CN114531334B

CN114531334B - Intent processing method, device, electronic equipment and readable storage medium

Info

Publication number: CN114531334B
Application number: CN202011215912.2A
Authority: CN
Inventors: 张文
Original assignee: Nanjing ZTE New Software Co Ltd
Current assignee: Nanjing ZTE New Software Co Ltd
Priority date: 2020-11-04
Filing date: 2020-11-04
Publication date: 2025-07-29
Anticipated expiration: 2040-11-04
Also published as: CN114531334A

Abstract

Analyzing the obtained buried point data and initial user intention to determine the user intention to be processed, wherein the initial user intention is used for representing the original requirement of a user; and determining a configuration scheme of the intention network according to the user intention to be processed. The method comprises the steps of obtaining buried point data, analyzing the buried point data and initial user intention, determining the user intention to be processed, simplifying the input information of a user, making up the defect of the initial user intention through the buried point data, expanding understanding of the user intention, eliminating ambiguity of the user intention, determining a configuration scheme of an intention network according to the user intention to be processed, establishing an association relation between the user intention to be processed and configuration information of the intention network, and improving accuracy of intention translation in the intention network.

Description

Intent processing method, device, electronic equipment and readable storage medium

Technical Field

The present application relates to the field of communications networks, and in particular, to an intent processing method, an intent processing device, an electronic device, and a readable storage medium.

Background

An intention Network (IBN) is a closed-loop Network architecture that is built and operated Based on human business intention under the condition of grasping its own "holographic state", and realizes automatic conversion from user intention to specific infrastructure, so that the overall performance of the Network can be monitored, the problems occurring in the Network can be identified and automatically solved without manual intervention. IBN includes intent translation and validation, automation, awareness of network state, assurances, and dynamic optimization/repair functions. Wherein intent translation enables the conversion of the intent of the user's natural language expression (User says) into a network-recognizable intent, a key element to ensure IBN.

At present, in the IBN solution, due to uncertainty of user input, ambiguity may exist in analysis of user intention by the system, and problems of networking structure diversity, complex network configuration parameters and the like are added, so that accuracy of the network configuration scheme is difficult to guarantee, and user experience is reduced.

Disclosure of Invention

The application provides an intent processing method, an intent processing device, an electronic device and a readable storage medium.

The embodiment of the application provides an intention processing method, which comprises the steps of analyzing acquired buried point data and initial user intention, determining the user intention to be processed, wherein the initial user intention is used for representing the original requirement of a user, and determining the configuration scheme of an intention network according to the user intention to be processed.

The embodiment of the application provides an intention processing device which comprises an analysis module and a processing module, wherein the analysis module is used for analyzing acquired buried point data and initial user intention to determine user intention to be processed, the initial user intention is used for representing original requirements of a user, and the processing module is used for determining a configuration scheme of an intention network according to the user intention to be processed.

The embodiment of the application provides electronic equipment, which comprises one or more processors and a memory, wherein one or more programs are stored in the memory, and when the one or more programs are executed by the one or more processors, the one or more processors realize any one of the intended processing methods in the embodiment of the application.

The embodiment of the application provides a readable storage medium storing a computer program which, when executed by a processor, implements any one of the intended processing methods of the embodiment of the application.

According to the intention processing method, the device, the electronic equipment and the readable storage medium, the user intention to be processed is determined by analyzing the obtained buried point data and the initial user intention, so that the input information of a user is simplified, the defect of the initial user intention can be made up through the buried point data, the understanding of the user intention is expanded, the ambiguity of the user intention is eliminated, the configuration scheme of an intention network is determined according to the user intention to be processed, the association relation between the user intention to be processed and the configuration information of the intention network is established, the configuration accuracy of the intention network is ensured, the accuracy of intention translation in the intention network is improved, and the user experience satisfaction is improved.

With respect to the above embodiments and other aspects of the application and implementations thereof, further description is provided in the accompanying drawings, detailed description and claims.

Drawings

Fig. 1 shows a system framework diagram of an IBN in the present application.

FIG. 2 is a flow chart of an intent processing method in an embodiment of the present application.

Fig. 3 shows a schematic diagram of the structure of event metadata and user metadata in an embodiment of the present application.

Fig. 4 shows a schematic structural diagram of a knowledge graph in an embodiment of the present application.

Fig. 5 shows a flow diagram of an intent processing method in a further embodiment of the application.

Fig. 6 shows a block diagram of the intention processing device according to an embodiment of the application.

Fig. 7 shows a block diagram of the composition of the intended processing device in a further embodiment of the application.

Fig. 8 is a flowchart of a method for processing a user intention by the intention processing device based on the data embedded point in the embodiment of the application.

Fig. 9 is a flowchart of a method for complementing user intent based on buried data in an embodiment of the present application.

Fig. 10 is a schematic diagram illustrating intent characteristics obtained by parsing user information in an embodiment of the present application.

FIG. 11 shows a flow diagram of a method of translating pending user intent into a configuration of an intent network in an embodiment of the application.

Fig. 12 shows a block diagram of an exemplary hardware architecture of an electronic device capable of implementing the intent processing method and apparatus according to an embodiment of the application.

Detailed Description

For the purpose of making the objects, technical solutions and advantages of the present application more apparent, embodiments of the present application will be described in detail hereinafter with reference to the accompanying drawings. It should be noted that, without conflict, the embodiments of the present application and features of the embodiments may be arbitrarily combined with each other.

Fig. 1 shows a system framework diagram of an IBN in the present application. As shown in FIG. 1, the IBN may include devices such as a clouding management and control integrated platform 10, an intent engine 20, a Software-defined network (Software-defined Networking, SDN) controller 30, a network infrastructure device 40, and a database 50.

The user 11 inputs the intention information of the user 11 to the clouding management and control integrated platform 10 by operating the clouding management and control integrated platform 10, so that the clouding management and control integrated platform 10 can interact with the SDN controller 30 and the intention engine 20 through a standard interface to monitor the running condition of the IBN network in real time. For example, the user 11 is assisted by software such as "smart assistant" to complete input of intention information and the like.

The intention engine 20 performs semantic analysis on the user intention input by the cloud management and control integrated platform 10 through a northbound interface (Northbound Interface, NBI), converts the information input by the user 11 into the user intention, converts the user intention into a network policy through intention translation and verification, and checks the integrity of the network policy.

The SDN controller 30 is an application in a software defined network responsible for flow control to ensure intelligent operation of the network. The SDN controller interacts with the network base device 40 based on a protocol such as an internet communication protocol (OpenFlow), outputs a user instruction input by the cloud management and control integrated platform 10 to the network base device 40, and forwards a network intention configuration scheme input by the intention engine 20 to the cloud management and control integrated platform 10, so that the cloud management and control integrated platform 10 can display the network intention configuration scheme to the user 11, the user 11 updates the network intention configuration scheme according to own personalized requirements, and the network intention configuration scheme after feedback update by the cloud management and control integrated platform 10 is issued to the device through the SDN controller 30, thereby completing the complete process of intention input, translation, automatic implementation and deployment issuing.

The network infrastructure 40 is used to transmit information in the IBN, for example, the network infrastructure 40 may be a fiber optic device to enable rapid transmission of information in the IBN, etc.

The database 50 is used for storing natural language information input by the user 11 to assist the intent engine 20 in rendering user intent.

However, in the prior art, the intent engine 20 mainly extracts the feature information in the natural language information of the user 11 directly for translating the user intent, and does not refer to other information related to the user 11, which easily results in poor accuracy of translating the natural language information of the user 11, so that the user 11 cannot obtain the network service through the IBN network. For example, in the application of IBN solutions in communication networks, due to the complexity of network architecture and business logic, the used intent translation network device may translate the instruction information in error when obtaining a certain business processing intent instruction input by the user through voice, so that the user cannot obtain the expected physical configuration scheme, thereby reducing the satisfaction of the user experience.

For the purpose of making the objects, technical solutions and advantages of the present application more apparent, the embodiments of the present application will be described in further detail with reference to the accompanying drawings.

Fig. 2 shows a flow diagram of an intent processing method in accordance with an embodiment of the present application. The intention processing method is applicable to an intention processing device. As shown in fig. 2, the intent processing method in the embodiment of the present application may include the following steps.

Step 110, analyzing the obtained buried point data and the initial user intention to determine the user intention to be processed.

Where the initial user intent is used to characterize the user's original needs, for example, the initial user intent is "user A opens a 10GE service" to express what user A wishes to do.

In some implementations, the buried point data is data extracted from log buried point information. For example, the embedded point data related to the intention of the initial user is extracted by collecting the embedded point information of the log and sorting the embedded point information of the log.

