CN112905902A

CN112905902A - Cloud computing and artificial intelligence oriented session intention extraction method and system

Info

Publication number: CN112905902A
Application number: CN202110364193.9A
Authority: CN
Inventors: 卞美玲
Original assignee: Individual
Current assignee: Individual
Priority date: 2020-09-11
Filing date: 2020-09-11
Publication date: 2021-06-04
Also published as: CN112199580B; CN112905901A; CN112199580A

Abstract

The embodiment of the application provides a session intention extraction method and system facing cloud computing and artificial intelligence, subscription content nodes of task subscription items corresponding to tasks issued by the cloud computing under target reference interactive services corresponding to periodically collected data sets are considered, then subscription content nodes under each target reference interactive service are classified based on preset target content label classification, so that the difference between different target reference interactive services and the target content label classification is considered, artificial intelligence analysis is performed on the subscription associated intention data set classified by each target content label based on the request subscription content characteristics corresponding to the target content label classification, the accuracy of the artificial intelligence analysis can be improved in a decision mode, and the artificial intelligence analysis result can be matched with an actual service scene better.

Description

Cloud computing and artificial intelligence oriented session intention extraction method and system

Technical Field

The application relates to the technical field of cloud computing and big data positioning, in particular to a session intention extraction method and system oriented to cloud computing and artificial intelligence.

Background

With the rapid development of cloud computing and big data technology, the application range of the cloud computing and big data technology is wider and wider, corresponding cloud computing issued tasks are distributed for each business processing terminal through the strong cloud computing capability of the cloud artificial intelligence platform, so that the business processing terminals are instructed to upload big data business information needing cloud computing analysis, the big data business information uploaded by the business processing terminals is analyzed, the intention development rules of a large number of users are recognized, and the follow-up business service updating and product technology research and development are facilitated.

For the big data service track information of each service processing terminal, the interest items of the user using the service processing terminal in using the application internet service can be reflected to a certain extent, and if the interest items can be accurately analyzed, various related subscription services can be recommended for the related users, so that the application range of various online data services is expanded, which also becomes a technical problem to be solved in the field.

Disclosure of Invention

In order to overcome at least the above-mentioned deficiencies in the prior art, the present application aims to provide a session intention extraction method and system facing cloud computing and artificial intelligence, by analyzing an intention potential point of big data service track information of each service processing terminal and determining a service interest item set corresponding to the intention potential point according to historical session intention information of the intention potential point, subscription services corresponding to the subscription service interest item set can be requested from service processing equipment, the application range of various on-line data services is improved, and further, a cloud computing issued task issued next to the service processing equipment can be updated according to a target subscription service selected by the service processing equipment from the subscription services corresponding to the service interest item set, so that the matching degree of the analyzed big data service track information and the actual intention of a user can be improved by a closed-loop feedback manner, thereby continuously optimizing the subsequently recommended subscription services.

In a first aspect, the present application provides a cloud computing and artificial intelligence oriented session intention extraction method, which is applied to an artificial intelligence platform, where the artificial intelligence platform is in communication connection with a plurality of service processing devices, and the method includes:

receiving big data service track information aiming at a cloud computing issued task and sent by the service processing equipment, and extracting target intention labeling information of the big data service track information; wherein the target intention marking information comprises target intention potential points;

determining a service interest item set corresponding to the target intention latent point according to historical conversation intention information of the target intention latent point, wherein the historical conversation intention information is obtained by the artificial intelligence platform performing artificial intelligence analysis on a periodically collected data set generated by business processing equipment in a subscription application scene and a target reference interaction service corresponding to the periodically collected data set;

requesting the service processing equipment to subscribe the subscription service corresponding to the service interest item set;

and updating the cloud computing issuing task issued to the service processing equipment next time according to the target subscription service selected by the service processing equipment from the subscription services corresponding to the service interest item set.

In a possible implementation manner of the first aspect, the determining, according to the historical conversation intention information of the target intention potential point, a service interest item set corresponding to the target intention potential point includes:

acquiring reference problem description representation and non-text factor information of the reference problem description representation from historical conversation intention information of the target intention potential point, wherein the non-text factor information represents a service item attribute state corresponding to each description vector combination in the reference problem description representation;

processing the reference problem description representation according to the non-text factor information to generate response multi-valued attribute information of the reference problem description representation;

extracting the point of interest scene parameters of the reference problem description representation and the response multivalued attribute information, and determining a second point of interest scene vector set corresponding to the first point of interest scene vector set corresponding to the response multivalued attribute information from the extracted current point of interest scene parameter information;

performing feature fusion on the first point of interest scene vector set and the second point of interest scene vector set to obtain a third point of interest scene vector set;

and outputting a target service interest item set corresponding to the reference problem description representation according to the third interest point scene vector set.

In a possible implementation manner of the first aspect, the step of processing the reference question description representation according to the non-text factor information and generating response multi-value attribute information of the reference question description representation includes:

extracting interest point scene parameters of the reference problem description representation, carrying out scene dynamic track identification on first interest point scene parameters corresponding to the obtained reference problem description representation, and obtaining a first scene interaction topic set corresponding to the reference problem description representation according to the identified scene dynamic track;

extracting interest point scene parameters of the non-text factor information, identifying scene dynamic tracks of second interest point scene parameters corresponding to the obtained non-text factor information, and obtaining a second scene interaction topic set corresponding to the non-text factor information according to the identified scene dynamic tracks;

acquiring first topic evolution distribution information stored in the first scene interaction topic set, and converting the first topic evolution distribution information into a corresponding first topic evolution distribution vector;

acquiring second topic evolution distribution information respectively stored by a plurality of scene interaction topic objects in the second scene interaction topic set, and converting each second topic evolution distribution information into a corresponding second topic evolution distribution vector;

calculating a fusion distribution vector of each second topic evolution distribution vector and the first topic evolution distribution vector;

sequencing the fusion distribution vectors corresponding to each second topic evolution distribution vector, and selecting a plurality of similar topic evolution distribution vectors from the second topic evolution distribution vectors according to the sequencing result;

performing particle swarm algorithm processing on the evolution distribution vectors of the similar topics to obtain particle swarm characteristic vectors;

performing Gaussian probability density calculation on the topic feature vectors of the first scenario interaction topic set and the second scenario interaction topic set, and obtaining topic characterization parameter vectors according to the calculated Gaussian probability density; the topic representation parameter vector comprises influence parameters corresponding to all scenario interaction topic objects in the second scenario interaction topic set;

calculating a fused feature vector of the particle swarm feature vector and the topic representation parameter vector, and using a calculated result as topic evolution scene description of the first topic evolution distribution information;

mining the topic evolution scene description hidden Markov to a reference problem representation fragment set in the reference problem description representation to obtain initial topic hidden Markov conversation intention information;

carrying out scene dynamic track identification on the initial topic hidden Markov conversation intention information to obtain a reference scene dynamic track;

and obtaining the response multi-valued attribute information corresponding to the reference question description representation according to the first scenario interaction topic set, the second scenario interaction topic set and the reference scenario dynamic trajectory.

