第二网元获取与所述意图信息对应的意图达成的历史经验信息；然后根据所述决策信息以及所述意图达成的历史经验信息确定所述目标决策信息。
在具体实现中，在获取到所述意图达成的历史经验信息后，所述第二网元根据所述意图达成的历史经验信息确定所述决策信息中需要执行的操作；然后根据需要执行的操作确定所述目标决策信息。
在又一种可选的实施方式中，所述第二网元根据所述动作选择请求确定目标决策信息可以包括：所述意图驱动管理服务消费者将所述决策信息输入机器学习模型，并根据所述机器学习模型的输出确定所述目标决策信息。
具体的，所述机器学习模型可以是对历史经验的抽象建模，可以选用两层神经网络，输入为所述决策信息中的操作等信息，输出为可接受的或需要执行的操作，或者为可接受的或需要执行的操作标识。
可以理解的是步骤3032a执行的前提条件为：所述第二网元可以获取到所述业务诉求或者所述意图达成的历史经验信息，或者所述第二网元可以调用所述机器学习模型。若不满足所述前提条件，则执行步骤3032b。
3032b：第二网元根据所述动作选择请求确定目标决策信息。
具体的，所述第二网元根据所述动作选择请求确定目标决策信息的过程，具体可以包括如图4所示的3032b1-3032b3。
3032b1：第二网元向第三网元发送所述动作选择请求。
3032b2：第三网元根据所述动作选择请求确定目标决策信息。
在一种可选的实施方式中，所述第三网元根据所述动作选择请求确定目标决策信息可以包括：所述第三网元获取与所述意图信息对应的业务诉求；然后根据所述决策信息以及所述诉求信息确定所述目标决策信息。
在具体实现中，在获取到所述业务诉求信息后，所述第三网元根据所述业务诉求信息确定所述决策信息中需要执行的操作；然后根据需要执行的操作确定所述目标决策信息。
在另外一种可选的实施方式中，所述第三网元根据所述动作选择请求确定目标决策信息可以包括：所述第三网元获取与所述意图信息对应的意图达成的历史经验信息；然后根据所述决策信息以及所述意图达成的历史经验信息确定所述目标决策信息。
在具体实现中，在获取到所述意图达成的历史经验信息后，所述第三网元根据所述意图达成的历史经验信息确定所述决策信息中需要执行的操作；然后根据需要执行的操作确定所述目标决策信息。
在又一种可选的实施方式中，所述第三网元根据所述动作选择请求确定目标决策信息可以包括：所述意图驱动管理服务消费者将所述决策信息输入机器学习模型，并根据所述机器学习模型的输出确定所述目标决策信息。
3032b3：所述第三网元向所述第二网元发送动作选择响应。
所述第三网元根据所述目标决策信息生成所述动作选择响应，并向所述第二网元发送动作选择响应。所述动作选择响应用于所述第一网元根据所述动作选择响应中携带的信息确定目标动作。
所述动作选择响应携带所述目标决策信息，或者携带所述目标决策信息的标识信息，所述标识信息包括所述目标决策信息中的操作标识。
3033：所述第二网元向所述第一网元发送动作选择响应。
在一种可选的实施方式中，所述第二网元接收来自所述第三网元的动作选择响应，然后向所述第一网元发送动作选择响应。
在另一种可选的实施方式中，所述第二网元根据所述目标决策信息生成所述动作选择响应，并向所述第一网元发送动作选择响应。所述动作选择响应携带所述目标决策信息，或者携带所述目标决策信息的标识信息，所述标识信息包括所述目标决策信息中的操作标识。所述动作选择响应用于所述第一网元根据所述动作选择响应中携带的信息确定目标动作。
304：第一网元向网元发送操作执行请求。
在接收到所述动作选择响应后，所述第一网元根据所述动作选择响应中携带的信息确定目标动作。
其中，所述操作执行请求，携带所述目标动作信息的操作，即所述网元需要执行的操作，例如，该操作可以是MOD CELLULSCHALGO.UlEnhencedVoipSchSw:LocalCellId＝％cell_id；UlVoipSchOptSwitch＝ON。
在本申请实施例中，图4所示的步骤305-309的实现过程和图2所示的步骤105-109类似，因此不再赘述。
在本申请实施例中，当意图转译过程中存在多个方案(即多个动作信息)可以同时满足同一意图时，第一网元通过向第二网元发送携带多个动作的操作(即决策信息)的动作选择请求。第二网元通过意图的业务诉求或所述意图的意图达成的历史经验信息所述决策信息中的目标决策信息(即目标操作)，或者通过机器学习模型来确定所述决策信息中的目标决策信息。若第二网元不能确定目标决策信息，则将所述动作选择请求发送给第三网元，以使第三网元来确定所述目标决策信息(过程和第二网元确定目标决策信息相同)，并接收携带目标决策信息或目标决策信息的标识信息的动作选择响应。然后第二网元向第一网元 发送携带目标决策信息或目标决策信息的标识信息的动作选择响应，以使第一网元根据所述目标决策信息或所述信息标识确定所述多个动作信息中的目标动作信息。可以看出，相比于已有的随机选择动作信息或逐个尝试动作信息的方案，本申请实施例通过增加动作信息的选择流程来确定目标动作信息的方案，可以使的第一网元可以快速的确定目标动作信息，以及动作信息选择的准确性，进一步提升了第一网元意图转译的效率。
请参阅图5，图5是本申请实施例提供了一种信息处理方法的流程示意图，如图所示，本信息处理方法包括：
401：第一网元获取意图信息对应的多个动作信息。
在本申请实施例中，所述第一网元获取意图信息对应的多个动作信息的过程和图3所示的步骤201相同，此处不再赘述。
402：第一网元确定所述多个动作信息的决策信息。
所述决策信息中包括所述多个动作的优化增益以及与所述优化增益对应的优化增益标识、负面影响以及与所述负面影响对应的负面影响标识。
其中，所述优化增益标识为优化增益和动作信息之间的关联信息，即可以通过优化增益的优化增益标识确定与优化增益作对应的动作信息。
所述负面影响标识为负面影响和动作信息之间的关联信息，即可以通过负面影响的负面影响标识确定与负面影响作对应的动作信息。
可选的所述决策信息中包括所述多个动作的操作以及所述与所述操作对应的操作标识。
其中，所述操作标识为操作和动作信息之间的关联信息，即可以通过操作的操作标识确定与该操作对应的动作信息。
403：第一网元根据所述多个动作信息的决策信息确定目标动作信息。
在具体实现中，所述第一网元根据所述多个动作信息的决策信息确定目标动作信息过程，具体可以包括如图5所示的4031-4033。
4031：第一网元向第二网元发送动作选择请求。
其中，所述动作选择请求携带所述多个动作的决策信息。
4032a：第二网元根据所述动作选择请求确定目标决策信息。
在一种可选的实施方式中，所述第二网元根据所述动作选择请求确定目标决策信息可以包括：所述第二网元获取与所述意图信息对应的业务诉求；然后根据所述决策信息以及所述诉求信息确定所述目标决策信息。
在具体实现中，在获取到所述业务诉求信息后，在获取到所述业务诉求信息后，所述第二网元根据所述业务诉求信息确定所述决策信息中可接受的优化增益和负面影响，以及不可接受的负面影响；然后根据可接受的优化增益和负面影响，以及不可接受的负面影响确定所述目标决策信息。
例如，业务诉求为保障某场馆演唱会，该业务诉求不能接受的负面影响为小区容量降低，其他的负面影响和优化增益都可接受，在确定目标动作信息时，则只需要将多个动作信息中负面影响包括小区容量降低的动作信息排除，剩下的动作信息则为所述目标动作信息。