In some specific implementations, the method for determining the user intention to be processed includes the steps of analyzing the obtained buried point data and the initial user intention, extracting a first intention feature set in the initial user intention, determining an intention feature set to be analyzed according to the first intention feature set and a preset intention feature set in step 111, determining the user intention to be supplemented according to the buried point data and the intention feature set to be analyzed in step 113, and filling the user intention to be supplemented into the initial user intention to obtain the user intention to be processed in step 114.

For example, the first intention feature set is compared with the preset intention feature total set to determine an intention feature set to be analyzed, the intention feature shared by the intention feature set to be analyzed and the buried point data is extracted, the user intention to be supplemented is determined according to the shared intention feature, and the user intention to be supplemented is filled into the initial user intention to ensure the integrity and the accuracy of the user intention and improve the accuracy of the result of the subsequent translation of the user intention.

Step 120, determining the configuration scheme of the intention network according to the user intention to be processed.

For example, through an intention translation mode, the user intention to be processed is translated into a configuration scheme of an intention network, so that the accuracy of the configuration of the intention network is ensured.

In some specific implementations, the configuration scheme of the intention network is determined according to the user intention to be processed, and the method comprises the steps of determining a set of initial configuration schemes according to a knowledge graph, a preset reasoning algorithm and the user intention to be processed, calculating a weighted value of each initial configuration scheme in the set of initial configuration schemes according to a preset service type, and determining the configuration scheme of the intention network according to the weighted value of each initial configuration scheme and the set of initial configuration schemes, wherein the step 121 is performed.

The method and the system can further refine the screening of each initial configuration scheme by calculating the weighted values of the initial configuration schemes corresponding to different preset service types, and can accurately screen the initial configuration scheme corresponding to a specific service type when the intention of a user to be processed is used for representing the specific service type in the preset service types, so that the screening of each initial configuration scheme in an initial configuration scheme set is quickened, the generation efficiency of the configuration scheme of an intention network is improved, and the user experience satisfaction is improved.

In the embodiment, the user intention to be processed is determined by analyzing the obtained buried point data and the initial user intention, so that the input information of the user is simplified, the defect of the initial user intention can be made up by the buried point data, the understanding of the user intention is expanded, the ambiguity of the user intention is eliminated, the configuration scheme of the intention network is determined according to the user intention to be processed, the association relation between the user intention to be processed and the configuration information of the intention network is established, the configuration accuracy of the intention network is ensured, the accuracy of intention translation in the intention network is improved, and the user experience satisfaction degree is improved.

In one specific implementation, the determining the user intention to be supplemented according to the buried point data and the intention feature set to be analyzed in step 113 includes the following steps:

Step 1131, analyzing the event metadata and the user metadata in the buried point data to obtain buried point information.

It should be noted that the buried data includes two different models, for example, a model of event metadata and a model of user metadata. Fig. 3 shows a schematic diagram of the structure of event metadata and user metadata in an embodiment of the present application. Wherein the event metadata 210 includes a preset event 211, a virtual event 212, and a custom event 213, and the user metadata 220 includes a preset attribute 221 and a custom attribute 222. The attribute information of the preset event 211 in the event metadata 210 may include a user identification 2111, an event identification 2112, location information 2113, behavior information 2114, context information 2115, and custom attribute information 2116. The preset attributes 221 in the user metadata 220 include a user identification 2211, natural attribute information 2212, user usage habit information 2213, and account attribute information 2214. The user usage habit information 2213 may be information such as a long-term usage alert monitoring function, etc., and the natural attribute information 2212 may be information such as age, sex, etc. of the user.

The event metadata is data for representing the behavior of a user, the user metadata is data for representing attribute characteristics of the user, the embedded point information comprises an embedded point intention characteristic set and an embedded point event set, and the embedded point intention characteristic set comprises embedded point intention characteristics. By classifying the buried point information and extracting different types of information, the processing of buried point data can be quickened, the subsequent processing of the buried point information is convenient, and the processing efficiency is improved.

Step 1132, under the condition that it is determined that the buried point event set has a matching event matching with the preset mapping relationship, extracting the matching event in the buried point event set, and generating a matching event set.

The preset mapping relation is a mapping relation between a preset event and a preset intention feature. For example, the preset event is an alarm inquiry, the preset intention features comprise an alarm identifier, an alarm type and an alarm generation time, and by comparing the event in the buried point event set with the preset event, when the alarm inquiry exists in the buried point event set, the existence of a matching event in the buried point event set is indicated, and the matching event is the alarm inquiry, so that the preset intention features corresponding to the matching event can be rapidly determined to comprise the alarm identifier, the alarm type and the alarm generation time according to the preset mapping relation. The extraction speed of the intention characteristics of the matching event is improved, and the screening of the matching event is quickened.

Step 1133, determining the user intention to be supplemented according to the matching event set, the buried point intention feature set and the intention feature set to be analyzed.

In some specific implementations, determining the user intention to be supplemented according to the matching event set, the buried point intention feature set and the intention feature set to be analyzed comprises extracting a second intention feature from the buried point intention feature set according to the matching event in the matching event set, wherein the second intention feature is the buried point intention feature corresponding to the matching event, searching the intention feature set to be analyzed according to the second intention feature to obtain a searching result, calculating a feature value corresponding to the second intention feature when the searching result is determined to be the intention feature identical to the second intention feature in the intention feature set to be analyzed, and determining the user intention to be supplemented according to the second intention feature and the feature value corresponding to the second intention feature.

For example, a machine learning Algorithm (MACHINE LEARNING Algorithm, MLA) may be used to calculate the feature value corresponding to the second intent feature, so that the intent processing device may more intuitively understand the embedded point intent feature, and ensure accuracy of processing the embedded point intent feature in the embedded point intent feature set. Then, according to the second intention characteristic and the corresponding characteristic value thereof, the intention of the user to be supplemented is determined in an intention translation mode, so that the integrity of the intention of the user is ensured, and the subsequent analysis of the intention of the user can be accurate and more fit with the requirement of the user.

The method comprises the steps of extracting a matching event matched with a preset mapping relation from a buried point event set, accelerating the processing of buried point data, improving the processing efficiency of the data, extracting a second intention characteristic corresponding to the matching event from a buried point intention characteristic set, calculating a characteristic value of the second intention characteristic, enabling an intention processing device to understand the buried point intention characteristic more intuitively, improving the processing accuracy of equipment, determining the intention of a user to be supplemented in an intention translation mode, ensuring the integrity of the intention of the user, and enabling the subsequent analysis of the intention of the user to be accurate and more fit the requirement of the user.

In some specific implementations, the step 114 of filling the user intention to be supplemented into the initial user intention to obtain the user intention to be processed includes extracting a second intention feature in the user intention to be supplemented, calculating to obtain a feature value corresponding to the second intention feature, filling the second intention feature and the feature value corresponding to the second intention feature into a first intention feature set in the initial user intention to generate a to-be-processed intention feature set, obtaining a to-be-processed event corresponding to the intention feature in the to-be-processed intention feature set, and determining the user intention to be processed according to the to-be-processed intention feature set and the to-be-processed event.

For example, if the user inputs "user a wants to open a 10GE service", the initial intention feature of the user may be represented as { person, service type, layer rate, time }, where person is "user a", service type is "client layer", layer rate is "10GE", and time is "9 months in 2020, 24 days". And filling the second intention feature { source, sink, protection type, intelligence degree } and the corresponding feature value thereof into the first intention feature set in the intention of the initial user to generate an intention feature set to be processed, wherein the intention feature set to be processed can be expressed as { character, business type, layer rate, source, sink, protection type, intelligence degree, time }.

Through the intention feature set to be processed, the intention of the user to be processed is determined, the intention of the user is expanded, the intention to be expressed by the user can be more fully understood by the intention processing device, and the completeness of the intention of the user is ensured. More reliable data is provided for the subsequent translation of the user intention to be processed into the configuration scheme of the intention network, and the translation accuracy of the intention network is improved.

In some embodiments, determining the set of initial configuration schemes according to the knowledge graph, the preset inference algorithm, and the user intention to be processed in step 121 includes generating a set of first configuration schemes based on the knowledge graph and the preset inference algorithm, wherein the knowledge graph includes entity elements, attribute elements, and relationship elements, training schemes in the set of first configuration schemes according to the user intention to be processed to obtain training results, sorting the training results according to a scoring function to obtain a first sorting result, wherein the scoring function is a function determined according to a degree of association between the entity elements and the relationship elements, and determining the set of initial configuration schemes according to the first sorting result.