In a possible implementation manner of the first aspect, the step of obtaining the response multivalued attribute information corresponding to the reference question description according to the first scenario interaction topic set, the second scenario interaction topic set, and the reference scenario dynamic trajectory includes:

mapping the first scenario interaction topic set and the second scenario interaction topic set to each dynamic track node in the reference scenario dynamic track respectively to obtain mapping attribute information of each dynamic track node corresponding to the first scenario interaction topic set and the second scenario interaction topic set respectively;

and summarizing the mapping attribute information of each dynamic track node corresponding to the first scenario interaction topic set and the second scenario interaction topic set respectively to obtain the response multi-valued attribute information corresponding to the reference problem description representation.

In a possible implementation manner of the first aspect, the step of extracting the point of interest contextual parameters from the reference problem description representation and the response multi-valued attribute information, and determining a second point of interest contextual vector set corresponding to the first point of interest contextual vector set corresponding to the response multi-valued attribute information from the extracted current point of interest contextual parameter information includes:

extracting the point of interest contextual parameters of the reference question description representation and the response multi-valued attribute information to obtain current point of interest contextual parameter information mapped in the point of interest contextual parameters of the reference question description representation and the response multi-valued attribute information; the current interest point scene parameter information comprises scene description information of a plurality of interest point scene elements;

determining similar scene description information of the first point of interest scene vector set from the scene description information of the plurality of point of interest scene elements contained in the current point of interest scene parameter information, and taking the similar scene description information as the second point of interest scene vector set.

In a possible implementation manner of the first aspect, the step of performing feature fusion on the first point of interest scene vector set and the second point of interest scene vector set to obtain a third point of interest scene vector set includes:

respectively inputting the first interest point scene vector set and the second interest point scene vector set into a preset artificial intelligence analysis model, so that the artificial intelligence analysis model respectively outputs decision interest point scene vector sets of the first interest point scene vector set and the second interest point scene vector set, and a first target interest point scene vector set and a second target interest point scene vector set are obtained;

performing particle swarm algorithm calculation on the first target interest point scene vector set to obtain first subject particle swarm calculation information; extracting the interest point scene parameters of the first target interest point scene vector set, performing particle swarm algorithm calculation on the extracted interest point scene vector set to obtain second subject particle swarm calculation information, calculating fusion calculation information of the first subject particle swarm calculation information and the second subject particle swarm calculation information, and obtaining a first decision scene vector set corresponding to the first target interest point scene vector set;

performing particle swarm algorithm calculation on the second target interest point scene vector set to obtain third subject particle swarm calculation information; extracting the interest point scene parameters of the second target interest point scene vector set, performing particle swarm algorithm calculation on the extracted interest point scene vector set to obtain fourth subject particle swarm calculation information, and calculating fusion calculation information of the third subject particle swarm calculation information and the fourth subject particle swarm calculation information to obtain a second decision scene vector set corresponding to the second target interest point scene vector set;

and calculating a fusion feature set of the first decision scene vector set and the second decision scene vector set, and taking the obtained fusion feature set as the third interest point scene vector set.

In a possible implementation manner of the first aspect, the step of outputting the set of target service interest items corresponding to the reference problem description representation according to the third set of point of interest context vectors includes:

acquiring a plurality of interest point scene description vectors in the third interest point scene vector set and an artificial intelligence analysis strategy corresponding to each interest point scene description vector combination, wherein the interest point scene description vectors comprise a first interest point scene description vector and a second interest point scene description vector, and the first interest point scene description vector and the second interest point scene description vector are associated with each other through mapping;

performing hidden Markov excavation on the third interest point scene vector set to output a first reference hidden Markov excavation object corresponding to each first interest point scene description vector combination and a target hidden Markov excavation object corresponding to the third interest point scene vector set;

calculating the degree of association between the target hidden Markov mining object and each first reference hidden Markov mining object to obtain the hidden Markov mining object similarity parameter between each corresponding first interest point scene description vector and the third interest point scene vector set;

identifying all scene service data segments in each first interest point scene description vector, and calculating the Gaussian probability density corresponding to each first interest point scene description vector;

generating a scene service distribution map of corresponding interest point scene description vectors according to all the scene service data segments and the Gaussian probability density;

generating a distribution map label corresponding to each interest point scene description vector combination according to the artificial intelligence analysis strategy;

obtaining a first scene cluster of each interest point scene description vector by using each scene service distribution map corresponding to the distribution map label;

clustering according to the first reference hidden Markov mining objects corresponding to each interest point scene description vector combination of the second interest point scene description vector combination to obtain a second scene cluster of each interest point scene description vector;

determining a target interest point scene description vector from the plurality of interest point scene description vectors according to the first scene cluster and the second scene cluster;

and taking a first reference hidden Markov excavation object corresponding to the target interest point scene description vector combination as a second hidden Markov excavation object, and performing hidden Markov excavation on the second hidden Markov excavation object to the third interest point scene vector set so as to output a target service interest item set corresponding to the reference problem description representation.

In a possible implementation manner of the first aspect, the artificial intelligence analysis policy includes a vector structure clustering policy on the interest point scenario description vector and a feature value clustering policy corresponding to the interest point scenario description vector combination;

the step of generating a distribution map label corresponding to each point of interest scene description vector combination according to the artificial intelligence analysis strategy comprises the following steps:

carrying out characteristic value clustering on the interest point scene description vectors according to the characteristic value clustering strategy to generate service interest item set characteristic values corresponding to the interest point scene description vector combinations;

carrying out vector structure clustering on the interest point scene description vectors according to the vector structure clustering strategy to generate vector structure clustering results corresponding to the interest point scene description vector combination;

and generating a distribution map label corresponding to the interest point scene description vector according to the service interest item set characteristic value and the vector structure clustering result.

In a possible implementation manner of the first aspect, the method further includes:

acquiring historical conversation intention information of the target intention potential point;

the step of obtaining the historical session intention information of the target intention potential point comprises the following steps:

acquiring a periodic collection data set generated by the service processing equipment in a subscription application scene and a target reference interactive service corresponding to the periodic collection data set from the service processing equipment;

acquiring subscription content nodes of task subscription items corresponding to the cloud computing issued tasks under the target reference interactive services, classifying the subscription content nodes under the target reference interactive services according to predetermined target content label classifications, and respectively generating a subscription content node set classified by each target content label;

for each target content label classification, acquiring subscription associated intention data of each subscription content node in a subscription content node set classified by the target content label, wherein the subscription associated intention data is matched with the periodically collected data set, and carrying out artificial intelligent analysis on the subscription associated intention data set classified by each target content label based on the subscription content requesting characteristics corresponding to the target content label classification to obtain historical conversation intention information classified by each target content label;

and obtaining the historical conversation intention information of the marked target content label classification included in the target intention potential point from the historical conversation intention information of each target content label classification.

In a second aspect, an embodiment of the present application further provides a cloud computing and artificial intelligence oriented session intention extraction apparatus, which is applied to an artificial intelligence platform, where the artificial intelligence platform is in communication connection with a plurality of service processing devices, and the apparatus includes:

the receiving module is used for receiving big data service track information aiming at a cloud computing issued task and sent by the service processing equipment and extracting target intention labeling information of the big data service track information; wherein the target intention marking information comprises target intention potential points;

a determining module, configured to determine, according to historical conversation intention information of the target intention potential point, a service interest item set corresponding to the target intention potential point, where the historical conversation intention information is obtained by performing artificial intelligence analysis on a periodically collected data set generated by a service processing device in a subscription application scene and a target reference interaction service corresponding to the periodically collected data set by the artificial intelligence platform;

a subscription request module, configured to request the service processing device to subscribe to a subscription service corresponding to the service interest item set;

and the updating module is used for updating the cloud computing issuing task issued to the service processing equipment next time according to the target subscription service selected by the service processing equipment from the subscription services corresponding to the service interest item set.