可选的，当所述决策信息还包括所述多个动作的操作时，则在所述第二网元根据所述 业务诉求信息确定所述决策信息中可接受的优化增益和负面影响，以及不可接受的负面影响之前，所述第二网元可以根据所述述业务诉求确定可接受的操作，若所述可接受的操作为一个时，则确定所述可接受的操作为所述目标决策信息；若所述可接受的操作为多个时，则继续执行所述第二网元根据所述业务诉求信息确定所述决策信息中可接受的优化增益和负面影响的操作。
在另外一种可选的实施方式中，所述第二网元根据所述动作选择请求确定目标决策信息可以包括：所述第二网元获取与所述意图信息对应的意图达成的历史经验信息；然后根据所述决策信息以及所述意图达成的历史经验信息确定所述目标决策信息。
在具体实现中，在获取到所述意图达成的历史经验信息后，所述第二网元根据所述意图达成的历史经验信息确定所述决策信息中可接受的优化增益和负面影响，以及不可接受的负面影响；然后根据可接受的优化增益和负面影响，以及不可接受的负面影响确定所述目标决策信息。
可选的，当所述决策信息还包括所述多个动作的操作时，则在所述第二网元根据所述意图达成的历史经验信息确定所述决策信息中可接受的优化增益和负面影响，以及不可接受的负面影响之前，所述第二网元可以根据所述述业务诉求确定可接受的操作，若所述可接受的操作为一个时，则确定所述可接受的操作为所述目标决策信息；若所述可接受的操作为多个时，则继续执行所述第二网元根据所述意图达成的历史经验信息确定所述决策信息中可接受的优化增益和负面影响，以及不可接受的负面影响的操作。
在又一种可选的实施方式中，所述第二网元根据所述动作选择请求确定目标决策信息可以包括：所述意图驱动管理服务消费者将所述决策信息输入机器学习模型，并根据所述机器学习模型的输出确定所述目标决策信息。
具体的，所述机器学习模型可以是对历史经验的抽象建模，可以选用两层神经网络，输入为所述决策信息中的优化增益，负面影响和操作等信息，输出为可接受的优化增益、可接受的负面影响或需要执行的操作。
可以理解的是步骤4032a执行的前提条件为：所述第二网元可以获取到所述业务诉求或者所述意图达成的历史经验信息，或者所述第二网元可以调用所述机器学习模型。若不满足所述前提条件，则执行步骤4032b。
4032b：第二网元根据所述动作选择请求确定目标决策信息。
具体的，所述第二网元根据所述动作选择请求确定目标决策信息的过程，具体可以包括如图4所示的3032b1-3032b3。
4032b1：第二网元向第三网元发送所述动作选择请求。
4032b2：第三网元根据所述动作选择请求确定目标决策信息。
在一种可选的实施方式中，所述第三网元根据所述动作选择请求确定目标决策信息可以包括：所述第三网元获取与所述意图信息对应的业务诉求；然后根据所述决策信息以及所述诉求信息确定所述目标决策信息。
在具体实现中，在获取到所述业务诉求信息后，所述第三网元根据所述业务诉求信息确定所述决策信息中可接受的优化增益和负面影响，以及不可接受的负面影响；然后根据可接受的优化增益和负面影响，以及不可接受的负面影响确定所述目标决策信息。
例如，业务诉求为保障某场馆演唱会，该业务诉求不能接受的负面影响为小区容量降低，其他的负面影响和优化增益都可接受，在确定目标动作信息时，则只需要将多个动作信息中负面影响包括小区容量降低的动作信息排除，剩下的动作信息则为所述目标动作信息。
可选的，当所述决策信息还包括所述多个动作的操作时，则在所述第三网元根据所述业务诉求信息确定所述决策信息中可接受的优化增益和负面影响，以及不可接受的负面影响之前，所述第三网元可以根据所述述业务诉求确定可接受的操作，若所述可接受的操作为一个时，则确定所述可接受的操作为所述目标决策信息；若所述可接受的操作为多个时，则继续执行所述第三网元根据所述业务诉求信息确定所述决策信息中可接受的优化增益和负面影响的操作。
在另外一种可选的实施方式中，所述第三网元根据所述动作选择请求确定目标决策信息可以包括：所述第三网元获取与所述意图信息对应的意图达成的历史经验信息；然后根据所述决策信息以及所述意图达成的历史经验信息确定所述目标决策信息。
在具体实现中，在获取到所述意图达成的历史经验信息后，所述第三网元根据所述意图达成的历史经验信息确定所述决策信息中可接受的优化增益和负面影响，以及不可接受的负面影响；然后根据可接受的优化增益和负面影响，以及不可接受的负面影响确定所述目标决策信息。
可选的，当所述决策信息还包括所述多个动作的操作时，则在所述第三网元根据所述意图达成的历史经验信息确定所述决策信息中可接受的优化增益和负面影响，以及不可接受的负面影响之前，所述第三网元可以根据所述述业务诉求确定可接受的操作，若所述可接受的操作为一个时，则确定所述可接受的操作为所述目标决策信息；若所述可接受的操作为多个时，则继续执行所述第三网元根据所述意图达成的历史经验信息确定所述决策信息中可接受的优化增益和负面影响，以及不可接受的负面影响的操作。
在又一种可选的实施方式中，所述第三网元根据所述动作选择请求确定目标决策信息可以包括：所述意图驱动管理服务消费者将所述决策信息输入机器学习模型，并根据所述机器学习模型的输出确定所述目标决策信息。
具体的，所述机器学习模型可以是对历史经验的抽象建模，可以选用两层神经网络，输入为所述决策信息中的优化增益，负面影响和操作等信息，输出为可接受的优化增益、可接受的负面影响或需要执行的操作。
4032b3：所述第三网元向所述第二网元发送动作选择响应。
具体的，所述第三网元根据所述决策信息生成所述动作选择响应，并向所述第二网元发送动作选择响应，所述动作选择响应携带所述目标决策信息，或者携带所述决策信息的标识信息。
其中，若所述决策信息包括操作，标识信息包括所述操作的操作标识；若所述决策信息包括可接受的优化增益和/或负面影响，以及包括不可接受的优化增益和/或负面影响，则所述表示信息包括所述可接受的优化增益标识和/或负面影响标识，以及不可接受的优化增益标识和/或负面影响标识。
4033：所述第二网元向所述第一网元发送动作选择响应。
在一种可选的实施方式中，所述第二网元接收来自所述第三网元的动作选择响应，然后向所述第一网元发送动作选择响应。
在另一种可选的实施方式中，所述第二网元根据所述目标决策信息生成所述动作选择响应，并向所述第一网元发送动作选择响应。
在具体实现中，所述第二网元根据所述目标决策信息生成所述动作选择响应，可以包括：所述第二网元根据所述决策信息生成所述动作选择响应，所述动作选择响应携带所述目标决策信息，或携带所述决策信息的标识信息。
其中，若所述决策信息包括操作，标识信息包括所述操作的操作标识；若所述决策信息包括可接受的优化增益和/或负面影响，以及包括不可接受的优化增益和/或负面影响，则所述表示信息包括所述可接受的优化增益标识和/或负面影响标识，以及不可接受的优化增益标识和/或负面影响标识。
404：第一网元向网元发送操作执行请求。
在接收到所述动作选择响应后，所述第一网元根据所述动作选择响应中携带的信息确定目标动作。
其中，所述操作执行请求，携带所述目标动作信息的操作，即所述网元需要执行的操作，例如，该操作可以是MOD CELLULSCHALGO.UlEnhencedVoipSchSw:LocalCellId＝％cell_id；UlVoipSchOptSwitch＝ON。
在本申请实施例中，图5所示的步骤405-409的实现过程和图2所示的步骤105-109类似，因此不再赘述。