For example, fig. 4 shows a schematic structural diagram of a knowledge graph in an embodiment of the present application. Wherein, a knowledge graph can be expressed as a triplet (subject, relationship, object). For example, as shown in fig. 4, the son of the daming is the junior, and is represented as the triplet (daming, son, junior). The subject and the object are collectively referred to as an entity (entity). The relationship is irreversible, meaning that the subject and object cannot be reversed. The collection of knowledge maps are linked into a graph, each node is an entity, each edge is a relationship or a fact (fact), and the meaning of each edge representation is that the subject points to the object. Thus, the knowledge graph is a directed graph.

And then training the schemes in the set of the first configuration schemes by taking the user intention to be processed as an input parameter to obtain a training result, so that the training result can be more in line with the user intention to be processed. And taking the training result ranked at the front in the first sorting result as an initial configuration scheme to obtain a plurality of initial configuration schemes, for example, extracting the training result ranked at the front 10 bits in the first sorting result to generate a set of initial configuration schemes so as to improve the configuration scheme of the intention network to be more consistent with the intention of the user to be processed and improve the accuracy of translating the intention of the user into the configuration scheme of the intention network.

In some implementations, calculating the weighted value of each initial configuration scheme in the set of initial configuration schemes according to the preset service type in step 122 includes:

step 1221, extracting configuration parameter information of each initial configuration scheme and initial event information corresponding to the configuration parameter information in sequence.

Step 1222, determining target event index information corresponding to the preset service type according to the preset service type.

It should be noted that, the target event index information corresponding to different service types is also different. The preset service type may be service issuing or processing of alarm information, and the above description is only illustrative for the preset service type, and may be specifically set according to actual needs, and other undescribed preset service types are also within the protection scope of the present application, which is not described herein.

Step 1223, calculating the weighted value of each initial configuration scheme according to the target event index information and the initial event information.

For example, the target event index information and the initial event information are compared, when the related information in the initial event information meets the target event index, the weighting value of the corresponding initial configuration scheme is higher, and when the related information in the initial event information is far away from the target event index, the weighting value of the corresponding initial configuration scheme is lower. Different initial configuration schemes can be ensured to determine the weighted value of the initial configuration scheme according to the degree that the corresponding initial event information is away from the target event index information, so that the quality degree of the different initial configuration schemes is distinguished, the initial configuration schemes can be conveniently screened, and the configuration scheme of the intention network which meets the requirements of users can be obtained.

In some embodiments, the calculating the weighted value of each initial configuration scheme according to the target event index information and the initial event information in step 1223 includes determining a conversion rate of successful service delivery of each initial configuration scheme according to the first target event index information and each initial event information corresponding to the service delivery in case that the preset service type is determined to be the service delivery, determining the weighted value of each initial configuration scheme according to the conversion rate of successful service delivery of each initial configuration scheme in case that the preset service type is determined to be the alarm pressure reduction, determining an alarm retention rate of each initial configuration scheme according to the second target event index information and each initial event information corresponding to the alarm pressure reduction in case that the preset service type is determined to be the alarm pressure reduction, and determining the weighted value of each initial configuration scheme according to the alarm retention rate of each initial configuration scheme.

For example, the alarm pressure reduction levels are divided into a first level, a second level and a third level, and the lower the number of levels, the higher the corresponding alarm retention rate, for example, when the alarm pressure reduction level is the first level, the corresponding alarm retention rate is 30%, when the alarm pressure reduction level is the second level, the corresponding alarm retention rate is 60%, and when the alarm pressure reduction level is the third level, the corresponding alarm retention rate is 75%. I.e. the user's tolerance to alarm pressure reduction is different.

The method comprises the steps of sequentially extracting configuration parameter information of each initial configuration scheme and initial event information corresponding to the configuration parameter information, determining target event index information corresponding to the preset service type according to the preset service type, calculating weighted values of each initial configuration scheme according to the target event index information and the initial event information, reflecting the quality degree of the configuration scheme corresponding to different service types through the weighted values, ensuring the configuration accuracy of an intention network and improving the user experience satisfaction.

In some embodiments, determining the configuration scheme of the intent network according to the weighted value of each initial configuration scheme and the set of initial configuration schemes in step 123 includes re-ordering each initial configuration scheme in the set of initial configuration schemes according to the first ordering result and the weighted value of each initial configuration scheme to obtain a second ordering result, and determining the configuration scheme of the intent network according to the second ordering result.

The weighted value of each initial configuration scheme is used as a reference value of the second order, so that the second order can be more accurately matched with the configuration schemes of the intention networks corresponding to different service types, and the configuration accuracy of the intention networks is improved.

Fig. 5 shows a flow diagram of an intent processing method in a further embodiment of the application. The intention processing method is applicable to an intention processing device. As shown in fig. 5, the intent processing method in the embodiment of the present application may include the following steps.

Step 410, obtaining user information according to the intention starting event identification and the intention ending event identification.

Wherein the user information includes operation information of the user and input information of the user. For example, the operation information of the user may be opening an operation web page to browse some web pages, and the input information of the user may be information such as a user name and a password input when the user logs in some web pages. When the user enters the intention starting page (i.e., the operation of "user enters the intention starting page" is identified as an intention starting event), a session of the user's intention is opened, and a new intention session is generated. When the user enters a certain intention termination page (namely, the operation of 'entering the intention termination page by the user', the intention processing device can select user operation information between the intention starting event identification and the intention termination event identification and input information of the user as user information of the current intention session so as to facilitate extraction of the user information.

Step 420, determining the initial user intention according to the operation information of the user and the input information of the user.

Wherein the input information includes intent text and/or voice information, the initial user intent being a machine-recognizable intent language.

For example, when the input information is an intention text, the intention text is converted into an initial user intention, i.e., a machine-recognizable intention language, by a machine learning technique such as command parsing. When the user information is voice information, the voice information needs to be converted into text information, and then the text information is converted into initial user intention according to the processing mode of the text information. The input information in different forms can be converted into intention voice which can be recognized by a machine, and the intention processing device can conveniently process the input information further.

Step 430, obtaining the embedded point data from the data server according to the intended initiating event identification and the embedded point port information of the data server.

It should be noted that, when the user starts the intention session (for example, the intention start event identifier is taken as an identifier for starting to extract the embedded point data), the embedded point data stored in the data server is obtained through the embedded point port (for example, the port number is 4500, etc.) of the data server at the same time, and the embedded point data supplements the user intention, so that the user intention can be expressed as machine-recognizable information more fully, and the accuracy of the configuration scheme for translating the user intention into the intention network is improved.

Step 440, analyzing the obtained buried point data and the initial user intention to determine the user intention to be processed.

Step 450, determining the configuration scheme of the intention network according to the user intention to be processed.

It should be noted that, steps 440 to 450 in the present embodiment are the same as steps 110 to 120 in the previous embodiment, and are not described herein.

In the embodiment, the user information is acquired according to the intention starting event identification and the intention ending event identification, the embedded point data is acquired from the data server according to the intention starting event identification and the embedded point port information of the data server, the acquired embedded point data and the initial user intention are analyzed, and the user intention to be processed is determined, so that the user intention to be processed is more sufficient, the requirements of the user can be expressed more accurately, and the completeness and the accuracy of the user intention are ensured. And determining a configuration scheme of the intention network according to the user intention to be processed. The configuration accuracy of the intention network is guaranteed, the expansibility of the intention network is improved, and the user experience satisfaction is improved.

In some implementations, after the step of determining the configuration scheme of the intent network according to the user intent to be processed, the method further includes feeding back the configuration scheme of the intent network to the user, obtaining modification parameters determined by the user based on the configuration scheme of the intent network, and updating the configuration scheme of the intent network according to the modification parameters and the configuration scheme of the intent network, wherein the configuration scheme of the intent network includes network configuration parameters.

The user can update the configuration scheme of the intention network according to personalized requirements by acquiring the modification parameters determined by the configuration scheme of the user based on the intention network, so that the accuracy of the configuration parameters of the intention network is fully ensured.

In some implementations, after the step of updating the configuration scheme of the intent network in accordance with the modification parameters and the configuration scheme of the intent network, the method further includes configuring the intent network in accordance with the updated configuration scheme of the intent network, and deleting the cached initial user intent and buried point data if it is determined that the configuration of the intent network is successful.

The method comprises the steps of updating the configuration scheme of the intention network, configuring the intention network, ensuring that the configuration of the intention network meets the requirements of users and is closer to the intention of the users, deleting the cached initial user intention and buried point data under the condition that the successful configuration of the intention network is confirmed, saving storage space, facilitating the next processing of the user intention and the corresponding buried point data, avoiding confusion of the data and improving the processing efficiency of the data.