In a third aspect, an embodiment of the present application further provides a cloud computing and artificial intelligence oriented session intention extraction system, where the cloud computing and artificial intelligence oriented session intention extraction system includes an artificial intelligence platform and a plurality of business processing devices communicatively connected to the artificial intelligence platform;

the artificial intelligence platform is used for receiving big data service track information aiming at a cloud computing issued task and sent by the service processing equipment and extracting target intention labeling information of the big data service track information; wherein the target intention marking information comprises target intention potential points;

the artificial intelligence platform is used for determining a service interest item set corresponding to the target intention latent point according to historical conversation intention information of the target intention latent point, wherein the historical conversation intention information is obtained by performing artificial intelligence analysis on a periodically collected data set generated by business processing equipment in a subscription application scene and a target reference interaction service corresponding to the periodically collected data set by the artificial intelligence platform;

the artificial intelligence platform is used for requesting the business processing equipment to subscribe the subscription service corresponding to the service interest item set;

and the artificial intelligence platform is used for updating a cloud computing issuing task issued to the service processing equipment next time according to a target subscription service selected by the service processing equipment from the subscription services corresponding to the service interest item set.

In a fourth aspect, an embodiment of the present application further provides an artificial intelligence platform, where the artificial intelligence platform includes a processor, a machine-readable storage medium, and a network interface, where the machine-readable storage medium, the network interface, and the processor are connected through a bus system, the network interface is configured to be in communication connection with at least one business processing device, the machine-readable storage medium is configured to store a program, an instruction, or a code, and the processor is configured to execute the program, the instruction, or the code in the machine-readable storage medium to perform the cloud computing and artificial intelligence oriented session intention extraction method in the first aspect or any one of the possible implementations of the first aspect.

In a fifth aspect, an embodiment of the present application provides a computer-readable storage medium, where instructions are stored in the computer-readable storage medium, and when the instructions are executed, the computer executes the cloud computing-oriented and artificial intelligence-oriented session intention extraction method in the first aspect or any one of the possible implementation manners of the first aspect.

Based on any one of the above aspects, the application analyzes the intention potential point of the big data service track information of each service processing terminal, and determining a service interest item set corresponding to the target intention latent point according to the historical conversation intention information of the target intention latent point, therefore, the subscription service corresponding to the subscription service interest item set can be requested from the business processing equipment, the application range of various online data services is improved, and can further select a target subscription service from the subscription services corresponding to the service interest item set according to the service processing device, and updating the cloud computing issued task issued to the service processing equipment next time, so that the matching degree of the analyzed big data service track information and the actual intention of the user can be improved in a closed-loop feedback mode, and the follow-up recommended subscription service is continuously optimized.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that need to be called in the embodiments are briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained according to the drawings without inventive efforts.

Fig. 1 is a schematic view of an application scenario of a session intention extraction system for cloud computing and artificial intelligence provided in an embodiment of the present application;

fig. 2 is a schematic flowchart of a session intention extraction method for cloud computing and artificial intelligence according to an embodiment of the present application;

fig. 3 is a schematic functional module diagram of a session intention extraction apparatus for cloud computing and artificial intelligence provided in an embodiment of the present application;

fig. 4 is a schematic block diagram of structural components of an artificial intelligence platform for implementing the cloud computing and artificial intelligence oriented session intention extraction method according to the embodiment of the present application.

Detailed Description

The present application will now be described in detail with reference to the drawings, and the specific operations in the method embodiments may also be applied to the apparatus embodiments or the system embodiments.

Fig. 1 is an interaction diagram of a cloud computing and artificial intelligence oriented session intention extraction system 10 according to an embodiment of the present application. The cloud computing and artificial intelligence oriented session intention extraction system 10 may include an artificial intelligence platform 100 and a business processing device 200 communicatively coupled to the artificial intelligence platform 100. The cloud computing and artificial intelligence oriented session intent extraction system 10 shown in fig. 1 is only one possible example, and in other possible embodiments, the cloud computing and artificial intelligence oriented session intent extraction system 10 may also include only a portion of the components shown in fig. 1 or may also include other components.

In this embodiment, the business processing device 200 may comprise a mobile device, a tablet computer, a laptop computer, etc., or any combination thereof. In some embodiments, the mobile device may include an internet of things device, a wearable device, a smart mobile device, a virtual reality device, an augmented reality device, or the like, or any combination thereof. In some embodiments, the internet of things device may include a control device of a smart appliance device, a smart monitoring device, a smart television, a smart camera, and the like, or any combination thereof. In some embodiments, the wearable device may include a smart bracelet, a smart lace, smart glass, a smart helmet, a smart watch, a smart garment, a smart backpack, a smart accessory, or the like, or any combination thereof. In some embodiments, the smart mobile device may include a smartphone, a personal digital assistant, a gaming device, and the like, or any combination thereof. In some embodiments, the virtual reality device and the augmented reality device may include a virtual reality helmet, virtual reality glass, a virtual reality patch, an augmented reality helmet, augmented reality glass, an augmented reality patch, or the like, or any combination thereof. For example, virtual reality devices and augmented reality devices may include various virtual reality products and the like.

In this embodiment, the artificial intelligence platform 100 and the business processing device 200 in the cloud computing and artificial intelligence oriented session intention extraction system 10 may cooperatively perform the cloud computing and artificial intelligence oriented session intention extraction method described in the following method embodiment, and the following detailed description of the method embodiment may be referred to for the specific steps performed by the artificial intelligence platform 100 and the business processing device 200.

Based on the inventive concept of the technical solution provided by the present application, the artificial intelligence platform 100 provided by the present application can be applied to scenes such as smart medical care, smart city management, smart industrial internet, general service monitoring management, etc. in which a big data technology or a cloud computing technology is applied, and for example, can also be applied to scenes such as but not limited to new energy automobile system management, smart cloud office, cloud platform data processing, cloud game data processing, cloud live broadcast processing, cloud automobile management platform, block chain financial data service platform, etc., but not limited thereto.

In order to solve the technical problem in the foregoing background, fig. 2 is a flowchart illustrating a cloud computing and artificial intelligence oriented session intention extraction method provided in an embodiment of the present application, where the cloud computing and artificial intelligence oriented session intention extraction method provided in this embodiment may be executed by the artificial intelligence platform 100 shown in fig. 1, and the cloud computing and artificial intelligence oriented session intention extraction method is described in detail below.

Step S110 is to receive the big data service trajectory information for the cloud computing issued task sent by the service processing device 200, and extract the target intention labeling information of the big data service trajectory information.

And step S120, determining a service interest item set corresponding to the target intention potential point according to the historical conversation intention information of the target intention potential point.

Step S130, requesting the service processing device 200 for a subscription service corresponding to the subscription service interest item set.