在本申请实施例中，当意图转译过程中存在多个方案(即多个动作信息)可以同时满足同一意图时，第一网元通过向第二网元发送携带多个动作的优化增益和负面影响(即决策信息)的动作选择请求。第二网元通过意图的业务诉求或所述意图的意图达成的历史经验信息所述决策信息中的目标决策信息(即可接受的优化增益和负面影响以及不可接受的负面影响)，或者通过机器学习模型来确定所述决策信息中的目标决策信息。若第二网元不能确定目标决策信息，则将所述动作选择请求发送给第三网元，以使第三网元来确定所述目标决策信息(过程和第二网元确定目标决策信息相同)，并接收携带目标决策信息或目标决策信息的标识信息的动作选择响应。然后第二网元向第一网元发送携带目标决策信息或目标决策信息的标识信息的动作选择响应，以使第一网元根据所述目标决策信息或所述信息标识确定所述多个动作信息中的目标动作信息。可以看出，相比于已有的随机选择动作信息或逐个尝试动作信息的方案，本申请实施例通过增加动作信息的选择流程来确定目标动作信息的方案，可以使的第一网元可以快速的确定目标动作信息，以及动作信息选择的准确性，进一步提升了第一网元意图转译的效率。
请参阅图6，图6是本申请实施例提供了一种信息处理方法的流程示意图，如图所示，本信息处理方法包括：
501：第一网元获取意图信息对应的多个动作信息。
在本申请实施例中，所述第一网元获取意图信息对应的多个动作信息的过程和图2所示的步骤101相同，此处不再赘述。
502：第一网元确定所述多个动作信息的决策信息。
所述决策信息中包括所述多个动作的操作以及所述与所述操作对应的操作标识。
其中，所述操作标识为操作和动作信息之间的关联信息，即可以通过操作的操作标识确定与该操作对应的动作信息。
503：第一网元根据所述多个动作信息的决策信息确定目标动作信息。
在具体实现中，所述第一网元根据所述多个动作信息的决策信息确定目标动作信息过程，具体可以包括如图6所示的3031-3033。
5031：第一网元向第三网元发送动作选择请求。
其中，所述动作选择请求携带所述多个动作的决策信息。
5032：第三网元根据所述动作选择请求确定目标决策信息。
在一种可选的实施方式中，所述第三网元根据所述动作选择请求确定目标决策信息可以包括：所述第三网元获取与所述意图信息对应的业务诉求；然后根据所述决策信息以及所述诉求信息确定所述目标决策信息。
在具体实现中，在获取到所述业务诉求信息后，所述第三网元根据所述业务诉求信息确定所述决策信息中需要执行的操作；然后根据需要执行的操作确定所述目标决策信息。
在另外一种可选的实施方式中，所述第三网元根据所述动作选择请求确定目标决策信息可以包括：所述第三网元获取与所述意图信息对应的意图达成的历史经验信息；然后根据所述决策信息以及所述意图达成的历史经验信息确定所述目标决策信息。
在具体实现中，在获取到所述意图达成的历史经验信息后，所述第三网元根据所述意图达成的历史经验信息确定所述决策信息中需要执行的操作；然后根据需要执行的操作确定所述目标决策信息。
在又一种可选的实施方式中，所述第三网元根据所述动作选择请求确定目标决策信息可以包括：所述意图驱动管理服务消费者将所述决策信息输入机器学习模型，并根据所述机器学习模型的输出确定所述目标决策信息。
5033：所述第三网元向所述第一网元发送动作选择响应。
所述第三网元根据所述目标决策信息生成所述动作选择响应，并向所述第一网元发送动作选择响应。所述动作选择响应用于所述第一网元根据所述动作选择响应中携带的信息确定目标动作。
所述动作选择响应携带所述目标决策信息，或者携带所述目标决策信息的标识信息，所述标识信息包括所述目标决策信息中的操作标识。
504：第一网元向网元发送操作执行请求。
在接收到所述动作选择响应后，所述第一网元根据所述动作选择响应中的携带的信息确定所述目标动作。
其中，所述操作执行请求，携带所述目标动作信息的操作，即所述网元需要执行的操作，例如，该操作可以是MOD CELLULSCHALGO.UlEnhencedVoipSchSw:LocalCellId＝％cell_id；UlVoipSchOptSwitch＝ON。
在本申请实施例中，图6所示的步骤605-609的实现过程和图2所示的步骤105-109类似，因此不再赘述。
在本申请实施例中，当意图转译过程中存在多个方案(即多个动作信息)可以同时满足同一意图时，第一网元通过向第三网元发送携带多个动作的操作(即决策信息)的动作选择请求。第三网元通过意图的业务诉求或所述意图的意图达成的历史经验信息所述决策信息中的目标决策信息(即目标操作)，或者通过机器学习模型来确定所述决策信息中的目标决策信息。然后第三网元向第一网元发送携带目标决策信息或目标决策信息的标识信息的动作选择响应，以使第一网元根据所述目标决策信息或所述信息标识确定所述多个动作信息中的目标动作信息。可以看出，相比于已有的随机选择动作信息或逐个尝试动作信息的方案，本申请实施例通过增加动作信息的选择流程来确定目标动作信息的方案，可以使的第一网元可以快速的确定目标动作信息，以及动作信息选择的准确性，进一步提升了第一网元意图转译的效率。
请参阅图7，图7是本申请实施例提供了一种信息处理方法的流程示意图，如图所示，本信息处理方法包括：
601：第一网元获取意图信息对应的多个动作信息。
在本申请实施例中，所述第一网元获取意图信息对应的多个动作信息的过程和图3所示的步骤201相同，此处不再赘述。
602：第一网元确定所述多个动作信息的决策信息。
所述决策信息中包括所述多个动作的优化增益以及与所述优化增益对应的优化增益标识、负面影响以及与所述负面影响对应的负面影响标识。
其中，所述优化增益标识为优化增益和动作信息之间的关联信息，即可以通过优化增益的优化增益标识确定与优化增益作对应的动作信息。
所述负面影响标识为负面影响和动作信息之间的关联信息，即可以通过负面影响的负面影响标识确定与负面影响作对应的动作信息。
可选的所述决策信息中包括所述多个动作的操作以及所述与所述操作对应的操作标识。
其中，所述操作标识为操作和动作信息之间的关联信息，即可以通过操作的操作标识确定与该操作对应的动作信息。
603：第一网元根据所述多个动作信息的决策信息确定目标动作信息。
在具体实现中，所述第一网元根据所述多个动作信息的决策信息确定目标动作信息过程，具体可以包括如图7所示的6031-6033。
6031：第一网元向第三网元发送动作选择请求。
其中，所述动作选择请求携带所述多个动作的决策信息。
6032：第三网元根据所述动作选择请求确定目标决策信息。
在一种可选的实施方式中，所述第三网元根据所述动作选择请求确定目标决策信息可以包括：所述第三网元获取与所述意图信息对应的业务诉求；然后根据所述决策信息以及所述诉求信息确定所述目标决策信息。
在具体实现中，在获取到所述业务诉求信息后，在获取到所述业务诉求信息后，所述第三网元根据所述业务诉求信息确定所述决策信息中可接受的优化增益和负面影响，以及不可接受的负面影响；然后根据可接受的优化增益和负面影响，以及不可接受的负面影响确定所述目标决策信息。
例如，业务诉求为保障某场馆演唱会，该业务诉求不能接受的负面影响为小区容量降低，其他的负面影响和优化增益都可接受，在确定目标动作信息时，则只需要将多个动作信息中负面影响包括小区容量降低的动作信息排除，剩下的动作信息则为所述目标动作信息。