An apparatus according to an embodiment of the present application will be described in detail below with reference to the accompanying drawings. Fig. 6 shows a block diagram of the intention processing device according to an embodiment of the application. As shown in fig. 6, the intention processing device may include the following modules:

the system comprises an analysis module 510 for analyzing the obtained buried point data and an initial user intention to determine a user intention to be processed, wherein the initial user intention is used for representing the original requirement of a user, and a processing module 520 for determining a configuration scheme of an intention network according to the user intention to be processed.

In some implementations, the analysis module 510 includes an extraction sub-module configured to extract a first intention feature set in an initial user intention, a first determination sub-module configured to determine an intention feature set to be analyzed according to the first intention feature set and a preset intention feature set, a second determination sub-module configured to determine a user intention to be supplemented according to buried point data and the intention feature set to be analyzed, and a filling sub-module configured to fill the user intention to be supplemented into the initial user intention to obtain the user intention to be processed.

In some embodiments, the second determining submodule is configured to analyze event metadata and user metadata in the embedded point data to obtain embedded point information, where the event metadata is data for characterizing a behavior of a user, the user metadata is data for characterizing an attribute feature of the user, the embedded point information includes an embedded point intention feature set and an embedded point event set, the embedded point intention feature set includes an embedded point intention feature, and if it is determined that a matching event matching a preset mapping relationship exists in the embedded point event set, the matching event in the embedded point event set is extracted to generate a matching event set, where the preset mapping relationship is a mapping relationship between a preset event and a preset intention feature, and determine a user intention to be supplemented according to the matching event set, the embedded point intention feature set, and the intention feature set to be analyzed.

In some specific implementations, the determining of the user intention to be supplemented according to the matching event set, the buried point intention feature set and the intention feature set to be analyzed in the second determining submodule comprises extracting a second intention feature from the buried point intention feature set according to the matching event in the matching event set, wherein the second intention feature is the buried point intention feature corresponding to the matching event, searching the intention feature set to be analyzed according to the second intention feature to obtain a searching result, calculating a feature value corresponding to the second intention feature when the searching result is that the intention feature identical to the second intention feature exists in the intention feature set to be analyzed, and determining the user intention to be supplemented according to the second intention feature and the feature value corresponding to the second intention feature.

In some implementations, a filling sub-module is used for extracting second intention features in user intention to be supplemented, calculating and obtaining feature values corresponding to the second intention features, filling the second intention features and the feature values corresponding to the second intention features into a first intention feature set in initial user intention to generate a to-be-processed intention feature set, acquiring to-be-processed events corresponding to the intention features in the to-be-processed intention feature set, and determining to-be-processed user intention according to the to-be-processed intention feature set and the to-be-processed events.

In some embodiments, the processing module 520 includes an initial configuration scheme determining submodule configured to determine a set of initial configuration schemes according to a knowledge graph, a preset inference algorithm, and a user intention to be processed, a weighted value determining submodule configured to calculate a weighted value of each initial configuration scheme in the set of initial configuration schemes according to a preset service type, and a configuration scheme determining submodule configured to determine a configuration scheme of the intention network according to the weighted value of each initial configuration scheme and the set of initial configuration schemes.

In some implementations, an initial configuration scheme determining submodule is configured to generate a set of first configuration schemes based on a knowledge graph and a preset reasoning algorithm, wherein the knowledge graph includes entity elements, attribute elements and relationship elements, train schemes in the set of first configuration schemes according to user intention to be processed to obtain training results, sort the training results according to a scoring function to obtain a first sorting result, wherein the scoring function is a function determined according to a degree of association between the entity elements and the relationship elements, and determine the set of initial configuration schemes according to the first sorting result.

In some implementations, the configuration scheme determining submodule is configured to reorder each initial configuration scheme in the set of initial configuration schemes according to the first ordering result and the weighted value of each initial configuration scheme to obtain a second ordering result, and determine a configuration scheme of the intended network according to the second ordering result.

In some implementations, the weighting value determining submodule is used for sequentially extracting configuration parameter information of each initial configuration scheme and initial event information corresponding to the configuration parameter information, determining target event index information corresponding to the preset service type according to the preset service type, and calculating the weighting value of each initial configuration scheme according to the target event index information and the initial event information.

In some specific implementations, the weighting value determining sub-module calculates the weighting value of each initial configuration scheme according to the target event index information and the initial event information, and comprises determining the conversion rate of successful service issuance of each initial configuration scheme according to the first target event index information and the initial event information corresponding to the service issuance when the preset service type is determined to be the service issuance, determining the weighting value of each initial configuration scheme according to the conversion rate of successful service issuance of each initial configuration scheme, determining the alarm retention rate of each initial configuration scheme according to the second target event index information and the initial event information corresponding to the alarm pressure reduction when the preset service type is determined to be the alarm pressure reduction, and determining the weighting value of each initial configuration scheme according to the alarm retention rate of each initial configuration scheme.

In some implementations, the intent processing device further includes an acquisition user information module configured to acquire user information according to the intent initiation event identification and the intent termination event identification, wherein the user information includes operation information of a user and input information of the user, a determination module configured to determine an initial user intent according to the operation information of the user and the input information of the user, wherein the input information includes intent text and/or voice information, the initial user intent is a machine-recognizable intent language, and an acquisition buried data module configured to acquire buried data from the data server according to the intent initiation event identification and buried port information of the data server.

In some specific implementations, the intent processing device further comprises a feedback module for feeding back the configuration scheme of the intent network to the user, a parameter modification module for obtaining modification parameters determined by the user based on the configuration scheme of the intent network, and an updating module for updating the configuration scheme of the intent network according to the modification parameters and the configuration scheme of the intent network, wherein the configuration scheme of the intent network comprises network configuration parameters.

In some specific implementations, the intention processing device further comprises a configuration module for configuring the intention network according to the updated configuration scheme of the intention network, and a deletion module for deleting the cached initial user intention and buried point data under the condition that the intention network is determined to be successfully configured.

In some implementations, the embedded point data in the intended processing device is data extracted from log embedded point information.

According to the intention processing device, the analysis module is used for analyzing the obtained buried point data and the initial user intention, determining the user intention to be processed, simplifying the input information of a user, making up the defect of the initial user intention through the buried point data, expanding the understanding of the user intention, eliminating the ambiguity of the user intention, determining the configuration scheme of an intention network according to the user intention to be processed, establishing the association relation between the user intention to be processed and the configuration information of the intention network, ensuring the configuration accuracy of the intention network, improving the accuracy of intention translation in the intention network and improving the user experience satisfaction.

It should be clear that the application is not limited to the specific arrangements and processes described in the foregoing embodiments and shown in the drawings. For convenience and brevity of description, detailed descriptions of known methods are omitted herein, and specific working processes of the systems, modules and units described above may refer to corresponding processes in the foregoing method embodiments, which are not repeated herein.

Fig. 7 shows a block diagram of the composition of the intended processing device in a further embodiment of the application. As shown in fig. 7, the intention processing device based on the data buried point includes an intention feedback module 610, an intention analysis module 620, a buried point analysis module 630, and a session control module 640, which are sequentially connected. The buried point analysis module 630 includes a preprocessing unit 631, a result reorganization unit 632, an event analysis unit 633, an index evaluation unit 634, and a data extraction unit 635, which are sequentially connected.

The intention feedback module 610 is configured to send information such as the user intention and the configuration scheme of the intention network corresponding to the user intention output by the intention analysis module 620 to the display interface, so that the user can obtain the configuration scheme of the intention network through the display interface.

The intent resolution module 620 is used to convert information entered by the user (e.g., intent text and/or voice information, etc.) into an initial user intent, i.e., a machine-recognizable intent language. For example, the intent resolution module 620 obtains the original intent text entered by the user and the behavioral events of the user, eliminates invalid text input by analyzing the behavioral events of the user and using text preprocessing functions, and converts the valid text input into the original user intent by machine learning techniques such as command resolution, semantic slot filling, and the like. The intention analysis module 620 invokes the buried point analysis module 630 to obtain the user intention to be processed after semantic slot filling returned by the result reorganization unit 632 in the buried point analysis module 630, so as to ensure the integrity and accuracy of the user intention. According to the user intention to be processed and the weighted value input by the index evaluation unit 634 in the buried point analysis module 630, a configuration scheme of the intention network is determined, and the configuration scheme of the intention network is fed back to the intention feedback module 610.

It should be noted that, the intent analysis module 620 processes the knowledge graph and the user intent to be processed (for example, through the process of generating the intent policy, network arrangement, and verification) by adopting a preset reasoning algorithm, and generates a set of initial configuration schemes. The method comprises the steps of carrying out evaluation optimization on each initial configuration scheme in a set of initial configuration schemes by combining weighted values input by an index evaluation unit 634, determining the configuration scheme of an intention network finally recommended to a user, outputting the configuration scheme of the intention network to an intention feedback module 610 so that the intention feedback module 610 feeds back to the user to obtain modification parameters determined by the user based on the configuration scheme of the intention network, and updating the configuration scheme of the intention network according to the modification parameters and the configuration scheme of the intention network. Ensuring the configuration accuracy of the intention network.