In this embodiment, the target intention tagging information may include, for example, a target intention potential point, which may be used to represent a data segment in which there is an intention tendency for the target service node. For example, in the process of extracting the target intention labeling information of the big data service track information, a data segment with a service track duration longer than a preset time period in the big data service track information may be extracted as the target intention labeling information.

In this embodiment, the historical session intention information may be obtained by performing artificial intelligence analysis on a periodically collected data set generated by the service processing device 200 in a subscription application scenario and a target reference interaction service corresponding to the periodically collected data set by the artificial intelligence platform. The periodic data collection may refer to that, for each service usage process, a plurality of service statistical units are usually included, and may include, for example and without limitation, a service click statistical unit, a service feedback statistical unit, and the like. The target reference interactive service may be used to represent a scene service acquired in a specific service interaction process, such as a video scene service, an audio scene service, an online chat scene service, and the like.

Based on the above steps, the present embodiment analyzes the intention potential point of the big data service track information of each service processing terminal, and determining a service interest item set corresponding to the target intention latent point according to the historical conversation intention information of the target intention latent point, therefore, the subscription service corresponding to the subscription service interest item set can be requested from the business processing equipment, the application range of various online data services is improved, and can further select a target subscription service from the subscription services corresponding to the service interest item set according to the service processing device, and updating the cloud computing issued task issued to the service processing equipment next time, so that the matching degree of the analyzed big data service track information and the actual intention of the user can be improved in a closed-loop feedback mode, and the follow-up recommended subscription service is continuously optimized.

In one possible implementation, step S120 may be implemented by the following exemplary substeps, which are described in detail below.

And a substep S121, obtaining the reference problem description representation and the non-text factor information of the reference problem description representation from the historical conversation intention information of the target intention potential point, wherein the non-text factor information can represent the service item attribute state corresponding to each description vector combination in the reference problem description representation.

And a substep S122, processing the reference question description representation according to the non-text factor information, and generating response multi-valued attribute information of the reference question description representation.

And a substep S123 of extracting the point of interest contextual parameters from the reference question description representation and the response multi-valued attribute information, and determining a second point of interest contextual vector set corresponding to the first point of interest contextual vector set corresponding to the response multi-valued attribute information from the extracted current point of interest contextual parameter information.

And a substep S124, performing feature fusion on the first interest point scene vector set and the second interest point scene vector set to obtain a third interest point scene vector set.

And a substep S125, outputting a set of target service interest items corresponding to the reference problem description representation according to the third point of interest scene vector set.

Exemplarily, in the sub-step S122, it can be realized by the following specific embodiments.

(1) Extracting interest point scene parameters from the reference problem description representation, carrying out scene dynamic track identification on first interest point scene parameters corresponding to the obtained reference problem description representation, and obtaining a first scene interaction topic set corresponding to the reference problem description representation according to the identified scene dynamic track.

(2) And extracting the interest point scene parameters of the non-text factor information, identifying scene dynamic tracks of second interest point scene parameters corresponding to the obtained non-text factor information, and obtaining a second scene interaction topic set corresponding to the non-text factor information according to the identified scene dynamic tracks.

(3) Obtaining first topic evolution distribution information stored in the first scene interaction topic set, and converting the first topic evolution distribution information into a corresponding first topic evolution distribution vector.

(4) Second topic evolution distribution information stored by a plurality of scene interaction topic objects in a second scene interaction topic set is obtained, and each second topic evolution distribution information is converted into a corresponding second topic evolution distribution vector.

(5) And calculating a fusion distribution vector of each second topic evolution distribution vector and the first topic evolution distribution vector.

(6) And sequencing the fusion distribution vectors corresponding to each second topic evolution distribution vector, and selecting a plurality of similar topic evolution distribution vectors from the plurality of second topic evolution distribution vectors according to the sequencing result.

(7) And carrying out particle swarm algorithm processing on the evolution distribution vectors of the similar topics to obtain particle swarm characteristic vectors.

(8) And performing Gaussian probability density calculation on the topic feature vectors of the first scene interactive topic set and the second scene interactive topic set, and obtaining topic characterization parameter vectors according to the Gaussian probability density obtained by calculation. The topic representation parameter vector contains influence parameters corresponding to each scene interaction topic object in the second scene interaction topic set.

(9) And calculating a fused feature vector of the particle swarm feature vector and the topic representation parameter vector, and using the calculated result as topic evolution scene description of the first topic evolution distribution information.

(10) And mining the topic evolution scene description hidden Markov to a reference problem representation fragment set in the reference problem description representation to obtain initial topic hidden Markov conversation intention information.

(11) And carrying out scene dynamic track identification on the initial topic hidden Markov conversation intention information to obtain a reference scene dynamic track.

(12) And obtaining corresponding response multi-valued attribute information represented by the reference question description according to the first scene interactive topic set, the second scene interactive topic set and the reference scene dynamic track.

Exemplarily, in the sub-step S123, it can be realized by the following specific embodiments.

(1) And extracting the contextual parameters of the interest points from the reference question description representation and the response multi-valued attribute information to obtain the current contextual parameter information of the interest points mapped in the contextual parameters of the reference question description representation and the response multi-valued attribute information.

In this embodiment, the current point of interest context parameter information includes context description information of a plurality of point of interest context elements.

(2) And determining similar scene description information of the first point of interest scene vector set from the scene description information of the plurality of point of interest scene elements contained in the current point of interest scene parameter information, and taking the similar scene description information as a second point of interest scene vector set.

Exemplarily, in the sub-step S124, it can be realized by the following specific embodiments.

(1) And respectively inputting the first interest point scene vector set and the second interest point scene vector set into a preset artificial intelligence analysis model, so that the artificial intelligence analysis model respectively outputs decision interest point scene vector sets of the first interest point scene vector set and the second interest point scene vector set, and a first target interest point scene vector set and a second target interest point scene vector set are obtained.

(2) And performing particle swarm algorithm calculation on the first target interest point scene vector set to obtain first subject particle swarm calculation information. Extracting the interest point scene parameters of the first target interest point scene vector set, performing particle swarm algorithm calculation on the extracted interest point scene vector set to obtain second subject particle swarm calculation information, calculating fusion calculation information of the first subject particle swarm calculation information and the second subject particle swarm calculation information, and obtaining a first decision scene vector set corresponding to the first target interest point scene vector set.

(3) And performing particle swarm algorithm calculation on the second target interest point scene vector set to obtain third subject particle swarm calculation information. And extracting the interest point scene parameters of the second target interest point scene vector set, performing particle swarm algorithm calculation on the extracted interest point scene vector set to obtain fourth subject particle swarm calculation information, calculating fusion calculation information of the third subject particle swarm calculation information and the fourth subject particle swarm calculation information, and obtaining a second decision scene vector set corresponding to the second target interest point scene vector set.

(4) And calculating a fusion feature set of the first decision scene vector set and the second decision scene vector set, and taking the obtained fusion feature set as a third interest point scene vector set.

Exemplarily, in the sub-step S125, it can be realized by the following specific embodiments.

(1) And acquiring a plurality of interest point scene description vectors in a third interest point scene vector set and an artificial intelligence analysis strategy corresponding to each interest point scene description vector combination, wherein the interest point scene description vectors comprise a first interest point scene description vector and a second interest point scene description vector, and the first interest point scene description vector and the second interest point scene description vector are associated with each other in a mapping manner.