可选的，当所述决策信息还包括所述多个动作的操作时，则在所述第三网元根据所述业务诉求信息确定所述决策信息中可接受的优化增益和负面影响，以及不可接受的负面影响之前，所述第三网元可以根据所述述业务诉求确定可接受的操作，若所述可接受的操作为一个时，则确定所述可接受的操作为所述目标决策信息；若所述可接受的操作为多个时，则继续执行所述第三网元根据所述业务诉求信息确定所述决策信息中可接受的优化增益和负面影响的操作。
在另外一种可选的实施方式中，所述第三网元根据所述动作选择请求确定目标决策信息可以包括：所述第三网元获取与所述意图信息对应的意图达成的历史经验信息；然后根据所述决策信息以及所述意图达成的历史经验信息确定所述目标决策信息。
在具体实现中，在获取到所述意图达成的历史经验信息后，所述第三网元根据所述意图达成的历史经验信息确定所述决策信息中可接受的优化增益和负面影响，以及不可接受的负面影响；然后根据可接受的优化增益和负面影响，以及不可接受的负面影响确定所述目标决策信息。
可选的，当所述决策信息还包括所述多个动作的操作时，则在所述第三网元根据所述意图达成的历史经验信息确定所述决策信息中可接受的优化增益和负面影响，以及不可接受的负面影响之前，所述第三网元可以根据所述述业务诉求确定可接受的操作，若所述可接受的操作为一个时，则确定所述可接受的操作为所述目标决策信息；若所述可接受的操作为多个时，则继续执行所述第三网元根据所述意图达成的历史经验信息确定所述决策信息中可接受的优化增益和负面影响，以及不可接受的负面影响的操作。
在又一种可选的实施方式中，所述第三网元根据所述动作选择请求确定目标决策信息可以包括：所述意图驱动管理服务消费者将所述决策信息输入机器学习模型，并根据所述机器学习模型的输出确定所述目标决策信息。
具体的，所述机器学习模型可以是对历史经验的抽象建模，可以选用两层神经网络，输入为所述决策信息中的优化增益，负面影响和操作等信息，输出为可接受的优化增益、可接受的负面影响或需要执行的操作。
6033：所述第三网元向所述第一网元发送动作选择响应。
在一种可选的实施方式中，所述第三网元根据所述决策信息生成所述动作选择响应，并向所述第一网元发送动作选择响应，所述动作选择响应携带所述目标决策信息，或者携带所述目标决策信息的标识信息。
其中，若所述决策信息包括操作，标识信息包括所述操作的操作标识；若所述决策信息包括可接受的优化增益和/或负面影响，以及包括不可接受的优化增益和/或负面影响，则所述表示信息包括所述可接受的优化增益标识和/或负面影响标识，以及不可接受的优化增益标识和/或负面影响标识。
604：第一网元向网元发送操作执行请求。
在接收到所述动作选择响应后，所述第一网元根据所述动作选择响应中携带的信息确定目标动作。
其中，所述操作执行请求，携带所述目标动作信息的操作，即所述网元需要执行的操作，例如，该操作可以是MOD CELLULSCHALGO.UlEnhencedVoipSchSw:LocalCellId＝％cell_id；UlVoipSchOptSwitch＝ON。
在本申请实施例中，图7所示的步骤605-609的实现过程和图2所示的步骤105-109类似，因此不再赘述。
在本申请实施例中，当意图转译过程中存在多个方案(即多个动作信息)可以同时满足同一意图时，第一网元通过向第三网元发送携带多个动作的优化增益和负面影响(即决策信息)的动作选择请求。第三网元通过意图的业务诉求或所述意图的意图达成的历史经验信息所述决策信息中的目标决策信息(即可接受的优化增益和负面影响以及不可接受的负面影响)，或者通过机器学习模型来确定所述决策信息中的目标决策信息。然后第三网元向第一网元发送携带目标决策信息或目标决策信息的标识信息的动作选择响应，以使第一网元根据所述目标决策信息或所述信息标识确定所述多个动作信息中的目标动作信息。可以看出，相比于已有的随机选择动作信息或逐个尝试动作信息的方案，本申请实施例通过增加动作信息的选择流程来确定目标动作信息的方案，可以使的第一网元可以快速的确定目标动作信息，以及动作信息选择的准确性，进一步提升了第一网元意图转译的效率。
如图8所示，为本申请实施例提供的IDM系统分层架构示意图。其中，通信服务消费者可以包括最终用户、中小型企业、大型企业、垂直行业以及其他通信服务提供者等。IDMS消费者主要下发意图，IDMS提供者主要完成意图转译，将意图转换成网络管理任务或策略。根据IDMS用户的不同，可将意图分为通信服务消费者意图、通信服务提供者意图和网络运营商意图三类。如图8所示，每个上下层之间可以互为一对IDMS消费者和IDMS提供者，上层为IDMS消费者，下层为IDMS提供者。
在本申请实施例中，IDMS消费者，主要下发意图，同时新增辅助转译功能，其可部署在业务支撑系统(Business Support System,BSS)、网络管理系统(Network Management System，NMS)、网元管理系统(Element Management System,EMS)、RAN集群控制器(Cluster RAN Controller,CLRC)上。
IDMS的提供者，主要完成意图转译及意图执行和位置，其可部署在NMS、EMS、CLRC、网元(Network Element,NE)上。
知识库主要存储网络运维知识，其可部署在BSS、NMS、EMS，也独立部署在服务器或运营商数据中心；
网元可以是4G LTE制式下的网元eNodeB，5G的网元CU/DU/gNodeB。
逻辑网元辅助转译系统主要完成辅助转译，可部署在BSS、NMS、EMS、CLRC，也独立部署在服务器或运营商数据中心。
上述详细阐述了本发明实施例的方法，下面为了便于更好地实施本发明实施例的上述方案，相应地，下面还提供用于配合实施上述方案的相关装置。
上述本申请提供的实施例中，分别从各个网元本身、以及从各个网元之间交互的角度对本申请实施例提供的通信方法的各方案进行了介绍。可以理解的是，各个网元，例如上 述第一网元、第二网元、第三网元为了实现上述功能，其包含了执行各个功能相应的硬件结构和/或软件模块。本领域技术人员应该很容易意识到，结合本文中所公开的实施例描述的各示例的单元及算法步骤，本申请能够以硬件或硬件和计算机软件的结合形式来实现。某个功能究竟以硬件还是计算机软件驱动硬件的方式来执行，取决于技术方案的特定应用和设计约束条件。专业技术人员可以对每个特定的应用来使用不同方法来实现所描述的功能，但是这种实现不应认为超出本申请的范围。
请参见图9，图9为本申请实施例提供的一种信息处理装置的结构示意图，该信息处理装置可用于执行上述图2-图7中第一网元的操作。如图9所示，所述信息处理装置900至少包括：获取模块910，第一确定模块920以及第二确定模块930；其中，
获取模块910，用于获取意图的多个动作信息；
第一确定模块920，用于确定所述多个动作信息的决策信息，所述决策信息为所述多个动作信息中包含的信息；
第二确定模块930，用于根据所述多个动作信息的决策信息确定目标动作信息，所述目标动作信息用于实现所述意图。
本申请实施例中所述信息处理装置通过获取意图的多个动作信息，然后从多个信息中提取出决策信息，并根据得到的决策信息来确定所述多个动作信息中的目标动作信息。可见，所述信息处理装置通过增加辅助选择功能以及流程来快速确定实现意图的目标动作信息，从而提升意图转译的效率。