The embedded point analysis module 630 is configured to read embedded point data from the data server, analyze the embedded point data, determine a user intention to be supplemented, and feed back the user intention to be supplemented to the intention analysis module 620. Meanwhile, according to the different service types, the weighting value of each initial configuration scheme is obtained through calculation, and the weighting value is fed back to the intention analysis module 620.

The preprocessing unit 631 is configured to perform preprocessing operations such as feature extraction, attribute comparison, and the like on the initial user intention input by the intention analysis module 620, and obtain a first intention feature set, where the first intention feature set includes a plurality of first intention features. Comparing the first intention characteristic with the intention characteristic in the preset intention characteristic total set, and determining an intention characteristic set to be analyzed.

The result reorganizing unit 632 is configured to obtain the set of buried point intention features input by the event analyzing unit 633 and feature values corresponding to the respective second intention features, where the set of buried point intention features includes the second intention features. The second intention feature and the corresponding feature value thereof are processed, converted into a user intention to be supplemented, and the user intention to be supplemented is sent to the intention analysis module 620.

The event analysis unit 633 is configured to analyze the buried-point intention feature set, the buried-point event set and the preset mapping relationship input by the data extraction unit 635, generate a matching event set (for example, in a case where it is determined that there is a matching event matching the preset mapping relationship in the buried-point event set, the matching event set is generated by extracting a matching event in the buried-point event set), where the buried-point intention feature set includes a buried-point intention feature, the preset mapping relationship is a mapping relationship between the preset event and the preset intention feature, extract a second intention feature from the buried-point intention feature set according to the matching event in the matching event set, where the second intention feature is a buried-point intention feature corresponding to the matching event, search the intention feature set to be analyzed input by the preprocessing unit 631 according to the second intention feature, obtain a search result, calculate a feature value corresponding to the second intention feature if it is determined that there is an intention feature identical to the second intention feature in the intention feature set to be analyzed, and output the second intention feature value corresponding to the result recombination unit 632.

The index evaluation unit 634 is configured to obtain the set of buried point intention characteristics, the set of buried point events, and the set of preset mapping relationships from the data extraction unit 635. The user intention to be processed, which is input by the intention analysis module 620, is combined for comprehensive analysis, corresponding weighted values are generated according to different service types, and the weighted values are fed back to the intention analysis module 620.

For example, when the preset service type is determined to be service release, setting the first target event index information corresponding to service release to be the conversion rate of the page with successful service release, and determining the conversion rate of the page with successful service release within a preset time period (for example, within 10 minutes) through a funnel analysis method and an evaluation algorithm. When the preset service type is determined to be the alarm pressure reduction, the correlation analysis is adopted to analyze the user who has performed the alarm confirmation and/or the alarm clearing and other event operations within the preset time, so that the alarm retention rate corresponding to different alarm pressure reduction rules is evaluated when the user expresses the user intention of the 'quick pressure reduction' alarm during the alarm peak period. (for example, the setting of the minimum or maximum alarm amount is the Mth week, and the second target event index information corresponding to the alarm pressure reduction is the association relationship between the "Mth week alarm retention" and the "alarm pressure reduction" event). And determining a conversion trend according to the conversion rate or the alarm retention rate of the page with successful service distribution, and taking the conversion trend as a weighted value of the configuration scheme of the intention network.

The data extraction unit 635 is configured to process the buried data, and convert the buried data into buried information, where the buried information includes a buried intention feature set and a buried event set. Setting a preset mapping relation, wherein the preset mapping relation is a mapping relation between a preset event and a preset intention characteristic.

For example, the embedded point data (e.g., event metadata, user metadata, etc.) is read from the data server, then parsed to obtain JS object tag (JavaScript Object Notation, JSON) data, the user identifier and the embedded point event corresponding to the preset mapping relationship are extracted from the JSON data according to the preset conditions, the embedded point event is filtered, de-duplicated, etc. to generate processed JSON data, and the processed JSON data is sent to the index evaluation unit 634 for analysis. The JSON data is easy to read and write by people, and is easy to analyze and generate by machines, so that the network transmission efficiency is effectively improved.

Session control module 640 is configured to maintain a processing flow of user intention to be processed, management of context dialogue information, and the like, so as to track user behavior and events.

In this embodiment, the information input by the user is converted into the initial user intention by the intention analysis module, so that the machine can recognize the user intention which the user wants to express. Analyzing and processing the buried point data through each unit in the buried point analysis module to obtain the user intention to be processed, which is returned by the result recombination unit in the buried point analysis module and is filled by the semantic slot, so as to ensure the integrity and the accuracy of the user intention. According to the user intention to be processed and the weighted value input by the index evaluation unit in the buried point analysis module, determining the configuration scheme of the intention network, and feeding back the configuration scheme of the intention network to the intention feedback module, so that the user can timely obtain the matched configuration scheme of the intention network, the user can conveniently and rapidly and accurately configure the intention network, the accuracy of intention translation in the intention network is improved, and the user experience satisfaction is improved.

Fig. 8 is a flowchart of a method for processing a user intention by the intention processing device based on the data embedded point in the embodiment of the application. As shown in fig. 8, the intention processing device based on the data buried point adopts the following steps to realize the processing of the intention of the user.

Step 701, obtaining user information in response to an intent session request sent by a user.

Wherein the user information includes operation information of the user and input information of the user.

For example, when the user enters the intention starting page (i.e., the operation of "user enters the intention starting page" is regarded as a starting Event (start_event)), a session of the user's intention is started, and a new intention session is generated. Meanwhile, an acquisition switch for the embedded point data is started (for example, the acquisition of the embedded point data is started by adopting a software development kit (Software Development Kit, SDK)) to automatically collect embedded point information corresponding to the current intention session, and when a user enters a certain (or a plurality of) intention termination pages (namely, the operation of 'the user entering the intention termination page' is used as a termination Event (end_event)), or when the running time of the current intention session exceeds a preset session time (for example, 6 hours and the like), the current intention session is closed. It should be noted that, different intention sessions are distinguished by session identification and service domain identification, so as to facilitate extraction and processing of relevant information of disagreement to the graph session.

Step 702, buried point data associated with user information is obtained.

For example, a semi-automatic embedded point mode is adopted to standardize part of manual work to form an SDK, the SDK is embedded in a product, and when the user information comprises an initial event, the embedded point analysis module reads embedded point data from a data server in a mode of embedding the SDK.

Wherein the buried data includes event metadata and user metadata. The event metadata comprises any one or more of a preset event, a virtual event and a custom event, and the user metadata comprises any one or more of user identification (UserID), user attribute information (such as age, gender and the like of a user), user usage habit information (such as long-term usage alarm monitoring and the like) and account attribute information. The preset event includes any one or more of user identification, event identification, location information, behavior information, context information, and event attribute information. The custom Event may be an Event acquired through a trace Event (track_event) interface of the SDK, which is main data for analyzing user behavior and buried point data.

It should be noted that the event metadata and the user metadata are associated by a unique identification (e.g., userID, or device identification, etc.). The unique identification is also different depending on the difference in calling mode of the interface.

In some specific implementations, according to the service type intended by the user, the custom Event may be any one or more of behavior events such as browsing operation (ViewTopo _event) of a certain topological graph object, wire configuration operation (LinkCfg _event) between submitted network elements, board installation operation (SetupBoard _event) and network element configuration operation (NetCfg _event), and the custom Event may also include attribute information of events such as preset wire type, network element name, wire name, operation time, and the like. Then, the event metadata is stored in a data server in the form of event metadata for the calling analysis of the intention processing device, so that the subsequent processing is convenient. It should be noted that the foregoing examples of the custom event are only illustrative, and other non-illustrative custom events are also within the protection scope of the present application, and may be specifically set according to specific situations, which are not described herein.

In step 703, the user information is analyzed to obtain an initial user intent.

Under the condition that the user information is the intention text, invalid text information is removed by using a text preprocessing function, effective text information is obtained, and the effective text information is converted into the initial user intention, namely the intention language which can be recognized by a machine through technical means of machine learning such as command analysis or semantic slot filling.

In the case where it is determined that the user information is speech information, the original speech signal is converted to text information by an automatic speech recognition (Automatic Speech Recognition, ASR) technique, and the text information is converted to a frame semantic representation by a natural language understanding (Natural Language Processing, NLU) technique. For example, text information is converted into a frame semantic representation using a developer-oriented natural language semantic understanding service (Language Understanding INTELLIGENT SERVICE, LUIS) platform.