(2) And performing hidden Markov excavation on the third interest point scene vector set to output a first reference hidden Markov excavation object corresponding to each first interest point scene description vector combination and a target hidden Markov excavation object corresponding to the third interest point scene vector set.

(3) And calculating the relevance between the target hidden Markov mining object and each first reference hidden Markov mining object to obtain the hidden Markov mining object similarity parameter between each corresponding first interest point scene description vector and the third interest point scene vector set.

(4) Identifying all scene service data segments in each first interest point scene description vector, and calculating the respective corresponding Gaussian probability density of each first interest point scene description vector.

(5) And generating a scene service distribution map of the corresponding interest point scene description vector according to all the scene service data segments and the Gaussian probability density.

(6) And generating a distribution map label corresponding to each point of interest scene description vector combination according to an artificial intelligence analysis strategy.

As an example, the artificial intelligence analysis policy may include a vector structure clustering policy for the interest point scenario description vectors, and a feature value clustering policy corresponding to a combination of the interest point scenario description vectors. Based on the method, the interest point scene description vectors can be subjected to characteristic value clustering according to a characteristic value clustering strategy to generate service interest item set characteristic values corresponding to the interest point scene description vector combination, then the interest point scene description vectors are subjected to vector structure clustering according to a vector structure clustering strategy to generate vector structure clustering results corresponding to the interest point scene description vector combination, and therefore distribution map labels corresponding to the interest point scene description vectors can be generated according to the service interest item set characteristic values and the vector structure clustering results.

(7) And obtaining respective first scene clusters of the scene description vectors of each interest point by using each scene service distribution map corresponding to the distribution map labels.

(8) And clustering according to the first reference hidden Markov mining objects corresponding to each interest point scene description vector combination of the second interest point scene description vector combination to obtain a second scene cluster of each interest point scene description vector.

(9) And determining a target interest point scene description vector from the plurality of interest point scene description vectors according to the first scene cluster and the second scene cluster.

(10) And taking the first reference hidden Markov excavation object corresponding to the target interest point scene description vector combination as a second hidden Markov excavation object, and excavating the second hidden Markov excavation object to a third interest point scene vector set so as to output a target service interest item set corresponding to the reference problem description representation.

Exemplarily, the interest point scene description vector further includes a third interest point scene description vector, before (7), a hidden markov mining translation parameter of each first reference hidden markov mining object relative to each reference hidden markov mining object in a reference hidden markov mining object set corresponding to the third interest point scene description vector may be further calculated, and then all the hidden markov mining translation parameters corresponding to each first reference hidden markov mining object are fused to obtain a fused hidden markov mining parameter corresponding to each first reference hidden markov mining object.

On the reference, all the first reference hidden markov excavation objects may be arranged in sequence according to the fusion hidden markov excavation parameters corresponding to each first reference hidden markov excavation object, the respective priority parameter of each first reference hidden markov excavation object is determined according to the sequence of the arrangement of each first reference hidden markov excavation object, and then the fusion hidden markov excavation parameter corresponding to each first reference hidden markov excavation object is processed according to the respective priority parameter of each first reference hidden markov excavation object, so as to generate the weighted fusion hidden markov excavation parameter of each interest point context description vector.

Thus, in step (7), each scene service distribution map corresponding to the distribution map label can be used to cluster the weighted fusion hidden markov mining parameters of each interest point scene description vector, so as to obtain a first scene cluster corresponding to each interest point scene description vector combination.

In a possible implementation manner, before step S120, the cloud computing and artificial intelligence oriented session intention extraction method provided by this embodiment may further include step S101 of obtaining historical session intention information of the target intention potential.

For example, step S101 may be specifically realized by the following substeps.

In sub-step S1011, a periodic collected data set generated by each service processing device 200 in a subscription application scenario and a target reference interactive service corresponding to the periodic collected data set are obtained from each service processing device 200.

Step S1012, obtaining the subscription content nodes of the task subscription items corresponding to the cloud computing issued tasks under the target reference interactive services, classifying the subscription content nodes under each target reference interactive service according to the predetermined target content label classification, and generating a subscription content node set classified by each target content label.

The subscription content node may be configured to represent a content statistical tag of a specific application of a task subscription item corresponding to the cloud-computing task issuing under each target reference interactive service for periodically collecting data, for example, a smart voice content statistical tag, an online shopping content statistical tag, and the like.

In this embodiment, the predetermined target content tag classification may be flexibly selected according to actual design requirements, and is mainly used for representing a subscription request subscription selection menu provided for different users, which is not limited in detail herein.

Step S1013, for each target content label classification, obtaining subscription associated intention data of each subscription content node in the subscription content node set classified by the target content label matching with the periodically collected data set, and performing artificial intelligence analysis on the subscription associated intention data set classified by each target content label based on the request subscription content characteristics corresponding to the target content label classification to obtain historical conversation intention information classified by each target content label.

Step 1014, obtaining the historical conversation intention information classified by the marked target content label included in the target intention potential point from the historical conversation intention information classified by each target content label.

Based on the above steps, in this embodiment, subscription content nodes of task subscription items corresponding to the cloud-computing issued tasks under the target reference interactive services corresponding to the periodically collected data sets are considered, and then the subscription content nodes under each target reference interactive service are classified based on the predetermined target content label classification, so that differences between different target reference interactive services and the target content label classification are considered, and therefore, the subscription association intention data set classified by each target content label is subjected to artificial intelligence analysis based on the subscription request content characteristics corresponding to the target content label classification, so that the accuracy of the artificial intelligence analysis can be improved, and the artificial intelligence analysis result can be better matched with an actual service scene.

In one possible implementation, for example, with respect to step S1013, in the process of obtaining the subscription-related intention data that each subscription content node in the set of subscription content nodes categorized by the target content tag matches with the periodic collection data set, the following exemplary sub-steps may be further implemented, which are described in detail below.

And a substep S10131, obtaining the subscription content construction characteristics related to each subscription content node in the subscription content node set classified by the target content label.

And a substep S10132 of matching corresponding target data areas from the periodically collected data set according to the subscription content construction characteristics related to each subscription content node.

And a substep S10133, constructing a subscription associated intention attribute corresponding to each service record plate in the feature-matched target data area according to the subscription content related to each subscription content node, and determining that each subscription content node in the subscription content node set classified by the target content label is matched with the subscription associated intention data of the periodically collected data set.

In one possible implementation, such as for step S1013, in the process of performing artificial intelligence analysis on the subscription association intention data set categorized by each target content tag based on the corresponding request subscription content feature of the target content tag categorization, the following sub-steps may be implemented.

And a substep S10134, determining a subscription-requesting tag information parameter of each subscription-requesting tag information of each target content tag classification and a tag coverage object covered by the subscription-requesting tag information based on the subscription-requesting content characteristics corresponding to the target content tag classification.

And a substep S10135, determining tracking parameters of an artificial intelligence analysis component required for artificial intelligence analysis of the request subscription label information in each target content label classification according to the request subscription label information parameters of the request subscription label information in each target content label classification and the label coverage object covered by the request subscription label information.

And a substep S10136, determining each artificial intelligence analysis component as a tracking object according to the tracking parameter of the artificial intelligence analysis component required by each request for subscribing the tag information, wherein the intention tracking information corresponding to the tracking object is intention tracking information other than intention tracking information classified by the currently configured target content tag contained in the request for subscribing the tag information.