在一种可能的实施方式中，所述第二确定模块930，具体用于获取所述意图的业务诉求或所述意图的意图达成的历史经验信息，所述业务诉求是指与所述意图对应的业务要求，所述历史经验信息为与当前场景相同的意图转译事件的信息，所述场景相同是指要实现的意图以及网络状态相同；根据所述意图的业务诉求或所述意图达成的历史经验信息以及所述决策信息确定所述目标动作信息。
在一种可能的实施方式中，所述第二确定模块930，具体用于将所述多个动作的决策信息输入机器学习模型，得到所述目标动作信息。
在一种可能的实施方式中，所述第二确定模块930，具体用于向第二网元或第三网元发送动作选择请求，所述动作选择请求携带所述多个动作信息的决策信息，所述动作选择请求用于第二网元或第三网元根据所述多个动作信息的决策信息确定目标决策信息；接收来自所述第二网元或所述第三网元的动作选择响应，所述动作选择响应携带所述目标决策信息或所述目标决策信息的标识信息；根据所述目标决策信息或所述标识信息确定所述目标动作信息。
在一种可能的实施方式中，所述第一确定模块920，具体用于根据所述多个动作信息的优化增益和负面影响确定所多个动作的决策信息，所述优化增益包含动作和动作对象，所述负面影响包含动作和动作对象；或者根据所述多个动作信息的操作确定所多个动作的决策信息，所述操作信息包含动作和动作对象。
在一种可能的实施方式中，所述第一确定模块920，具体用于根据所述多个动作信息的优化增益和负面影响以及操作确定所多个动作的决策信息。
请参见图10，图10为本申请实施例提供的一种信息处理装置的结构示意图，该信息 处理装置可用于执行上述图2-图7中第二网元的操作。如图10所示，所述信息处理装置1000至少包括：接收模块1010、确定模块1020以及发送模块1030，其中，
所述接收模块1010，用于接收来自第一网元的动作选择请求，所述动作选择请求包括多个动作信息的决策信息，所述多个动作信息为所述第一网元根据意图获取的，所述决策信息为所述多个动作信息中包含的信息；
所述确定模块1020，用于根据多个动作信息的决策信息确定目标决策信息，所述目标决策信息用于确定所述目标动作；
所述发送模块1030，用于向所述第一网元发送动作选择响应，所述动作选择响应携带所述目标决策信息，或者携带所述目标决策信息的标识信息。
本申请实施例中所述信息处理装置通接收来自第一网元的动作选择请求，其中所动作选择请求携带决策信息；然后确定出所述决策信息中的目标决策信息，最后向所述第一网元发送携带目标决策信息生成动作选择响应，以使所述第一网元根据所述目标决策信息确定目标动作信息。可见，所述信息处理装置通过增加辅助选择功能以及流程来帮助所述第一网元确定实现意图的目标动作信息，从而提升意图转译的效率。
在一种可能的实施方式中，所述确定模块1020，具体用于在所述第二网元不能处理所述动作选择请求的情况下，所述第二网元向第三网元发送所述动作选择请求；接收来自所述第三网元的所述动作选择响应。
在一种可能的实施方式中，所述确定模块1020，具体用于获取所述意图的业务诉求或所述意图的意图达成的历史经验信息，所述业务诉求是指与所述意图对应的业务要求，所述历史经验信息为与当前场景相同的意图转译事件的信息，所述场景相同是指要实现的意图以及网络状态相同；根据所述多个动作的决策信息以及所述诉求信息或所述历史经验信息确定所述目标决策信息。
在一种可能的实施方式中，所述确定模块1020，具体用于所述第二网元将所述多个动作的决策信息输入机器学习模型，得到所述目标决策信息。
在一种可能的实施方式中，所述多个动作信息的决策信息包括所述多个动作信息的操作，或者所述多个动作信息的决策信息包括所述多个动作信息的优化增益和负面影响，或者所述多个动作信息的决策信息包括所述多个动作信息的操作、所述多个动作信息的优化增益和负面影响，所述优化增益包含动作和动作对象，所述负面影响包含动作和动作对象，所述操作信息包含动作和动作对象。
请参见图11，图11为本申请实施例提供的一种信息处理装置的结构示意图，该信息处理装置可用于执行上述图2-图7中第三网元的操作。如图11所示，所述信息处理装置1100至少包括：接收模块1110、确定模块1120以及发送模块1130，其中，
所述接收模块1110，用于接收来自第一网元或第二网元的动作选择请求，所述动作选择请求，所述动作选择请求包括多个动作信息的决策信息，所述多个动作信息为所述第一网元根据意图获取的，所述决策信息为所述多个动作信息中包含的信息；
所述确定模块1120，用于根据多个动作信息的决策信息确定目标决策信息，所述目标决策信息用于确定所述目标动作；
所述发送模块1130，用于向所述第一网元或第二网元发送动作选择响应，所述动作选 择响应携带所述目标决策信息，或者携带所述目标决策信息的标识信息。
本申请实施例中所述信息处理装置通接收来自第一网元或第二网元的动作选择请求，其中所动作选择请求携带决策信息；然后确定出所述决策信息中的目标决策信息，最后向所述第一网元或第二网元发送携带目标决策信息生成动作选择响应，以使所述第一网元根据所述目标决策信息确定目标动作信息。可见，所述信息处理装置通过增加辅助选择功能以及流程来帮助所述第一网元确定实现意图的目标动作信息，从而提升意图转译的效率。
在一种可能的实施方式中，所述确定模块1120，具体用于获取所述意图的业务诉求或所述意图的意图达成的历史经验信息，所述业务诉求是指与所述意图对应的业务要求，所述历史经验信息为与当前场景相同的意图转译事件的信息，所述场景相同是指要实现的意图以及网络状态相同；根据所述多个动作的决策信息以及所述诉求信息或所述历史经验信息确定所述目标决策信息。
在一种可能的实施方式中，所述确定模块1120，具体用于所述第二网元将所述多个动作的决策信息输入机器学习模型，得到所述目标决策信息。
在一种可能的实施方式中，所述多个动作信息的决策信息包括所述多个动作信息的操作，或者所述多个动作信息的决策信息包括所述多个动作信息的优化增益和负面影响，或者所述多个动作信息的决策信息包括所述多个动作信息的操作、所述多个动作信息的优化增益和负面影响，所述优化增益包含动作和动作对象，所述负面影响包含动作和动作对象，所述操作信息包含动作和动作对象。
需要说明的是，本申请实施例中的各个功能模块还可以根据上述方法实施例中的方法具体实现，在此不再赘述。
请参见图12，图12时本申请实施例提供的一种信息处理装置的结构示意图，所述信息处理装置1200至少包括处理器1210、收发器1220以及存储器1230，所述处理器1210、收发器1220以及存储器1230通过总线1240相互连接，其中，
所述处理器1210可以是中央处理器(central processing unit，CPU)，或者CPU和硬件芯片的组合。上述硬件芯片可以是专用集成电路(application-specific integrated circuit，ASIC)，可编程逻辑器件(programmable logic device，PLD)或其组合。上述PLD可以是复杂可编程逻辑器件(complex programmable logic device，CPLD)，现场可编程逻辑门阵列(field-programmable gate array，FPGA)，通用阵列逻辑(generic array logic,GAL)或其任意组合。
所述收发器1220可以包括一个接收器和一个发送器，例如，无线射频模块，以下描述的处理器1210接收或者发送某个消息，具体可以理解为该处理器1210通过该收发器来接收或者发送。