And step 704, supplementing the user intention to be supplemented, which is related to the user information, in the embedded data to the initial user intention to obtain the user intention to be processed.

For example, a preprocessing function is called, preprocessing operations such as cleaning and validity checking are performed on the buried data, and a buried point intention feature set and a buried point event set of the buried data are determined. Extracting a matching event in the buried point event set under the condition that the matching event matched with a preset mapping relation exists in the buried point event set, generating a matching event set, wherein the buried point intention feature set comprises buried point intention features, the preset mapping relation is a mapping relation between the preset event and the preset intention features, extracting a second intention feature from the buried point intention feature set according to the matching event in the matching event set, wherein the second intention feature is the buried point intention feature corresponding to the matching event, searching the intention feature set to be analyzed input by the preprocessing unit 631 according to the second intention feature, obtaining a searching result, calculating a feature value corresponding to the second intention feature under the condition that the searching result is the intention feature set to be analyzed has the same intention feature as the second intention feature, and filling the second intention feature and the feature value corresponding to the second intention feature into an initial user intention to obtain the user intention to be processed.

Step 705, determining the configuration scheme of the intention network according to the user intention to be processed.

In the semantic network, the knowledge graph is composed of three elements of an entity, an attribute and a relation, and the knowledge graph can be expressed as a triplet (head entity, relation and tail entity). Wherein the head entity and the tail entity are collectively referred to as an entity, the entity including attribute information. The method comprises the steps of obtaining a set of initial configuration schemes by adopting a conversion embedding (TRANSLATING EMBEDDING, transE) algorithm and carrying out knowledge reasoning based on a knowledge graph, calculating the weighted value of each initial configuration scheme in the set of initial configuration schemes according to a preset service type, sorting each initial configuration scheme according to the weighted value of each initial configuration scheme to obtain a sorting result, and determining the configuration scheme of an intention network according to the sorting result. The intention network is essentially a combination of network features, such as various combinations of topology nodes, tunnel policies, and the like.

Step 706, feeding back the configuration scheme of the intention network to the user, obtaining the modification parameters determined by the user based on the configuration scheme of the intention network, and updating the configuration scheme of the intention network according to the modification parameters and the configuration scheme of the intention network.

It should be noted that, after the configuration scheme of the intended network is obtained, some shortages may still exist, or a plurality of different configuration schemes exist at the same time, at this time, the plurality of different configuration schemes need to be fed back to the user, so that the user can select according to the own needs, or the user can modify relevant parameters in the configuration schemes according to the personalized needs. When the modification parameters determined by the user based on the configuration scheme of the intention network are obtained, the intention processing device corrects the configuration scheme of the intention network selected by the user again according to the modification parameters to obtain the final configuration scheme of the intention network, so that the accuracy of the configuration parameters of the intention network is fully ensured.

Step 707, configuring the intention network by using the updated configuration scheme of the intention network.

When the configuration of the intent network is completed and the configuration success message fed back by the underlying network is received, the intent processing device needs to clear the relevant cache data (e.g., the cached initial user intent, and buried point data related to the initial user intent, etc.) of the intent session.

In this embodiment, the machine is enabled to recognize the user intention that the user wishes to express by converting the information input by the user into the initial user intention. And supplementing the user intention to be supplemented, which is related to the user information, in the embedded data to the initial user intention to obtain the user intention to be processed so as to ensure the integrity and the accuracy of the user intention. According to the preset service types, calculating the weighted value of each initial configuration scheme in the set of the initial configuration schemes, sorting each initial configuration scheme according to the weighted value of each initial configuration scheme to obtain a sorting result, determining the configuration scheme of the intention network according to the sorting result, feeding back the configuration scheme of the intention network to a user, facilitating the user to rapidly and accurately configure the intention network, establishing the association relation between the intention of the user to be processed and the configuration information of the intention network, ensuring the configuration accuracy of the intention network, improving the accuracy of intention translation in the intention network and improving the user experience satisfaction.

Fig. 9 is a flowchart of a method for complementing user intent based on buried data in an embodiment of the present application. As shown in fig. 9, the following steps are included.

Step 801, obtaining an initial user intention in user information, and extracting a first intention feature set in the initial user intention.

In user information, the user's intent is often expressed in a manner similar to "what i want to do" or "how i want to do". For example, fig. 10 shows a schematic diagram of intent features obtained by parsing user information in an embodiment of the present application. The user input "I want to open a 10GE service" can be resolved into the Intent features shown in FIG. 10, where domain = { service }, intent (Intent) = { open service }, semantic slot = { person, service type, layer rate, time }. The persona is "wang someplace", the business type is "client layer", the layer rate is "10GE", and the time is "month B and C days a" (e.g., month 9 and 24 days 2020). By looking at fig. 10, the initial user intent in the user information, as well as the first set of intent features (i.e., semantic slots) in the initial user intent, can be clearly obtained.

Step 802, determining an intention feature set to be analyzed according to the first intention feature set and the preset intention feature total set.

For example, the method comprises the steps of determining an intention feature set corresponding to the field = { service }, and intention = { open service }, as a preset intention feature set through searching a configuration file, and then comparing the intention feature (namely { character, service type, layer rate and time }) in a semantic slot with the preset intention feature set to obtain an intention feature set to be analyzed. It should be noted that, the number of the intention features in the preset intention feature set is greater than the number of the intention features in the first intention feature set, for example, the preset intention feature set may include 16 different intention features, and the first intention feature set includes only 4 intention features of people, service types, layer rates and time, so that the number of the intention features in the intention feature set to be analyzed is 12.

Step 803, analyzing the event metadata and the user metadata in the buried point data to obtain buried point information.

The event metadata is used for representing data of behaviors of the user, the user metadata is used for representing data of attribute characteristics of the user, the embedded point information comprises an embedded point intention characteristic set and an embedded point event set, and the embedded point intention characteristic set comprises embedded point intention characteristics.

Step 804, determining whether there is a matching event matching with the preset mapping relationship in the buried point event set.

The preset mapping relation is a mapping relation between a preset event and a preset intention feature. And if the event which is the same as the preset event in the buried point event set is determined, indicating that the matched event which is matched with the preset mapping relation exists in the buried point event set, otherwise, indicating that the matched event which is matched with the preset mapping relation does not exist in the buried point event set.

If it is determined that there is a matching event matching the preset mapping relationship in the buried point event set, step 805 is executed, otherwise, if it is determined that there is no matching event matching the preset mapping relationship in the buried point event set, it is indicated that the initial user intention obtained at this time is sufficient to characterize the desired information of the user, and it is not necessary to add some intention features in the buried point data to the initial user intention, and step 811 is executed to determine the user intention to be processed according to the initial user intention of the user.

Step 805, extracting matching events in the buried point event set, and generating a matching event set.

Step 806, extracting a second intention feature from the buried point intention feature set according to the matching event in the matching event set.

Wherein the second intent feature is a buried point intent feature corresponding to the matching event. For example, the second intent feature may be { source, sink, protection type, degree of intelligence }, or the like intent feature. The intelligent degree indicates the degree of the capacity of dynamically distributing and flexibly controlling bandwidth, rapidly generating service, providing protection and restoration, dynamically expanding capacity and the like in an intention network, and the protection type indicates the protection mode of the service in the intention network, such as linear protection, ring network protection and the like.

Step 807, searching the intention feature set to be analyzed according to the second intention feature, and obtaining a searching result.

Step 808, calculating a feature value corresponding to the second intention feature if it is determined that the search result is that the intention feature identical to the second intention feature exists in the set of intention features to be analyzed.

For example, the second intention feature is calculated by a machine learning algorithm to obtain a feature value corresponding to the second intention feature. Or, through the intention characteristic of the source end and the intention characteristic of the destination end, carrying out data analysis on the corresponding network element configuration event and the link event between the network elements, and determining the characteristic value corresponding to the second intention characteristic. For the intended features for which the feature value cannot be obtained through calculation, data cleaning and discarding are performed.

Step 809, determining the user intention to be supplemented according to the second intention feature and the corresponding feature value.

Step 810, filling the user intention to be supplemented into the initial user intention to obtain the user intention to be processed.

For example, the second intention feature corresponding to the user intention to be supplemented and the corresponding feature value thereof are filled into the semantic slot corresponding to the initial user intention, so that the semantic slot= { character, service type, layer rate, source end, sink end, protection type, intelligence degree and time } corresponding to the user intention to be processed is generated, and the user intention to be processed is determined through each intention feature in the semantic slot corresponding to the user intention to be processed, so that the integrity and accuracy of the user intention are ensured.

It should be noted that, when there is ambiguity in the voice information or text information input by the user (for example, the text information input by the user is "i want to build a protected service"), multiple intention features (for example, linear protection or ring network protection) corresponding to the semantic feature of "protected" are obtained by analyzing the text information input by the user, and through multiple similar intention features, it is not known exactly what type of protected service needs to be built by the user, and also the corresponding service type cannot be known (for example, the user may need to build a traditional protected service or may need to build an intelligent recovery protected service). Therefore, when the text information input by the user is analyzed, and a plurality of similar intention features exist in the obtained semantic slots, detailed analysis is required to be performed on the operation event (for example, the page content browsed by the user and/or the object operated by the user), and the corresponding relationship between the operation event and the intention features is determined, so that the user intention to be supplemented is further determined according to the corresponding relationship.

Step 811, outputting the user intention to be processed.

In the embodiment, the intention feature set to be analyzed is determined by comparing the initial user intention with the preset intention feature complete set, then the buried point intention feature corresponding to the matching event matched with the preset mapping relation in the buried point event set is extracted (namely the second intention feature), the second intention feature is supplemented into the intention feature set corresponding to the initial user intention, the user intention to be processed is generated, and the completeness of the user intention is ensured. According to the user intention to be processed, the configuration scheme of the intention network is determined, and the translation accuracy of the intention network is improved.

FIG. 11 shows a flow diagram of a method of translating pending user intent into a configuration of an intent network in an embodiment of the application. As shown in fig. 11, the following steps are included.

Step 1001, obtaining a user intention to be processed.

Wherein the user intent to be processed includes an initial user intent and a user intent to be supplemented. The completeness of the user intention is guaranteed, and the accurate understanding of the user intention is improved.

Step 1002, determining a set of initial configuration schemes according to the knowledge graph, a preset reasoning algorithm and the user intention to be processed.

For example, the knowledge graph is divided according to the values of domain and Intent in the frame semantics of the user intention to generate a training set and a testing set, the user intention to be processed is used as an input parameter for training to obtain training results, the training results are evaluated and scored according to a scoring function (such as a loss function) to obtain the score of each training result, and then the ranking corresponding to each training result is obtained. From the ranking, a set of initial configuration schemes may be determined.

It should be noted that, the scoring function is used to characterize the association degree between the entity and the relationship in the triplet, and the entity vector and the relationship vector can be continuously adjusted through the scoring function, so that the training result better accords with the intention of the user to be processed.

Step 1003, extracting configuration parameter information of each initial configuration scheme and initial event information corresponding to the configuration parameter information in sequence.

Step 1004, determining target event index information corresponding to the preset service type according to the preset service type.

For example, when the preset service type is service release, the target event index information corresponding to the service release may be a conversion rate of successful service release, and when the preset service type is alarm pressure reduction, the target event index information corresponding to the alarm pressure reduction may be an alarm retention rate. The target event index information corresponding to different preset service types is determined, so that the standard which is specifically expected to be reached can be definitely and specifically achieved, subsequent processing is facilitated, and the processing efficiency is improved.

In step 1005, a weighting value of each initial configuration scheme is calculated according to the target event index information and the initial event information.

For example, in the case that the preset service type is determined to be service release, the conversion rate of successful service release of each initial configuration scheme is determined according to the first target event index information corresponding to service release and each initial event information, and the weighting value of each initial configuration scheme is determined according to the conversion rate of successful service release of each initial configuration scheme.

When domain is a service and the preset service type is a service release, a funnel analysis model is established according to each event in the service release process, a conversion period of service release, preset filtering conditions, a user to be analyzed (for example, an optical transport network (Optical Transport Network, OTN) user) and a funnel analysis algorithm. And analyzing the first target event index information and each initial event information corresponding to the service release according to the funnel analysis model, determining the conversion rate of successful service release of each initial configuration scheme, and further reacting the conversion trend through the conversion rate.

For example, in the case that the preset service type is determined to be the alarm pressure reduction, the alarm retention rate of each initial configuration scheme is determined according to the second target event index information corresponding to the alarm pressure reduction and each initial event information, and the weighting value of each initial configuration scheme is determined according to the alarm retention rate of each initial configuration scheme. The alarm pressure reduction comprises alarm pressure reduction levels, different alarm pressure reduction levels and corresponding alarm retention rates. For example, the alarm pressure reduction levels are divided into a first level, a second level and a third level, and the lower the number of levels, the higher the corresponding alarm retention rate, for example, when the alarm pressure reduction level is the first level, the corresponding alarm retention rate is 30%, when the alarm pressure reduction level is the second level, the corresponding alarm retention rate is 60%, and when the alarm pressure reduction level is the third level, the corresponding alarm retention rate is 75%. I.e. the user's tolerance to alarm pressure reduction is different.

Step 1006, determining the configuration scheme of the intention network according to the weighted value of each initial configuration scheme and the set of initial configuration schemes.

For example, each initial configuration scheme in the set of initial configuration schemes is reordered according to the first ordering result and the weighted value of each initial configuration scheme to obtain a second ordering result, and the configuration scheme of the intention network is determined according to the second ordering result.

In this embodiment, since the static knowledge graph is difficult to describe the distribution and variation of the user intention, in the knowledge reasoning process of the knowledge graph, the operation behaviors of different users and the behavior differences of the same user at different moments cannot reflect the influence of the operation behaviors on the configuration scheme of the intention network. The method comprises the steps of obtaining configuration schemes of an intention network expected to be obtained by a user and the configuration schemes of the intention network actually obtained have deviation, sequentially extracting configuration parameter information of each initial configuration scheme and initial event information corresponding to the configuration parameter information, determining target event index information corresponding to a preset service type according to the preset service type, calculating weighted values of each initial configuration scheme according to the target event index information and the initial event information, and sequencing the initial configuration schemes again according to the weighted values, so that a more accurate configuration scheme of the intention network is obtained. The configuration scheme of the intention network expected to be obtained by the user and the configuration scheme of the intention network actually obtained can be more attached, the configuration scheme of the intention network actually obtained can be better corrected, the correlation accuracy between the intention of the user to be processed and the configuration information of the intention network is improved, the configuration accuracy of the intention network is ensured, the accuracy of intention translation in the intention network is improved, and the user experience satisfaction is improved.

As shown in fig. 12, the electronic device 1100 includes an input device 1101, an input interface 1102, a central processor 1103, a memory 1104, an output interface 1105, and an output device 1106. The input interface 1102, the central processor 1103, the memory 1104, and the output interface 1105 are connected to each other through a bus 1107, and the input device 1101 and the output device 1106 are connected to the bus 1107 through the input interface 1102 and the output interface 1105, respectively, and further connected to other components of the electronic device 1100.

Specifically, the input device 1101 receives input information from the outside and transmits the input information to the central processor 1103 through the input interface 1102, the central processor 1103 processes the input information based on computer-executable instructions stored in the memory 1104 to generate output information, temporarily or permanently stores the output information in the memory 1104, and then transmits the output information to the output device 1106 through the output interface 1105, and the output device 1106 outputs the output information to the outside of the electronic device 1100 for use by a user. The electronic device 1100 may be used to perform the intent processing method described in the above embodiments.

In one embodiment, the electronic device shown in FIG. 12 may be implemented as an intent processing system that may include a memory configured to store a program and a processor configured to execute the program stored in the memory to perform the intent processing method described in the above embodiments.

The foregoing description is only exemplary embodiments of the application and is not intended to limit the scope of the application. In general, the various embodiments of the application may be implemented in hardware or special purpose circuits, software, logic or any combination thereof. For example, some aspects may be implemented in hardware, while other aspects may be implemented in firmware or software which may be executed by a controller, microprocessor or other computing device, although the application is not limited thereto.

Embodiments of the application may be implemented by a data processor of a mobile device executing computer program instructions, e.g. in a processor entity, either in hardware, or in a combination of software and hardware. The computer program instructions may be assembly instructions, instruction Set Architecture (ISA) instructions, machine-related instructions, microcode, firmware instructions, state setting data, or source or object code written in any combination of one or more programming languages.

The block diagrams of any of the logic flows in the figures of this application may represent program steps, or may represent interconnected logic circuits, modules, and functions, or may represent a combination of program steps and logic circuits, modules, and functions. The computer program may be stored on a memory. The memory may be of any type suitable to the local technical environment and may be implemented using any suitable data storage technology, such as, but not limited to, read Only Memory (ROM), random Access Memory (RAM), optical storage devices and systems (digital versatile disk DVD or CD optical disk), etc. The computer readable medium may include a non-transitory storage medium. The data processor may be of any type suitable to the local technical environment, such as, but not limited to, general purpose computers, special purpose computers, microprocessors, digital Signal Processors (DSPs), application Specific Integrated Circuits (ASICs), programmable logic devices (FGPAs), and processors based on a multi-core processor architecture.

The foregoing detailed description of exemplary embodiments of the application has been provided by way of exemplary and non-limiting examples. Various modifications and adaptations to the above embodiments may become apparent to those skilled in the art without departing from the scope of the application, which is defined in the accompanying drawings and claims. Accordingly, the proper scope of the application is to be determined according to the claims.

Claims

1. A method of intent processing, the method comprising:

analyzing the obtained buried point data and initial user intention, and determining the user intention to be processed, wherein the initial user intention is used for representing the original requirement of a user;

determining a configuration scheme of an intention network according to the user intention to be processed;

The analyzing the obtained buried point data and the initial user intention to determine the user intention to be processed comprises the following steps:

extracting a first intention feature set in the initial user intention;

determining an intention feature set to be analyzed according to the first intention feature set and a preset intention feature total set;

determining user intention to be supplemented according to the buried point data and the intention feature set to be analyzed;

filling the user intention to be supplemented into the initial user intention to obtain the user intention to be processed;

determining the user intention to be supplemented according to the buried point data and the intention feature set to be analyzed, including:

Analyzing event metadata and user metadata in the embedded point data to obtain embedded point information, wherein the event metadata are data used for representing behaviors of the user, the user metadata are data used for representing attribute characteristics of the user, the embedded point information comprises an embedded point intention characteristic set and an embedded point event set, and the embedded point intention characteristic set comprises embedded point intention characteristics;

Under the condition that a matching event matched with a preset mapping relation exists in a buried point event set, extracting the matching event in the buried point event set to generate a matching event set, wherein the preset mapping relation is a mapping relation between a preset event and a preset intention feature;

and determining the user intention to be supplemented according to the matching event set, the buried point intention characteristic set and the intention characteristic set to be analyzed.

2. The method of claim 1, wherein the determining the user intent to supplement from the set of matching events, the set of buried point intent features, and the set of intent features to analyze comprises:

Extracting a second intention feature from the buried point intention feature set according to the matching event in the matching event set, wherein the second intention feature is a buried point intention feature corresponding to the matching event;

searching the intention feature set to be analyzed according to the second intention feature to obtain a searching result;

calculating a feature value corresponding to the second intention feature under the condition that the search result is determined to be that the intention feature identical to the second intention feature exists in the intention feature set to be analyzed;

and determining the user intention to be supplemented according to the second intention characteristic and the corresponding characteristic value thereof.

3. The method of claim 1, wherein the populating the initial user intent with the user intent to be supplemented to obtain the user intent to be processed comprises:

extracting a second intention characteristic in the user intention to be supplemented;

Calculating to obtain a feature value corresponding to the second intention feature;

Filling the second intention feature and the corresponding feature value thereof into a first intention feature set in the initial user intention to generate an intention feature set to be processed;

acquiring an event to be processed corresponding to the intention feature in the intention feature set to be processed;

and determining the intention of the user to be processed according to the intention characteristic set to be processed and the event to be processed.

4. The method according to claim 1, wherein said determining a configuration of an intent network based on said pending user intent comprises:

determining a set of initial configuration schemes according to the knowledge graph, a preset reasoning algorithm and the user intention to be processed;

Calculating the weighted value of each initial configuration scheme in the set of initial configuration schemes according to the preset service type;

and determining the configuration scheme of the intention network according to the weighted value of each initial configuration scheme and the set of the initial configuration schemes.

5. The method of claim 4, wherein determining the set of initial configuration schemes based on the knowledge-graph, the preset inference algorithm, and the user intent to be processed comprises:

Generating a set of first configuration schemes based on the knowledge graph and the preset reasoning algorithm, wherein the knowledge graph comprises entity elements, attribute elements and relation elements;

training schemes in the first configuration scheme set according to the user intention to be processed to obtain a training result;

Sorting the training results according to a scoring function to obtain a first sorting result, wherein the scoring function is a function determined according to the degree of association between the entity element and the relation element;

and determining the set of the initial configuration schemes according to the first sequencing result.

6. The method of claim 5, wherein said determining the configuration of the intended network based on the weighted value of each of the initial configuration and the set of initial configuration comprises:

Re-ordering each initial configuration scheme in the set of initial configuration schemes according to the first ordering result and the weighted value of each initial configuration scheme to obtain a second ordering result;

And determining a configuration scheme of the intention network according to the second sequencing result.

7. The method of claim 4, wherein calculating the weighted value of each initial configuration scheme in the set of initial configuration schemes according to the preset service type comprises:

extracting configuration parameter information of each initial configuration scheme and initial event information corresponding to the configuration parameter information in sequence;

determining target event index information corresponding to a preset service type according to the preset service type;

and calculating the weighted value of each initial configuration scheme according to the target event index information and the initial event information.

8. The method of claim 7, wherein calculating the weight value of each initial configuration scheme according to the target event index information and the initial event information comprises:

Under the condition that the preset service type is determined to be service release, determining the successful conversion rate of service release of each initial configuration scheme according to first target event index information corresponding to the service release and each initial event information;

Determining the weighted value of each initial configuration scheme according to the successful conversion rate of the service release of each initial configuration scheme;

Under the condition that the preset service type is determined to be the alarm pressure reduction, determining the alarm retention rate of each initial configuration scheme according to second target event index information corresponding to the alarm pressure reduction and each initial event information;

and determining the weighted value of each initial configuration scheme according to the alarm retention rate of each initial configuration scheme.

9. The method of claim 1, further comprising, prior to the step of analyzing the obtained buried point data and the initial user intent to determine the user intent to process:

Acquiring user information according to an intention starting event identifier and an intention ending event identifier, wherein the user information comprises operation information of the user and input information of the user;

Determining the initial user intention according to the operation information of the user and the input information of the user, wherein the input information comprises intention text and/or voice information, and the initial user intention is a machine-recognizable intention language;

And acquiring the buried point data from the data server according to the intention starting event identification and the buried point port information of the data server.

10. The method according to claim 1, further comprising, after said step of determining a configuration scheme of an intended network in accordance with said pending user intention:

feeding back a configuration scheme of the intention network to the user;

acquiring modification parameters determined by the user based on the configuration scheme of the intention network;

Updating the configuration scheme of the intention network according to the modification parameters and the configuration scheme of the intention network, wherein the configuration scheme of the intention network comprises network configuration parameters.

11. The method of claim 10, further comprising, after the step of updating the configuration scheme of the intended network in accordance with the modification parameters and the configuration scheme of the intended network:

Configuring the intention network according to the updated configuration scheme of the intention network;

and deleting the cached initial user intention and the buried point data under the condition that the intention network configuration is determined to be successful.

12. The method according to any one of claims 1 to 11, wherein the buried point data is data extracted from log buried point information.

13. An intent processing apparatus, characterized in that it comprises:

The analysis module is used for analyzing the obtained buried point data and the initial user intention, and determining the user intention to be processed, wherein the initial user intention is used for representing the original requirement of the user;

the processing module is used for determining a configuration scheme of an intention network according to the user intention to be processed;

The analysis module comprises an extraction submodule, a first determination submodule, a second determination submodule, a filling submodule and a processing submodule, wherein the extraction submodule is used for extracting a first intention feature set in initial user intention, the first determination submodule is used for determining an intention feature set to be analyzed according to the first intention feature set and a preset intention feature complete set, and the second determination submodule is used for determining user intention to be supplemented according to buried point data and the intention feature set to be analyzed;

The system comprises a first determining sub-module, a second determining sub-module and a third determining sub-module, wherein the first determining sub-module is used for analyzing event metadata and user metadata in the embedded point data to obtain embedded point information, the event metadata is data used for representing the behavior of a user, the user metadata is data used for representing the attribute characteristics of the user, the embedded point information comprises an embedded point intention characteristic set and an embedded point event set, the embedded point intention characteristic set comprises embedded point intention characteristics, under the condition that a matching event matched with a preset mapping relation exists in the embedded point event set is determined, the matching event in the embedded point event set is extracted to generate a matching event set, the preset mapping relation is the mapping relation between the preset event and the preset intention characteristic, and the user intention to be supplemented is determined according to the matching event set, the embedded point intention characteristic set and the intention characteristic set to be analyzed.

14. An electronic device, comprising:

One or more processors;

a memory having one or more programs stored thereon, which when executed by the one or more processors, cause the one or more processors to implement the intended processing method of any of claims 1-12.

15. A readable storage medium, characterized in that the readable storage medium stores a computer program which, when executed by a processor, implements the intended processing method according to any one of claims 1-12.