And a substep S10137, establishing an intention service node of the tracked object according to intention tracking information corresponding to the tracked object, determining a coverage item of the intention service node, and obtaining preliminary conversation intention information of artificial intelligence analysis on the subscription associated intention data set classified by each target content label by the first tracked object in the coverage item.

And a substep S10138, when screening the preliminary session intention information of each tracked object after the first tracked object in sequence according to the hierarchy of the tracked object, screening the tracked object and the preliminary session intention information of each tracked object after the tracked object, reestablishing an intention service node of the tracked object according to the screened preliminary session intention information, determining a coverage item of the reestablished intention service node, and obtaining screened preliminary session intention information of the tracked object in the coverage item of the reestablished intention service node.

And a substep S10139 of taking the screened preliminary conversation intention information of all the tracked objects as an artificial intelligence analysis result after the screened preliminary conversation intention information of all the tracked objects is obtained.

In one possible implementation, for example, after step S1013, the following steps may be further included:

step S1015, determining whether an updatable point of interest context vector set for indicating that the subscription content node has an updatable service exists during the artificial intelligence analysis, and extracting a first subscription association intention of a first subscription content node corresponding to the updatable point of interest context vector set analyzed by the artificial intelligence analysis and a second subscription association intention of at least one second subscription content node having an updatable service relationship with the first subscription content node when the updatable point of interest context vector set is detected;

step S1016, global historical conversation intention information between the first subscription association intention and the at least one second subscription association intention is determined according to a preset artificial intelligence model.

In one possible implementation, for example, for step S1015, a first subscription association intention of a first subscription content node corresponding to the updatable point of interest scenario vector set analyzed by the artificial intelligence analysis and a second subscription association intention of at least one second subscription content node having an updatable business relationship with the first subscription content node may be extracted from the artificial intelligence analysis record information generated in the artificial intelligence analysis process. Wherein the at least one second subscription content node having an updatable business relationship with the first subscription content node may refer to a second subscription content node having a linkage effect associated with the first subscription content node.

For example, if a subscribed content node needs to extend mining during mining of a first subscribed content node, the subscribed content node may be understood as a second subscribed content node having an updatable business relationship with the first subscribed content node.

In one possible implementation, such as for step S1016, this may be achieved by the following exemplary substeps, described in detail below.

And a substep S10161, fusing the first subscription associated intention with the subscription associated intention node corresponding to at least one second subscription associated intention according to each same subscription associated intention node to obtain a fused subscription associated intention.

The sub-step S10162 of adding the first subscription association intention and the at least one second subscription association intention to a preset intention distribution map queue, and establishing a plurality of first intention distribution map parameters of the first subscription association intention and a plurality of second intention distribution map parameters of the second subscription association intention based on the intention distribution map queue.

Substep S10163, determining first intention semantic information of a first subscription content node according to each first intention distribution map parameter, determining second intention semantic information of a second subscription content node according to each second intention distribution map parameter, mapping the first intention semantic information and the second intention semantic information to a session service feature model to obtain a first subscription associated intention feature corresponding to the first intention semantic information and a second subscription associated intention feature corresponding to the second intention semantic information, determining a plurality of session problem objects corresponding to a fusion subscription associated intention of the session service feature model, summarizing the plurality of session problem objects to obtain at least a plurality of session problem mining lists of different types, and mining a first session generation feature and a corresponding first session generation feature of each session problem object in the session problem mining lists in a preset artificial intelligence analysis process according to each session problem mining list A second session generation feature that subscribes to the associated intent feature.

And a substep S10164, according to the mining results of the first session generation characteristic and the second session generation characteristic corresponding to each session problem object in the session problem mining list, splicing the mining results according to the preset priority of knowledge expectation to generate an intention information stream, restoring the spliced intention information stream, and determining global historical session intention information of the first subscription content node and the at least one second subscription content node.

Therefore, the subsequent artificial intelligence analysis can be performed by taking the associated subscription content nodes as independent mining targets in a targeted manner in the actual artificial intelligence analysis process.

In one possible implementation, for step S10164, for example, in the process of restoring the spliced intention information stream and determining the global historical conversational intention information of the first subscription content node and the at least one second subscription content node, the spliced intention information stream may be reversely converted according to each corresponding simulation mining node to obtain the global historical conversational intention information of the first subscription content node and the at least one second subscription content node.

Fig. 3 is a schematic diagram of functional modules of the session intention extraction apparatus 300 for cloud computing and artificial intelligence according to the embodiment of the present disclosure, in this embodiment, the cloud computing and artificial intelligence-oriented session intention extraction apparatus 300 may be divided into the functional modules according to the method embodiment executed by the artificial intelligence platform 100, that is, the following functional modules corresponding to the cloud computing and artificial intelligence-oriented session intention extraction apparatus 300 may be used to execute each method embodiment executed by the artificial intelligence platform 100. The cloud computing and artificial intelligence oriented session intention extraction apparatus 300 may include a receiving module 310, a determining module 320, a request subscribing module 330 and an updating module 340, and the functions of the functional modules of the cloud computing and artificial intelligence oriented session intention extraction apparatus 300 are described in detail below.

A receiving module 310, configured to receive big data service trajectory information for a cloud computing issued task sent by the service processing device 200, and extract target intention labeling information of the big data service trajectory information; wherein the target intention marking information comprises target intention potential points. The receiving module 310 may be configured to perform the step S110, and the detailed implementation of the receiving module 310 may refer to the detailed description of the step S110.

A determining module 320, configured to determine, according to historical conversation intention information of the target intention potential point, a service interest item set corresponding to the target intention potential point, where the historical conversation intention information is obtained by performing artificial intelligence analysis on a periodically collected data set generated by the service processing device 200 in a subscription application scenario and a target reference interaction service corresponding to the periodically collected data set by the artificial intelligence platform. The determining module 320 may be configured to perform the step S120, and the detailed implementation of the determining module 320 may refer to the detailed description of the step S120.

A subscription requesting module 330, configured to request the service processing device 200 to subscribe to a subscription service corresponding to the service interest item set. The subscription requesting module 330 may be configured to perform the step S130, and the detailed implementation of the subscription requesting module 330 may refer to the detailed description of the step S130.

The updating module 340 is configured to update a cloud computing delivery task delivered to the service processing device next time according to a target subscription service selected by the service processing device from the subscription services corresponding to the service interest item set. The updating module 340 may be configured to perform the step S140, and the detailed implementation of the updating module 340 may refer to the detailed description of the step S140.

It should be noted that the division of the modules of the above apparatus is only a logical division, and the actual implementation may be wholly or partially integrated into one physical entity, or may be physically separated. And these modules may all be implemented in software invoked by a processing element. Or may be implemented entirely in hardware. And part of the modules can be realized in the form of calling software by the processing element, and part of the modules can be realized in the form of hardware. For example, the receiving module 310 may be a processing element separately set up, or may be implemented by being integrated into a chip of the apparatus, or may be stored in a memory of the apparatus in the form of program code, and the processing element of the apparatus calls and executes the functions of the receiving module 310. Other modules are implemented similarly. In addition, all or part of the modules can be integrated together or can be independently realized. The processing element described herein may be an integrated circuit having signal processing capabilities. In implementation, each step of the above method or each module above may be implemented by an integrated logic circuit of hardware in a processor element or an instruction in the form of software.