所述存储器1230包括但不限于是随机存取存储器(Random Access Memory，RAM)、只读存储器(Read-Only Memory，ROM)或可擦除可编程只读存储器(Erasable Programmable Read-Only Memory，EPROM或者快闪存储器)，该存储器1230用于存储相关指令及数据，并可以将存储的数据传输给处理器1210。
该信息处理装置1200中的处理器1210用于读取存储器1230中的相关指令执行以下操作：
处理器1210控制收发器1220中的接收器接收意图的多个动作信息；
所述处理器1210确定所述多个动作信息的决策信息，所述决策信息为所述多个动作信息中包含的信息；
处理器1210根据所述多个动作信息的决策信息确定目标动作信息，所述目标动作信息用于实现所述意图。
具体地，上述信息处理装置1200执行的各种操作的具体实现可参照上述方法实施例中第一网元的具体操作，在此不再赘述。
请参见图13，图13时本申请实施例提供的一种信息处理装置的结构示意图，所述信息处理装置1300至少包括处理器1310、收发器1320以及存储器1330，所述处理器1310、收发器1320以及存储器1330通过总线1340相互连接，其中，
所述处理器1310可以是中央处理器(central processing unit，CPU)，或者CPU和硬件芯片的组合。上述硬件芯片可以是专用集成电路(application-specific integrated circuit，ASIC)，可编程逻辑器件(programmable logic device，PLD)或其组合。上述PLD可以是复杂可编程逻辑器件(complex programmable logic device，CPLD)，现场可编程逻辑门阵列(field-programmable gate array，FPGA)，通用阵列逻辑(generic array logic,GAL)或其任意组合。
所述收发器1320可以包括一个接收器和一个发送器，例如，无线射频模块，以下描述的处理器1310接收或者发送某个消息，具体可以理解为该处理器1310通过该收发器来接收或者发送。
所述存储器1330包括但不限于是随机存取存储器(Random Access Memory，RAM)、只读存储器(Read-Only Memory，ROM)或可擦除可编程只读存储器(Erasable Programmable Read-Only Memory，EPROM或者快闪存储器)，该存储器1330用于存储相关指令及数据，并可以将存储的数据传输给处理器1310。
该信息处理装置1300中的处理器1310控制所述收发器1320中的接收器接收来自第一网元的动作选择请求，所述动作选择请求包括多个动作信息的决策信息，所述多个动作信息为所述第一网元根据意图获取的，所述决策信息为所述多个动作信息中包含的信息；
处理器1310根据多个动作信息的决策信息确定目标决策信息，所述目标决策信息用于确定所述目标动作；
所述处理器1310通过收发器1320中的发送器向所述第一网元发送动作选择响应，所述动作选择响应携带所述目标决策信息，或者携带所述目标决策信息的标识信息。
具体地，上述信息处理装置1300执行的各种操作的具体实现可参照上述方法实施例中第二网元的具体操作，在此不再赘述。
请参见图14，图14时本申请实施例提供的一种信息处理装置的结构示意图，所述信息处理装置1400至少包括处理器1410、收发器1420以及存储器1430，所述处理器1410、收发器1420以及存储器1430通过总线1440相互连接，其中，
所述处理器1410可以是中央处理器(central processing unit，CPU)，或者CPU和硬件芯片的组合。上述硬件芯片可以是专用集成电路(application-specific integrated circuit，ASIC)， 可编程逻辑器件(programmable logic device，PLD)或其组合。上述PLD可以是复杂可编程逻辑器件(complex programmable logic device，CPLD)，现场可编程逻辑门阵列(field-programmable gate array，FPGA)，通用阵列逻辑(generic array logic,GAL)或其任意组合。
所述收发器1420可以包括一个接收器和一个发送器，例如，无线射频模块，以下描述的处理器1410接收或者发送某个消息，具体可以理解为该处理器1410通过该收发器来接收或者发送。
所述存储器1430包括但不限于是随机存取存储器(Random Access Memory，RAM)、只读存储器(Read-Only Memory，ROM)或可擦除可编程只读存储器(Erasable Programmable Read-Only Memory，EPROM或者快闪存储器)，该存储器1430用于存储相关指令及数据，并可以将存储的数据传输给处理器1410。
该信息处理装置1400中的处理器1410用于读取存储器1430中的相关指令执行以下操作：
处理器1410通过收发器1420中的接收器接收收来自第一网元或第二网元的动作选择请求，所述动作选择请求，所述动作选择请求包括多个动作信息的决策信息，所述多个动作信息为所述第一网元根据意图获取的，所述决策信息为所述多个动作信息中包含的信息；
所述处理器1410根据多个动作信息的决策信息确定目标决策信息，所述目标决策信息用于确定所述目标动作；
所述处理器1410通过收发器1420中的发送器向所述第一网元或第二网元发送动作选择响应，所述动作选择响应携带所述目标决策信息，或者携带所述目标决策信息的标识信息。具体地，上述信息处理装置1400执行的各种操作的具体实现可参照上述方法实施例中第三网元的具体操作，在此不再赘述。
在上述实施例中，可以全部或部分地通过软件、硬件、固件、或者其任意组合来实现。当使用软件实现时，可以全部或部分地以计算机程序产品的形式实现。所述计算机程序产品包括一个或多个计算机指令。在计算机上加载和执行所述计算机程序指令时，全部或部分地产生按照本申请实施例所述的流程或功能。所述计算机可以是通用计算机、专用计算机、计算机网络、或者其他可编程装置。所述计算机指令可以存储在计算机可读存储介质中，或者从一个计算机可读存储介质向另一个计算机可读存储介质传输，例如，所述计算机指令可以从一个网站站点、计算机、服务器或数据中心通过有线(例如同轴电缆、光纤、数字用户线(DSL))或无线(例如红外、无线、微波等)方式向另一个网站站点、计算机、服务器或数据中心进行传输。所述计算机可读存储介质可以是计算机能够存取的任何可用介质或者是包含一个或多个可用介质集成的服务器、数据中心等数据存储设备。所述可用介质可以是磁性介质(例如软盘、硬盘、磁带)、光介质(例如，DVD)、或者半导体介质(例如固态硬盘Solid State Disk(SSD))等。
以上对本申请实施例进行了详细介绍，本文中应用了具体个例对本申请的原理及实施方式进行了阐述，以上实施例的说明只是用于帮助理解本申请的方法及其核心思想；同时，对于本领域的一般技术人员，依据本申请的思想，在具体实施方式及应用范围上均会有改变之处，综上所述，本说明书内容不应理解为对本申请的限制。

Claims (30)

  1. 一种信息处理方法，其特征在于，所述方法包括：
    第一网元获取意图的多个动作信息；
    所述第一网元确定所述多个动作信息的决策信息，所述决策信息为所述多个动作信息中包含的信息；
    所述第一网元根据所述多个动作信息的决策信息确定目标动作信息，所述目标动作信息用于实现所述意图。
  2. 根据权利要求1所述的方法，其特征在于，所述第一网元根据所述多个动作信息的决策信息确定目标动作信息，包括：
    所述第一网元获取所述意图的业务诉求或所述意图的意图达成的历史经验信息，所述业务诉求是指与所述意图对应的业务要求，所述历史经验信息为与当前场景相同的意图转译事件的信息，所述场景相同是指要实现的意图以及网络状态相同；
    所述第一网元根据所述意图的业务诉求或所述意图达成的历史经验信息以及所述决策信息确定所述目标动作信息。
  3. 根据权利要求1所述的方法，其特征在于，所述第一网元根据所述多个动作信息的决策信息确定目标动作信息，包括：
    所述第一网元将所述多个动作的决策信息输入机器学习模型，得到所述目标动作信息。
  4. 根据权利要求1所述的方法，其特征在于，所述根据所述多个动作信息的决策信息确定目标动作信息，包括：
    所述第一网元向第二网元或第三网元发送动作选择请求，所述动作选择请求携带所述多个动作信息的决策信息，所述动作选择请求用于第二网元或第三网元根据所述多个动作信息的决策信息确定目标决策信息；
    所述第一网元接收来自所述第二网元或所述第三网元的动作选择响应，所述动作选择响应携带所述目标决策信息或所述目标决策信息的标识信息；
    所述第一网元根据所述目标决策信息或所述标识信息确定所述目标动作信息。
  5. 根据权利要求1-4任一项所述的方法，其特征在于，所述第一网元确定所述多个动作信息的决策信息，包括：
    所述第一网元根据所述多个动作信息的优化增益和负面影响确定所多个动作的决策信息，所述优化增益包含动作和动作对象，所述负面影响包含动作和动作对象；
    或者所述第一网元根据所述多个动作信息的操作确定所多个动作的决策信息，所述操作信息包含动作和动作对象。
  6. 根据权利要求5所述的方法，其特征在于，所述第一网元根据所述多个动作信息的优化增益和负面影响确定所多个动作的决策信息，包括：
    所述第一网元根据所述多个动作信息的优化增益和负面影响以及操作确定所多个动作 的决策信息。
  7. 一种信息处理方法，其特征在于，所述方法包括：
    第二网元接收来自第一网元的动作选择请求，所述动作选择请求包括多个动作信息的决策信息，所述多个动作信息为所述第一网元根据意图获取的，所述决策信息为所述多个动作信息中包含的信息；
    所述第二网元根据多个动作信息的决策信息确定目标决策信息，所述目标决策信息用于确定所述目标动作；
    所述第二网元向所述第一网元发送动作选择响应，所述动作选择响应携带所述目标决策信息，或者携带所述目标决策信息的标识信息。
  8. 根据权利要求7所述的方法，其特征在于，所述第二网元根据多个动作信息的决策信息确定目标决策信息，包括：
    在所述第二网元不能处理所述动作选择请求的情况下，所述第二网元向第三网元发送所述动作选择请求；
    所述第二网元接收来自所述第三网元的所述动作选择响应。
  9. 根据权利要求7所述的方法，其特征在于，所述第二网元根据多个动作信息的决策信息确定目标决策信息，包括：
    所述第二网元获取所述意图的业务诉求或所述意图的意图达成的历史经验信息，所述业务诉求是指与所述意图对应的业务要求，所述历史经验信息为与当前场景相同的意图转译事件的信息，所述场景相同是指要实现的意图以及网络状态相同；
    所述第二网元根据所述多个动作的决策信息以及所述诉求信息或所述历史经验信息确定所述目标决策信息。
  10. 根据权利要求7所述的方法，其特征在于，所述第二网元根据多个动作信息的决策信息确定目标决策信息，包括：
    所述第二网元将所述多个动作的决策信息输入机器学习模型，得到所述目标决策信息。
  11. 根据权利要求7-10任一项所述的方法，其特征在于，所述多个动作信息的决策信息包括所述多个动作信息的操作，或者所述多个动作信息的决策信息包括所述多个动作信息的优化增益和负面影响，或者所述多个动作信息的决策信息包括所述多个动作信息的操作、所述多个动作信息的优化增益和负面影响，所述优化增益包含动作和动作对象，所述负面影响包含动作和动作对象，所述操作信息包含动作和动作对象。
  12. 一种信息处理方法，其特征在于，所述方法包括：
    第三网元接收来自第一网元或第二网元的动作选择请求，所述动作选择请求，所述动作选择请求包括多个动作信息的决策信息，所述多个动作信息为所述第一网元根据意图获 取的，所述决策信息为所述多个动作信息中包含的信息；
    第三网元根据多个动作信息的决策信息确定目标决策信息，所述目标决策信息用于确定所述目标动作；
    第三网元向所述第一网元或第二网元发送动作选择响应，所述动作选择响应携带所述目标决策信息，或者携带所述目标决策信息的标识信息。
  13. 根据权利要求12所述的方法，其特征在于，所述第三网元根据多个动作信息的决策信息确定目标决策信息，包括：
    所述第三网元获取所述意图的业务诉求或所述意图的意图达成的历史经验信息，所述业务诉求是指与所述意图对应的业务要求，所述历史经验信息为与当前场景相同的意图转译事件的信息，所述场景相同是指要实现的意图以及网络状态相同；
    所述第三网元根据所述多个动作的决策信息以及所述诉求信息或所述历史经验信息确定所述目标决策信息。
  14. 根据权利要求12所述的方法，其特征在于，所述第三网元根据多个动作信息的决策信息确定目标决策信息，包括：
    所述第二网元将所述多个动作的决策信息输入机器学习模型，得到所述目标决策信息。
  15. 根据权利要求12-14任一项所述的方法，其特征在于，所述多个动作信息的决策信息包括所述多个动作信息的操作，或者所述多个动作信息的决策信息包括所述多个动作信息的优化增益和负面影响，或者所述多个动作信息的决策信息包括所述多个动作信息的操作、所述多个动作信息的优化增益和负面影响，所述优化增益包含动作和动作对象，所述负面影响包含动作和动作对象，所述操作信息包含动作和动作对象。
  16. 一种信息处理装置，其特征在于，包括：
    获取模块，用于获取意图的多个动作信息；
    第一确定模块，用于确定所述多个动作信息的决策信息，所述决策信息为所述多个动作信息中包含的信息；
    第二确定模块，用于根据所述多个动作信息的决策信息确定目标动作信息，所述目标动作信息用于实现所述意图。
  17. 根据权利要求16所述的装置，其特征在于，所述第二确定模块，具体用于获取所述意图的业务诉求或所述意图的意图达成的历史经验信息，所述业务诉求是指与所述意图对应的业务要求，所述历史经验信息为与当前场景相同的意图转译事件的信息，所述场景相同是指要实现的意图以及网络状态相同；根据所述意图的业务诉求或所述意图达成的历史经验信息以及所述决策信息确定所述目标动作信息。
  18. 根据权利要求16所述的装置，其特征在于，所述第二确定模块，具体用于将所述 多个动作的决策信息输入机器学习模型，得到所述目标动作信息。
  19. 根据权利要求16所述的装置，其特征在于，所述第二确定模块，具体用于向第二网元或第三网元发送动作选择请求，所述动作选择请求携带所述多个动作信息的决策信息，所述动作选择请求用于第二网元或第三网元根据所述多个动作信息的决策信息确定目标决策信息；接收来自所述第二网元或所述第三网元的动作选择响应，所述动作选择响应携带所述目标决策信息或所述目标决策信息的标识信息；根据所述目标决策信息或所述标识信息确定所述目标动作信息。
  20. 根据权利要求16-19任一项所述的装置，其特征在于，所述第一确定模块，具体用于根据所述多个动作信息的优化增益和负面影响确定所多个动作的决策信息，所述优化增益包含动作和动作对象，所述负面影响包含动作和动作对象；或者根据所述多个动作信息的操作确定所多个动作的决策信息，所述操作信息包含动作和动作对象。
  21. 根据权利要求20所述的装置，其特征在于，所述第一确定模块，具体用于根据所述多个动作信息的优化增益和负面影响以及操作确定所多个动作的决策信息。
  22. 一种信息处理装置，其特征在于，包括：
    接收模块，用于接收来自第一网元的动作选择请求，所述动作选择请求包括多个动作信息的决策信息，所述多个动作信息为所述第一网元根据意图获取的，所述决策信息为所述多个动作信息中包含的信息；
    确定模块，用于根据多个动作信息的决策信息确定目标决策信息，所述目标决策信息用于确定所述目标动作；
    发送模块，用于向所述第一网元发送动作选择响应，所述动作选择响应携带所述目标决策信息，或者携带所述目标决策信息的标识信息。
  23. 根据权利要求22所述的装置，其特征在于，所述确定模块，具体用于在所述第二网元不能处理所述动作选择请求的情况下，所述第二网元向第三网元发送所述动作选择请求；接收来自所述第三网元的所述动作选择响应。
  24. 根据权利要求22所述的装置，其特征在于，所述确定模块，具体用于获取所述意图的业务诉求或所述意图的意图达成的历史经验信息，所述业务诉求是指与所述意图对应的业务要求，所述历史经验信息为与当前场景相同的意图转译事件的信息，所述场景相同是指要实现的意图以及网络状态相同；根据所述多个动作的决策信息以及所述诉求信息或所述历史经验信息确定所述目标决策信息。
  25. 根据权利要求22所述的装置，其特征在于，所述确定模块，具体用于所述第二网元将所述多个动作的决策信息输入机器学习模型，得到所述目标决策信息。
  26. 根据权利要求22-25任一项所述的装置，其特征在于，所述多个动作信息的决策信息包括所述多个动作信息的操作，或者所述多个动作信息的决策信息包括所述多个动作信息的优化增益和负面影响，或者所述多个动作信息的决策信息包括所述多个动作信息的操作、所述多个动作信息的优化增益和负面影响，所述优化增益包含动作和动作对象，所述负面影响包含动作和动作对象，所述操作信息包含动作和动作对象。
  27. 一种信息处理装置，其特征在于，包括：
    接收模块，用于接收来自第一网元或第二网元的动作选择请求，所述动作选择请求，所述动作选择请求包括多个动作信息的决策信息，所述多个动作信息为所述第一网元根据意图获取的，所述决策信息为所述多个动作信息中包含的信息；
    确定模块，用于根据多个动作信息的决策信息确定目标决策信息，所述目标决策信息用于确定所述目标动作；
    发送模块，用于向所述第一网元或第二网元发送动作选择响应，所述动作选择响应携带所述目标决策信息，或者携带所述目标决策信息的标识信息。
  28. 根据权利要求27所述的装置，其特征在于，所述确定模块，具体用于获取所述意图的业务诉求或所述意图的意图达成的历史经验信息，所述业务诉求是指与所述意图对应的业务要求，所述历史经验信息为与当前场景相同的意图转译事件的信息，所述场景相同是指要实现的意图以及网络状态相同；根据所述多个动作的决策信息以及所述诉求信息或所述历史经验信息确定所述目标决策信息。
  29. 根据权利要求27所述的装置，其特征在于，所述确定模块，具体用于所述第二网元将所述多个动作的决策信息输入机器学习模型，得到所述目标决策信息。
  30. 根据权利要求27-29任一项所述的装置，其特征在于，所述多个动作信息的决策信息包括所述多个动作信息的操作，或者所述多个动作信息的决策信息包括所述多个动作信息的优化增益和负面影响，或者所述多个动作信息的决策信息包括所述多个动作信息的操作、所述多个动作信息的优化增益和负面影响，所述优化增益包含动作和动作对象，所述负面影响包含动作和动作对象，所述操作信息包含动作和动作对象。
PCT/CN2020/117681 2019-09-30 2020-09-25 一种信息处理方法、相关设备及计算机存储介质 WO2021063254A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910945933.0 2019-09-30
CN201910945933.0A CN112583629B (zh) 2019-09-30 2019-09-30 一种信息处理方法、相关设备及计算机存储介质