For example, the above modules may be one or more integrated circuits configured to implement the above methods, such as: one or more Application Specific Integrated Circuits (ASICs), or one or more microprocessors (DSPs), or one or more Field Programmable Gate Arrays (FPGAs), among others. For another example, when some of the above modules are implemented in the form of a processing element scheduler code, the processing element may be a general-purpose processor, such as a Central Processing Unit (CPU) or other processor that can call program code. As another example, these modules may be integrated together, implemented in the form of a system-on-a-chip (SOC).

Fig. 4 is a schematic diagram illustrating a hardware structure of the artificial intelligence platform 100 for implementing the control device according to the embodiment of the disclosure, and as shown in fig. 4, the artificial intelligence platform 100 may include a processor 110, a machine-readable storage medium 120, a bus 130, and a transceiver 140.

In a specific implementation process, at least one processor 110 executes computer-executable instructions stored in the machine-readable storage medium 120 (for example, the receiving module 310, the determining module 320, the subscription requesting module 330, and the updating module 340 included in the cloud computing and artificial intelligence oriented session intention extraction apparatus 300 shown in fig. 3), so that the processor 110 may execute the cloud computing and artificial intelligence oriented session intention extraction method according to the above method embodiment, where the processor 110, the machine-readable storage medium 120, and the transceiver 140 are connected through the bus 130, and the processor 110 may be configured to control transceiving actions of the transceiver 140, so as to transceive data with the business processing device 200.

For a specific implementation process of the processor 110, reference may be made to the above-mentioned various method embodiments executed by the artificial intelligence platform 100, which implement principles and technical effects similar to each other, and details of this embodiment are not described herein again.

In the embodiment shown in fig. 4, it should be understood that the Processor may be a global business interactive matching process (CPU), other general purpose Processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), etc. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of a method disclosed in connection with the present invention may be embodied directly in a hardware processor, or in a combination of the hardware and software modules within the processor.

The machine-readable storage medium 120 may comprise high-speed RAM memory and may also include non-volatile storage NVM, such as at least one disk memory.

The bus 130 may be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, an Extended ISA (EISA) bus, or the like. The bus 130 may be divided into an address bus, a data bus, a control bus, and the like. For ease of illustration, the buses in the figures of the present application are not limited to only one bus or one type of bus.

In addition, an embodiment of the present application further provides a readable storage medium, where the readable storage medium stores computer-executable instructions, and when a processor executes the computer-executable instructions, the verification processing method based on the blockchain offline payment is implemented as above.

Having thus described the basic concept, it will be apparent to those skilled in the art that the foregoing detailed disclosure is to be regarded as illustrative only and not as limiting the present specification. Various modifications, improvements and adaptations to the present description may occur to those skilled in the art, although not explicitly described herein. Such modifications, improvements and adaptations are proposed in the present specification and thus fall within the spirit and scope of the exemplary embodiments of the present specification.

Also, the description uses specific words to describe embodiments of the description. Such as "one possible implementation," "one possible example," and/or "exemplary" means that a particular feature, structure, or characteristic described in connection with at least one embodiment of the specification is included. Therefore, it is emphasized and should be appreciated that two or more references to "one possible implementation," "one possible example," and/or "exemplary" in various places throughout this specification are not necessarily referring to the same embodiment. Furthermore, some features, structures, or characteristics of one or more embodiments of the specification may be combined as appropriate.

Moreover, those skilled in the art will appreciate that aspects of the present description may be illustrated and described in terms of several patentable species or situations, including any new and useful combination of processes, machines, manufacture, or materials, or any new and useful improvement thereof. Accordingly, aspects of this description may be performed entirely by hardware, entirely by software (including firmware, resident software, micro-code, etc.), or by a combination of hardware and software. The above hardware or software may be referred to as "data block," module, "" engine, "" unit, "" component, "or" system. Furthermore, aspects of the present description may be represented as a computer product, including computer readable program code, embodied in one or more computer readable media.

The computer storage medium may comprise a propagated data signal with the computer program code embodied therewith, for example, on baseband or as part of a carrier wave. The propagated signal may take any of a variety of forms, including electromagnetic, optical, etc., or any suitable combination. A computer storage medium may be any computer-readable medium that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code located on a computer storage medium may be propagated over any suitable medium, including radio, cable, fiber optic cable, RF, or the like, or any combination of the preceding.

Computer program code required for the operation of various portions of this specification may be written in any one or more programming languages, including an object oriented programming language such as Java, Scala, Smalltalk, Eiffel, JADE, Emerald, C + +, C #, VB.NET, Python, and the like, a conventional programming language such as C, Visual Basic, Fortran 2003, Perl, COBOL 2002, PHP, ABAP, a dynamic programming language such as Python, Ruby, and Groovy, or other programming languages. The program code may run entirely on the user's computer, or as a stand-alone software package on the user's computer, partly on the user's computer and partly on a remote computer or entirely on the remote computer or digital financial services terminal. In the latter scenario, the remote computer may be connected to the user's computer through any network format, such as a Local Area Network (LAN) or a Wide Area Network (WAN), or the connection may be made to an external computer (for example, through the Internet), or in a cloud computing environment, or as a service, such as a software as a service (SaaS).

Additionally, the order in which the elements and lists are processed, the use of alphanumeric characters, or other designations in this specification is not intended to limit the order in which the processes and methods of this specification are performed, unless otherwise specified in the claims. While various presently contemplated embodiments of the invention have been discussed in the foregoing disclosure by way of example, it is to be understood that such detail is solely for that purpose and that the appended claims are not limited to the disclosed embodiments, but, on the contrary, are intended to cover all modifications and equivalent arrangements that are within the spirit and scope of the embodiments herein. For example, although the system components described above may be implemented through interactive services, they may also be implemented through software-only solutions, such as installing the described system on an existing digital financial services terminal or mobile device.

Similarly, it should be noted that in the preceding description of embodiments of the present specification, various features are sometimes grouped together in a single embodiment, figure, or description thereof for the purpose of streamlining the disclosure aiding in the understanding of one or more of the embodiments. This method of disclosure, however, is not intended to imply that more features than are expressly recited in a claim. Indeed, the embodiments may be characterized as having less than all of the features of a single embodiment disclosed above.

It is to be understood that the descriptions, definitions and/or uses of terms in the accompanying materials of this specification shall control if they are inconsistent or contrary to the descriptions and/or uses of terms in this specification.

Finally, it should be understood that the embodiments described herein are merely illustrative of the principles of the embodiments of the present disclosure. Other variations are also possible within the scope of the present description. Thus, by way of example, and not limitation, alternative configurations of the embodiments of the specification can be considered consistent with the teachings of the specification. Accordingly, the embodiments of the present description are not limited to only those embodiments explicitly described and depicted herein.