Publications (1)

Publication Number Publication Date
WO2021063254A1 true WO2021063254A1 (zh) 2021-04-08

Family

ID=75117110

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/117681 WO2021063254A1 (zh) 2019-09-30 2020-09-25 一种信息处理方法、相关设备及计算机存储介质

Country Status (2)

Country Link
CN (1) CN112583629B (zh)
WO (1) WO2021063254A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2023125278A1 (zh) * 2021-12-29 2023-07-06 华为技术有限公司 意图冲突处理的方法及其装置

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115835237A (zh) * 2021-09-17 2023-03-21 华为技术有限公司 通信的方法及其装置
CN115996401A (zh) * 2021-10-20 2023-04-21 华为技术有限公司 一种意图预评估方法及装置

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109245916A (zh) * 2018-08-15 2019-01-18 西安电子科技大学 一种意图驱动的云化接入网系统及方法
CN110098950A (zh) * 2018-01-31 2019-08-06 慧与发展有限责任合伙企业 核实网络意图
CN110166305A (zh) * 2019-06-04 2019-08-23 电子科技大学 一种基于意图的应用QoE管理方法及系统

Family Cites Families (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170091629A1 (en) * 2015-09-30 2017-03-30 Linkedin Corporation Intent platform
US9838284B2 (en) * 2015-10-14 2017-12-05 At&T Intellectual Property I, L.P. Dedicated software-defined networking network for performance monitoring of production software-defined networking network
DK201770431A1 (en) * 2017-05-15 2018-12-20 Apple Inc. Optimizing dialogue policy decisions for digital assistants using implicit feedback
CN107369443B (zh) * 2017-06-29 2020-09-25 北京百度网讯科技有限公司 基于人工智能的对话管理方法及装置
US10944793B2 (en) * 2017-06-29 2021-03-09 Juniper Networks, Inc. Rules-based network security policy modification
CN110120882B (zh) * 2018-02-07 2020-10-09 华为技术有限公司 策略驱动方法和装置
CN110198237B (zh) * 2019-05-27 2021-05-04 北京邮电大学 一种无线意图驱动网络的配置方法
CN110278111B (zh) * 2019-05-29 2021-08-31 西安电子科技大学 一种意图驱动网络通用架构及其意图驱动网络转译方法
CN110247795B (zh) * 2019-05-30 2020-09-25 北京邮电大学 一种基于意图的云网资源服务链编排方法及系统

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110098950A (zh) * 2018-01-31 2019-08-06 慧与发展有限责任合伙企业 核实网络意图
CN109245916A (zh) * 2018-08-15 2019-01-18 西安电子科技大学 一种意图驱动的云化接入网系统及方法
CN110166305A (zh) * 2019-06-04 2019-08-23 电子科技大学 一种基于意图的应用QoE管理方法及系统

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2023125278A1 (zh) * 2021-12-29 2023-07-06 华为技术有限公司 意图冲突处理的方法及其装置

Also Published As

Publication number Publication date
CN112583629A (zh) 2021-03-30
CN112583629B (zh) 2022-06-10

Similar Documents

Publication Publication Date Title
WO2021063254A1 (zh) 一种信息处理方法、相关设备及计算机存储介质
US10313943B2 (en) Dynamic steering of traffic across radio access networks
WO2022011862A1 (zh) 一种o-ran与mec的通信方法及系统
US11290344B2 (en) Policy-driven method and apparatus
US11765584B2 (en) Message processing method and system, and user plane function device
US9392517B2 (en) Facilitation of adaptive traffic flow management by a power-limited mobile device
EP3780496B1 (en) Feature engineering programming method and apparatus
WO2020164590A1 (zh) 传输资源检测方法、传输资源确定方法和通信设备
Cheng et al. Heterogeneous statistical QoS provisioning for downlink transmissions over mobile wireless cellular networks
WO2022083508A1 (zh) 信道状态信息的报告方法、装置及终端
WO2019047913A1 (zh) 一种通信方法、相关设备和系统
US20220182909A1 (en) Device steering based on profile and historical behavior
KR20210100735A (ko) 사이드링크 통신의 서비스 품질을 제어하는 방법 및 장치, 매체, 및 전자 디바이스
JP2024515333A (ja) ネットワークスライス自己最適化方法、基地局、及び記憶媒体
US20160065480A1 (en) Controlling application traffic
WO2014129945A1 (en) Determination of network parameters in mobile communication networks
WO2011110104A1 (zh) 终端接入方法、设备及系统
WO2023045931A1 (zh) 一种网络性能异常分析方法、装置及可读存储介质
US20220021589A1 (en) Method and electronic device for placing micro network function
WO2023185711A1 (zh) 一种用于机器学习模型训练的通信方法及装置
US9769018B2 (en) Reporting technique for a telecommunications network
WO2023179709A1 (zh) 信息处理方法、装置、通信设备及可读存储介质
WO2024093876A1 (zh) 一种通信方法及装置
WO2024027427A1 (zh) 异常检测的方法和通信装置
WO2021204075A1 (zh) 一种网络自动化管理方法及装置

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20870606

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20870606

Country of ref document: EP

Kind code of ref document: A1