Claims

1. A session intention extraction method facing cloud computing and artificial intelligence is applied to an artificial intelligence platform, the artificial intelligence platform is in communication connection with a plurality of business processing devices, and the method comprises the following steps:

acquiring a periodic collected data set generated by the service processing equipment in the subscription application scene and a target reference interactive service corresponding to the periodic collected data set from the service processing equipment;

acquiring subscription content nodes of task subscription items corresponding to cloud computing issued tasks under the target reference interactive services, classifying the subscription content nodes under the target reference interactive services according to preset target content label classification, and respectively generating a subscription content node set classified by each target content label;

2. The cloud computing and artificial intelligence oriented session intention extraction method as claimed in claim 1, wherein the step of obtaining that each subscribed content node in the set of subscribed content nodes categorized by the target content tag matches the subscription associated intention data of the periodically collected data set comprises:

obtaining the subscription content construction characteristics related to each subscription content node in the subscription content node set classified by the target content label;

matching corresponding target data areas from the periodically collected data set according to the subscription content construction characteristics related to each subscription content node;

and constructing a subscription associated intention attribute corresponding to each service record plate in the target data area with matched characteristics according to the subscription content related to each subscription content node, and determining that each subscription content node in the subscription content node set classified by the target content label is matched with the subscription associated intention data of the periodically collected data set.

3. The cloud computing and artificial intelligence oriented session intention extraction method according to claim 1, wherein the step of performing artificial intelligence analysis on the subscription associated intention data set classified by each target content tag based on the corresponding request subscription content feature of the target content tag classification comprises:

determining a subscription request tag information parameter of each subscription request tag information of each target content tag classification and a tag coverage object covered by the subscription request tag information based on the subscription request content characteristics corresponding to the target content tag classification;

determining tracking parameters of an artificial intelligence analysis component required for artificial intelligence analysis on the request subscription label information in each target content label classification according to the request subscription label information parameters of the request subscription label information in each target content label classification and label covering objects covered by the request subscription label information;

determining each artificial intelligence analysis component as a tracking object according to the tracking parameter of the artificial intelligence analysis component required by each request for subscribing the label information, wherein the intention tracking information corresponding to the tracking object is intention tracking information except intention tracking information classified by a currently configured target content label contained in the request for subscribing the label information;

establishing an intention service node of the tracked object according to intention tracking information corresponding to the tracked object, determining a coverage item of the intention service node, and obtaining preliminary session intention information of artificial intelligent analysis of a subscription associated intention data set classified by each target content label by a first tracked object in the coverage item;

when the preliminary session intention information of each tracking object behind the first tracking object is sequentially screened according to the hierarchy of the tracking object, screening the tracking object and the preliminary session intention information of each tracking object behind the tracking object, reestablishing an intention service node of the tracking object according to the screened preliminary session intention information, determining a coverage item of the reestablished intention service node, and obtaining screened preliminary session intention information of the tracking object in the coverage item of the reestablished intention service node;

and after the screening preliminary session intention information of all the tracked objects is obtained, taking the screening preliminary session intention information of all the tracked objects as an artificial intelligence analysis result.

4. The cloud computing and artificial intelligence oriented conversation intention extraction method according to claim 1, wherein after the step of performing artificial intelligence analysis on the subscription associated intention data set of each target content tag classification based on the request subscription content feature corresponding to the target content tag classification to obtain the historical conversation intention information of each target content tag classification, the method further comprises the following steps:

judging whether an updatable interest point scene vector set for indicating that the subscription content nodes have updatable services exists or not in the artificial intelligence analysis process, and extracting a first subscription associated intention of a first subscription content node corresponding to the updatable interest point scene vector set analyzed by artificial intelligence and a second subscription associated intention of at least one second subscription content node having updatable service relationship with the first subscription content node when the updatable interest point scene vector set is detected;

global historical conversation intention information between the first subscription association intention and the at least one second subscription association intention is determined according to a preset artificial intelligence model.

5. The cloud-computing and artificial-intelligence oriented session intention extraction method according to claim 4, wherein the step of determining whether an updatable point of interest context vector set for indicating existence of an updatable service of the subscribed content nodes exists in the artificial-intelligence analysis process, and extracting a first subscription associated intention of a first subscribed content node corresponding to the updatable point of interest context vector set analyzed by the artificial-intelligence analysis and a second subscription associated intention of at least one second subscribed content node having an updatable service relationship with the first subscribed content node when the updatable point of interest context vector set is detected, comprises:

extracting a first subscription associated intention of a first subscription content node corresponding to an updatable interest point scene vector set analyzed by artificial intelligence and a second subscription associated intention of at least one second subscription content node having an updatable business relationship with the first subscription content node from artificial intelligence analysis record information generated in the artificial intelligence analysis process; at least one second subscription content node having an updatable business relationship with the first subscription content node may refer to a second subscription content node having a linkage effect associated with the first subscription content node;

if a subscribed content node needs to extend mining in the process of mining the first subscribed content node, the subscribed content node can be understood as a second subscribed content node having an updatable business relationship with the first subscribed content node.

6. The cloud computing and artificial intelligence oriented session intention extraction method as claimed in claim 4, wherein the step of determining global historical session intention information between the first subscription associated intention and the at least one second subscription associated intention according to a preset artificial intelligence model comprises:

fusing the first subscription associated intention with at least one subscription associated intention node corresponding to the second subscription associated intention according to each same subscription associated intention node to obtain a fused subscription associated intention;

adding the first subscription association intention and at least one second subscription association intention to a preset intention distribution map queue, and establishing a plurality of first intention distribution map parameters of the first subscription association intention and a plurality of second intention distribution map parameters of the second subscription association intention based on the intention distribution map queue;

determining first intention semantic information of a first subscription content node according to each first intention distribution map parameter, determining second intention semantic information of a second subscription content node according to each second intention distribution map parameter, mapping the first intention semantic information and the second intention semantic information to a session service feature model to obtain a first subscription associated intention feature corresponding to the first intention semantic information and a second subscription associated intention feature corresponding to the second intention semantic information, determining a plurality of session problem objects corresponding to a fusion subscription associated intention of the session service feature model, summarizing the plurality of session problem objects to obtain at least a plurality of session problem mining lists of different categories, and mining a first session generation feature corresponding to the first subscription associated intention feature and a second subscription associated feature corresponding to each session problem object in the session problem mining lists in a preset artificial intelligence analysis process according to each session problem mining list A second session generation feature of the intent feature;

and according to mining results of the first session generation characteristics and the second session generation characteristics corresponding to each session problem object in the session problem mining list, splicing the generated intention information streams according to the preset priority of knowledge expectation, restoring the spliced intention information streams, and determining global historical session intention information of the first subscription content node and the at least one second subscription content node.

7. The cloud computing and artificial intelligence oriented session intention extraction method according to claim 6, wherein the step of restoring the spliced intention information stream and determining global historical session intention information of the first subscription content node and the at least one second subscription content node includes:

and performing inverse conversion on the spliced intention information stream according to each corresponding simulation mining node to obtain global historical conversation intention information of the first subscription content node and the at least one second subscription content node.

8. The cloud computing and artificial intelligence oriented session intention extraction method according to any one of claims 1 to 7, wherein the method further comprises:

9. The cloud computing and artificial intelligence oriented conversation intention extraction method according to claim 1, wherein the step of determining the set of service interest items corresponding to the target intention potential point according to the historical conversation intention information of the target intention potential point comprises:

10. A cloud computing and artificial intelligence oriented conversation intention extraction system is characterized by comprising an artificial intelligence platform and a plurality of business processing devices in communication connection with the artificial intelligence platform;

the artificial intelligence platform is used for: