CN207818190U - A kind of information interaction system based on speech recognition - Google Patents
A kind of information interaction system based on speech recognition Download PDFInfo
- Publication number
- CN207818190U CN207818190U CN201721156798.4U CN201721156798U CN207818190U CN 207818190 U CN207818190 U CN 207818190U CN 201721156798 U CN201721156798 U CN 201721156798U CN 207818190 U CN207818190 U CN 207818190U
- Authority
- CN
- China
- Prior art keywords
- service
- user
- business processing
- model
- module
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Landscapes
- Telephonic Communication Services (AREA)
Abstract
The utility model embodiment discloses a kind of information interaction system based on speech recognition, which includes:Sound identification module, processor, speech production module and display module, wherein the sound identification module identifies the demand type of service of user for being based on voice input by user;The processor integrates result for obtaining and the corresponding business processing flow of the type of service and service model corresponding with the business processing flow;According to the business processing step and/or interactive information input by user being presently in, destination service model is determined in the service model integrates result;The speech production module, for generating and the interactive voice of the destination service Model Matching;The display module, for showing business processing interface.The utility model embodiment disclosure satisfy that user by the demand with financial self-service equipment progress speech exchange transacting business, promotes the intelligence degree of financial self-service equipment.
Description
Technical field
The utility model embodiment is related to technical field of voice interaction more particularly to a kind of information based on speech recognition is handed over
Mutual system.
Background technology
As information technology develops rapidly, society is constantly progressive, and the cloth of financial self-service equipment high-volume gradually increases, people couple
The utilization rate of financial self-service equipment is higher and higher, and the thing followed is that people are more next to financial self-service equipment intelligent Application demand
It is more urgent, the development of financial self-service equipment intelligence degree is promoted, the developing direction of financial self-service equipment is become.
Financial self-service equipment in the prior art is not carried out the work(of the speech exchange transacting business by user and equipment
Can, user is only to carry out business handling manually according to the prompt of financial self-service equipment, and during business handling, user cannot
It engages in the dialogue with financial self-service equipment, particularly with the user for partly having hand disease, cannot be satisfied it and set by self-service finance
The demand of standby transacting business, causes user experience poor.
Utility model content
The utility model embodiment provides a kind of information interaction system based on speech recognition, to solve to use in the prior art
Family can not with financial self-service equipment by speech exchange transacting business the problem of.
In a first aspect, the utility model embodiment provides a kind of information interaction system based on speech recognition, including:Language
Sound identification module, processor, speech production module and display module, wherein
The sound identification module is based on the voice, identifies the demand industry of user for receiving voice input by user
Service type, and the demand type of service of the user is sent to the processor;
The processor, the demand type of service of the user for obtaining sound identification module transmission;It obtains and the need
Seek the corresponding business processing flow of type of service, wherein the business processing flow includes:It is corresponding with the demand type of service
Whole business processing steps and different business processing links between logical relation;It obtains and the business processing flow
Corresponding service model integrates result, wherein the service model integrates result and includes:Service corresponding with business processing step
Logical relation between model, and different service models;It is inputted according to the business processing step being presently in and/or user
Interactive information, destination service model is determined in the service model integrates result, control display module show and the mesh
The matched business processing interface of service model is marked, and the destination service model is sent to speech production module, so that described
Speech production module exports interactive voice corresponding with the destination service model;
The speech production module generates and the destination service Model Matching for receiving the destination service model
Interactive voice, export the interactive voice to user;
The display module, for showing and the business processing interface of the destination service Model Matching.
Further, the sound identification module includes:Speech pattern recognition unit, linguistic interpretation unit and dialogue stream
Control unit, wherein
The speech pattern recognition unit, for receiving voice input by user, obtain the voice input by user with
The matching result of pattern model, and the matching result is sent to the linguistic interpretation unit, wherein the pattern model root
It is built according to the surface layer of the business operation term of collection, grammer layer and keyword;
The linguistic interpretation unit for executing encoding operation to the matching result, and carries out language to coding result
Analysis, after being verified to analysis result, sends the analysis result to the dialogue stream control unit, wherein described
Language analysis includes sentence structure analysis and semantic analysis;
The dialogue stream control unit, the demand type of service for identifying user according to the analysis result.
Further, the system also includes:
Task control module, the task control module are used for if it is determined that receiving the demand service class of the user
Type determines the corresponding industry of the demand type of service according to the demand type of service and pre-stored pending business information
Business process flow, sends the business processing flow to the processor, wherein pending business information includes all service class
Logical relation in type between the corresponding whole business processing steps of each type of service and different business processing links;
The processor is additionally operable to:The demand type of service of the user is sent to the task control module.
Further, the system also includes:
Service module, the service module are used to, according to the business processing flow, integrate the service model
Operation, integrates result by service model and is sent to the processor;
The processor is additionally operable to:The business processing flow is sent to the server, and receives server and is based on institute
The service model for stating business processing flow return integrates result.
Further, the service module includes:
Mapping relations acquiring unit, for obtaining the mapping between pre-stored service model and the business processing flow
Relationship;
Model extraction unit, for according to the mapping relations, extracting business processing step in the business processing flow
Corresponding service model;
Model integrated unit is used for according to the logical relation between different service models, at extraction, with the business
The service model of reason flow path match carries out integrated operation, obtains service model and integrates result.
Further, the system also includes:
Communication module, the communication module are connected with the server, for receiving the speech pattern recognition unit
The voice input by user institute that is sent to the server, and server is calculated based on the voice input by user
The matching result for stating voice and pattern model input by user is back to the speech pattern recognition unit;
The speech pattern recognition unit is additionally operable to:The matching result is received by communication module.
Further, the communication module also particularly useful for:
The destination service model is sent to the server, and the server is based on the destination service mould
Type, determining business processing interface are back to the processor;
The processor is additionally operable to:The business processing interface is received by the communication module.
Further, the system is applied to financial self-service equipment.
Second aspect, the utility model embodiment provide a kind of information interacting method based on speech recognition, including:
Obtain the demand type of service for the user that sound identification module is sent;
Obtain corresponding with demand type of service business processing flow, wherein the business processing flow includes:With
Logical relation between the corresponding whole business processing steps of the demand type of service and different business processing links;
It obtains service model corresponding with the business processing flow and integrates result, wherein the integrated knot of the service model
Fruit includes:Logical relation between service model corresponding with business processing step, and different service models;
It is integrated in the service model according to the business processing step and/or interactive information input by user being presently in
As a result destination service model is determined in, control display module is shown and the business processing interface of the destination service Model Matching,
And the destination service model is sent to speech production module, so that speech production module output and the destination service
The corresponding interactive voice of model.
The third aspect, the utility model embodiment additionally provide a kind of information interactive device based on speech recognition, including:
Type of service acquisition module, the demand type of service of the user for obtaining sound identification module transmission;
Business processing flow acquisition module, for obtaining business processing flow corresponding with the demand type of service,
In, the business processing flow includes:Whole business processing step corresponding with the demand type of service and different business
Logical relation between processing links;
Service model integrates result acquisition module, integrated for obtaining service model corresponding with the business processing flow
As a result, wherein the service model integrates result and includes:Service model corresponding with business processing step, and different services
Logical relation between model;
Destination service model determining module, for according to the business processing step and/or friendship input by user being presently in
Mutual information determines destination service model in the service model integrates result, and control display module is shown to be taken with the target
The business processing interface for Model Matching of being engaged in, and the destination service model is sent to speech production module, so that the voice
Generation module exports interactive voice corresponding with the destination service model.
The utility model embodiment is identified voice input by user by sound identification module, obtains user demand
Type of service, after the corresponding business processing flow of type of service for determining user demand, integrate and business processing flow
Matched service module is based on the service module, is engaged in the dialogue with user by speech production module, is used in further obtain
While the demand at family or selection, business processing interface is provided to the user by display module, realizes good human-computer interaction
Function.The utility model embodiment can accurately identify the demand of user, and provide the service of business handling to the user, be promoted
The intelligence degree of financial self-service equipment is done meanwhile, it is capable to meet user by carrying out speech exchange with financial self-service equipment
The demand of reason business, effectively the user experience is improved.
Description of the drawings
Fig. 1 is a kind of structural representation of information interaction system based on speech recognition in the utility model embodiment one
Figure;
Fig. 2 is a kind of structural representation of information interaction system based on speech recognition in the utility model embodiment two
Figure;
Fig. 3 is a kind of structural representation of information interaction system based on speech recognition in the utility model embodiment three
Figure;
Fig. 4 is a kind of structural representation of information interaction system based on speech recognition in the utility model embodiment four
Figure;
Fig. 5 is by a kind of applicable method of information interaction system based on speech recognition in the utility model embodiment five
Flow chart;
Fig. 6 is a kind of structural representation of information interactive device based on speech recognition in the utility model embodiment six
Figure.
Specific implementation mode
The utility model is described in further detail with reference to the accompanying drawings and examples.It is understood that herein
Described specific embodiment is used only for explaining the utility model, rather than the restriction to the utility model.It further needs exist for
It is bright, it illustrates only for ease of description, in attached drawing and the relevant part of the utility model rather than entire infrastructure.
Embodiment one
Fig. 1 is a kind of structural representation for information interaction system based on speech recognition that the utility model embodiment one provides
Figure, the present embodiment are applicable to through the information interaction system based on speech recognition come the case where carrying out business handling, the system
The mode that hardware and/or software may be used is realized.
The system of the utility model embodiment one specifically includes:Sound identification module 110, processor 120, speech production mould
Block 130 and display module 140.
The sound identification module 110 is based on the voice, identifies the need of user for receiving voice input by user
Type of service is sought, and the demand type of service of the user is sent to the processor.
Specifically, processor after receiving the speech recognition request of user, activates sound identification module.Speech recognition
Request can be that the volume that user inputs voice is more than setting decibel, or confirmly detects the presence of user, the utility model pair
This is not limited.Illustratively, setting decibel could be provided as 50 decibels, or the presence for passing through sensor senses user.Voice
Identification module receives voice input by user, after obtaining the matching result of voice and pattern model input by user, to
The voice of family input performs the encoding operation, and analyzes coding result, obtains the type of service of user demand.I.e. in determination
After the corresponding sentence content of user speech, the analysis of grammer, semanteme and sentence structure is carried out to sentence content, obtains user
Intention.User view, you can the voice that next may be inputted with the phonetic decision user currently inputted according to user and/
Or the type of service of user demand, the type of service of demand is sent to processor.Type of service may include transferring accounts, withdrawing the money
Pass through the type for the business that financial self-service equipment can be handled.
The processor 120, the demand type of service of the user for obtaining sound identification module transmission;Obtain with it is described
The corresponding business processing flow of demand type of service, wherein the business processing flow includes:With the demand type of service pair
Logical relation between the whole business processing steps and different business processing links answered;It obtains and the business processing stream
The corresponding service model of journey integrates result, wherein the service model integrates result and includes:Clothes corresponding with business processing step
The logical relation being engaged between model, and different service models;It is defeated according to the business processing step and/or user that are presently in
The interactive information entered, determines destination service model in the service model integrates result, control display module show with it is described
The business processing interface of destination service Model Matching, and the destination service model is sent to speech production module, so that institute
Predicate sound generation module exports interactive voice corresponding with the destination service model.
Specifically, processor receives the demand type of service for the user that sound identification module is sent, according to the demand of user
Type of service and pending business information determine the corresponding business processing flow of user demand type of service.Pending business letter
Breath can store in the processor, can also be stored in task control module, pending business information may include system energy
Logic between the corresponding all business processing steps of whole types of service and different business processing links that enough handle is closed
System.Determine the business processing flow and business of system jointly by the demand type of service of user and pending business information
Manage the corresponding business processing step of flow.At least one service model can store in the processor, can also be stored in service
In module, according to business processing step, at least one service model corresponding with business processing step is dynamically extracted, according to industry
The logical relation between logical relation and/or different service models between processing links of being engaged in, at least one clothes of extraction
Business model carries out operation that is integrated and integrating, to provide business processing flow corresponding integrity service content, and provides business
The corresponding specific service content of each business processing step of process flow.
Processor can be with whole business processing steps of monitoring business process flow, according to the business processing ring being presently in
Section and/or interactive information input by user determine that destination service model, i.e. basis are worked as in the service model integrates result
Preceding residing business processing step, provides specific service content corresponding with the business processing step being presently in.Wherein, user
The interactive information of input can be the interactive information of voice input, or the interaction being manually entered by business processing interface
Information, the utility model are without limitation.Business processing step may include all business processing steps and different industry
Logical relation between processing step of being engaged in.When processor determines destination service model, destination service model is sent to voice
Generation module, speech production module generate the corresponding interactive voice of destination service model, and human-computer interaction is carried out with user.User can
With the voice generated according to speech production module, voice input is further carried out by sound identification module, it is man-machine right to be formed
Words.Meanwhile processor provides business processing corresponding with destination service model interface, and receive user and be based on business processing interface
The data of input, business processing interface can be shown by display module.When determining that user's finishing service is handled, processing
Device can provide initialization display interface, and control display module display initialization display interface.
The speech production module 130 generates and the destination service model for receiving the destination service model
The interactive voice matched exports the interactive voice to user.
Specifically, after speech production module receives destination service model, the inverse of coding is executed to destination service model
Operation existing content will be reduced to voice in a coded form, and generate interactive voice corresponding with destination service model, and defeated
Go out to user.
The display module 140, for showing and the business processing interface of the destination service Model Matching.
Specifically, display module can be display screen, the business processing interface display for being provided processor by display screen
To user.
In the present embodiment, user can be inputted " I will transfer accounts " by sound identification module, and sound identification module, which receives, to be used
The voice of " I will transfer accounts " at family, is input to pattern model by the voice and matches, and obtains matching result, that is, it is defeated to obtain user
The sentence content of the voice entered, encodes sentence content, and after being further analyzed, and obtains the demand business of user
Type, i.e. transferred account service type.Processor obtains the corresponding industry of transferred account service type according to the transferred account service type of user demand
Logic between process flow of being engaged in and the whole business processing steps and different business processing links of business processing flow is closed
System.The integrated as a result, simultaneously according to the business processing step being presently in of service model is obtained according to business processing flow, is being serviced
Determine that destination service model, destination service model can provide at least one authentication mode in model integrated result, for example,
Password authentification and iris verification, speech production module can generate interactive voice corresponding with destination service model, such as " please select
Select the identification authentication mode of your needs:1, password authentification, 2, iris verification ", display module display target service model is corresponding
Business handling interface, user can show what screen inputted at business handling interface by way of voice input or by click
Mode carries out the selection of authentication mode.No matter whether the authentication of user succeeds, in service model integrates result all
It can determine corresponding destination service model.Under the conditions of subscriber authentication is successful, result can be integrated in service model
The corresponding destination service model of middle determination is to determine transfer account, and it is true that speech production module, which can be generated with destination service model,
Determine the corresponding interactive voice of transfer account, display module display target service model is to determine the corresponding business handling of transfer account
Interface, and so on, by the interaction between user and system, provide the business handling operation of demand type of service to the user.
The utility model embodiment one provides a kind of information interaction system based on speech recognition, passes through speech recognition mould
Voice input by user is identified in block, obtains the type of service of user demand, in the type of service pair for determining user demand
After the business processing flow answered, integrates and language is passed through based on the service module with business processing flow matched service module
Sound generation module engages in the dialogue with user, is to use by display module while further obtaining demand or the selection of user
Family provides business processing interface, realizes good human-computer interaction function.The utility model embodiment can accurately identify use
The demand at family, and provide the service of business handling to the user, the intelligence degree of financial self-service equipment is improved, meanwhile, it is capable to
Meets the needs of user with financial self-service equipment by carrying out speech exchange transacting business, effectively the user experience is improved.
Embodiment two
Fig. 2 is a kind of structural representation for information interaction system based on speech recognition that the utility model embodiment two provides
Figure, the utility model embodiment two is optimized based on embodiment one, specifically to sound identification module structure into
One-step optimization, as shown in Fig. 2, the sound identification module of the utility model embodiment two specifically includes:Speech pattern recognition unit
210, linguistic interpretation unit 220 and dialogue stream control unit 230.
The speech pattern recognition unit 210 obtains the voice input by user for receiving voice input by user
With the matching result of pattern model, and the matching result is sent to the linguistic interpretation unit, wherein the pattern model
It is built according to the surface layer of the business operation term of collection, grammer layer and keyword.
Specifically, collecting user's habitually business operation term in advance, surface layer, grammer layer are carried out to business operation term
And after the analysis of keyword, i.e., the elements such as voice, grammer, sentence structure, keyword is carried out to business operation term and divided
After analysis, forming types model.Pattern model can store in the processor, can also be stored in far-end server, this practicality
It is novel without limitation.Speech pattern recognition unit receives voice input by user, by voice input by user and pattern mould
Type is matched, and is obtained matching result, that is, is obtained the corresponding sentence content of voice input by user, and matching result is sent to
Linguistic interpretation unit.
The linguistic interpretation unit 220 for executing encoding operation to the matching result, and carries out language to coding result
Speech analysis, after being verified to analysis result, sends the analysis result to the dialogue stream control unit, wherein institute
Predicate speech analysis includes sentence structure analysis and semantic analysis.
Specifically, linguistic interpretation unit encodes matching result, voice is become to carry out sentence structure after text
And semantic analysis, and semantic and grammer inspection and verification are carried out to analysis result and also may be used with eliminating semantic ambiguity
With by the candidate text for establishing voice, to carry out semantic and grammer inspection and verification to analysis result, after verification
Analysis result be sent to dialogue stream control unit.
The dialogue stream control unit 230, the demand type of service for identifying user according to the analysis result.
Specifically, the extraction that dialogue stream control unit can carry out analysis result parameter is sentenced with obtaining the intention of user
Next voice that the type of service of disconnected user demand and/or user may input.Wherein, parameter can be according to sentence knot
Keyword determined by structure and grammer, for example, voice input by user be " I will withdraw the money ", by speech pattern recognition unit with
And after the processing of linguistic interpretation unit, the type of service that dialogue stream control unit can obtain user demand is withdrawal business,
And judge next voice that user may input for the specific amount of money of withdrawal.The parameter of extraction can be " wanting " and " take
Money ".Dialogue stream control unit may include type of service database and vocabulary, store what system can be handled in database
Type of service can obtain the mapping relations between surface layer word and Deep Semantics by vocabulary.
The utility model embodiment two provides a kind of information interaction system based on speech recognition, and user not only can be with
Business handling is carried out manually according to the prompt of financial self-service equipment, and during business handling, user can also be with self-service finance
Equipment engages in the dialogue, and disclosure satisfy that particularly with the user for partly having hand disease by handling for interactive voice finishing service
It improves the intelligence degree of financial self-service equipment, improves user's body by the demand of financial self-service equipment transacting business
It tests.
Embodiment three
Fig. 3 is a kind of structural representation for information interaction system based on speech recognition that the utility model embodiment three provides
Figure, the utility model embodiment three is optimized based on above-described embodiment, specifically to the information based on speech recognition
Interactive system structure advanced optimizes, as shown in figure 3, the system of the utility model embodiment three specifically includes:Speech recognition
Module 310, processor 320, service module 330, speech production module 340 and display module 350.
The sound identification module 310 is based on the voice, identifies the need of user for receiving voice input by user
Type of service is sought, and the demand type of service of the user is sent to the processor.
The processor 320, the demand type of service of the user for obtaining sound identification module transmission;Obtain with it is described
The corresponding business processing flow of demand type of service, wherein the business processing flow includes:With the demand type of service pair
Logical relation between the whole business processing steps and different business processing links answered;It obtains and the business processing stream
The corresponding service model of journey integrates result, wherein the service model integrates result and includes:Clothes corresponding with business processing step
The logical relation being engaged between model, and different service models;It is defeated according to the business processing step and/or user that are presently in
The interactive information entered, determines destination service model in the service model integrates result, control display module show with it is described
The business processing interface of destination service Model Matching, and the destination service model is sent to speech production module, so that institute
Predicate sound generation module exports interactive voice corresponding with the destination service model.
Service module 330, the service module are used to, according to the business processing flow, collect the service model
At operation, service model is integrated into result and is sent to the processor;
The processor is additionally operable to:The business processing flow is sent to the server, and receives server and is based on institute
The service model for stating business processing flow return integrates result.
Preferably, the service module 330 includes:
Mapping relations acquiring unit, for obtaining the mapping between pre-stored service model and the business processing flow
Relationship;
Model extraction unit, for according to the mapping relations, extracting business processing step in the business processing flow
Corresponding service model;
Model integrated unit is used for according to the logical relation between different service models, at extraction, with the business
The service model of reason flow path match carries out integrated operation, obtains service model and integrates result.
Specifically, processor after getting business processing flow, sends business processing flow to service module, by taking
Module of being engaged in completes the integrated operation to pre-stored at least one service model according to business processing flow, and by service model
Integrated result is back to processor.At least one service model being pre-stored in service module all has its corresponding purposes category
Property, purposes attribute, you can with the specific service content provided to the user.Each difference of service model and business processing flow
Business processing step between there are mapping relations, by mapping relations acquiring unit, obtain the mapping relations, and pass through model
Extraction unit, extract business processing flow in the matched at least one service model of business processing step.According to different services
The logical relation between logical relation and/or business processing step between model, at least one service model of extraction into
The integrated operation of row, i.e., according to the corresponding service order of business processing flow, by the corresponding partial service split of business processing step
At the integrity service needed for user, obtain the integrated of service model as a result, and service model integrated into result be sent to processor,
Processor integrates result according to service model and determines destination service model.
The speech production module 340 generates and the destination service model for receiving the destination service model
The interactive voice matched exports the interactive voice to user.
The display module 350, for showing and the business processing interface of the destination service Model Matching.
The utility model embodiment three provides a kind of information interaction system based on speech recognition, by identifying user's
Demand type of service obtains the corresponding business handling flow of demand type of service, and then provides and business handling flow path match
Service achievees the purpose that user's finishing service is helped to handle with this.
Example IV
Fig. 4 is a kind of structural representation for information interaction system based on speech recognition that the utility model embodiment four provides
Figure, the utility model embodiment four is optimized based on above-described embodiment, specifically to the information based on speech recognition
Interactive system structure advanced optimizes, as shown in figure 4, the system of the utility model embodiment four specifically includes:Speech recognition
Module 410, processor 420, task control module 430, service module 440, communication module 450, speech production module 460 and
Display module 470.
The sound identification module 410 is based on the voice, identifies the need of user for receiving voice input by user
Type of service is sought, and the demand type of service of the user is sent to the processor.
The processor 420, the demand type of service of the user for obtaining sound identification module transmission;Obtain with it is described
The corresponding business processing flow of demand type of service, wherein the business processing flow includes:With the demand type of service pair
Logical relation between the whole business processing steps and different business processing links answered;It obtains and the business processing stream
The corresponding service model of journey integrates result, wherein the service model integrates result and includes:Clothes corresponding with business processing step
The logical relation being engaged between model, and different service models;It is defeated according to the business processing step and/or user that are presently in
The interactive information entered, determines destination service model in the service model integrates result, control display module show with it is described
The business processing interface of destination service Model Matching, and the destination service model is sent to speech production module, so that institute
Predicate sound generation module exports interactive voice corresponding with the destination service model.
The task control module 430, for if it is determined that the demand type of service of the user is received, according to described
Demand type of service and pre-stored pending business information determine the corresponding business processing flow of the demand type of service,
The business processing flow is sent to the processor, wherein pending business information includes each industry in all types of service
Logical relation between the corresponding whole business processing steps of service type and different business processing links;
The processor is additionally operable to:The demand type of service of the user is sent to the task control module.
Specifically, after processor is if it is determined that receive demand type of service, demand type of service is sent to task
Control module.Task control module is pre-stored with pending business information, and pending business information includes that system is manageable
Logical relation between the corresponding all business processing steps of whole types of service and different business processing links.By required
Type of service and pending business information determine the business processing flow and the corresponding industry of business processing flow of system jointly
Be engaged in processing links, can also include business processing step in business processing step and different business processing step between patrol
The relationship of collecting.Task control module will include that the business processing flow of business processing step is sent to processor.
Service module 440, the service module are used to, according to the business processing flow, collect the service model
At operation, service model is integrated into result and is sent to the processor;
The processor is additionally operable to:The business processing flow is sent to the server, and receives server and is based on institute
The service model for stating business processing flow return integrates result.
Preferably, the service module includes:
Mapping relations acquiring unit, for obtaining the mapping between pre-stored service model and the business processing flow
Relationship;
Model extraction unit, for according to the mapping relations, extracting business processing step in the business processing flow
Corresponding service model;
Model integrated unit is used for according to the logical relation between different service models, at extraction, with the business
The service model of reason flow path match carries out integrated operation, obtains service model and integrates result.
Specifically, service module completes the integrated behaviour to pre-stored at least one service model according to business processing flow
Make, and the integrated result of service model is back to processor.
The communication module 450, the communication module are connected with the server, are used for the speech pattern recognition list
The voice input by user that member receives is sent to the server, and server is based on the voice meter input by user
The matching result of the voice input by user and pattern model calculated is back to the speech pattern recognition unit;
The speech pattern recognition unit is additionally operable to:The matching result is received by communication module.
Specifically, communication module one end is connected with the server in distal end, the other end is connected with system.The service of distal end
Device can when the speech pattern recognition unit of sound identification module receives voice input by user with memory module model
Voice input by user to be sent to the server of distal end by communication module, server is to voice input by user and pattern
Model carries out matching operation, obtains matching result, and matching result is back to the speech pattern recognition list of sound identification module
Member.
Preferably, the communication module also particularly useful for:
The destination service model is sent to the server, and the server is based on the destination service mould
Type, determining business processing interface are back to the processor;
The processor is additionally operable to:The business processing interface is received by the communication module.
Specifically, destination service model is sent to by processor after determining destination service model, by communication module
The server of distal end, server are based on destination service model and provide business processing interface, by communication module by business processing circle
Face is back to processor.The data that user is inputted based on business processing interface can also be sent to far-end server by processor,
The operations such as the transmission of result data that the verification of execution user information and customer service are handled.
The speech production module 460 generates and the destination service model for receiving the destination service model
The interactive voice matched exports the interactive voice to user.
The display module 470, for showing and the business processing interface of the destination service Model Matching.
Preferably, the system is applied to financial self-service equipment.
In the present embodiment, the information interaction system based on speech recognition can be applied to financial self-service equipment, for example, self-service
Cash dispenser.ATM can confirmly detect user there are when, activate sound identification module, pass through speech recognition mould
Block performs the encoding operation voice input by user, and carries out voice, semantic analysis, identifies the demand service class of user
Type, and next voice content that user is intended to input can be obtained.Demand type of service is sent to task control by processor
Type of service according to demand and pre-stored pending business information are determined system by module, task control module jointly
Logic between the business processing step that business processing flow and business processing flow include and different business processing links is closed
System is sent to processor, wherein is the relationship of parent and subclass between business processing flow and business processing step.Processor will
Business processing flow is sent to service module, and service module completes the collection to pre-stored service model according to business processing flow
It is back to processor at operation, and by the integrated result of service model.Processor according to the business processing step being presently in,
And/or interactive information input by user, destination service model is determined in service model integrates result, and destination service model is sent out
It send to speech production module, and destination service model is sent to by far-end server, speech production module by communication module
The corresponding interactive voice of destination service model is generated, and is exported to user, realizes the dialogue with user.Far-end server provide with
Business handling interface is back to processor, processor by the business handling interface of destination service Model Matching by communication module
Control display module display business handling interface is to user, while to realize that ATM carries out speech exchange with user,
Show the operation of business handling interface for users.
The utility model embodiment four provides a kind of information interaction system based on speech recognition, by self-service finance
The function that man machine language's interaction is realized in equipment meets the need that user passes through speech exchange transacting business with financial self-service equipment
It asks, improves the intelligence degree of financial self-service equipment, the user experience is improved.
Embodiment five
The side that Fig. 5 is applicable in by a kind of information interaction system based on speech recognition that the utility model embodiment five provides
The flow chart of method, the present embodiment can be applied to the information interaction system based on speech recognition of the utility model embodiment offer
In, suitable for carrying out voice transacting business situation by the information interaction system based on speech recognition, this method can be by one
The information interactive device based on speech recognition is planted to execute, which may be used software and/or the mode of hardware is realized, generally
It is integrated in the processor of the information interaction system based on speech recognition.
The method of the utility model embodiment five specifically includes:
S510, judge whether to get the demand type of service of the user of sound identification module transmission, if so, executing
Otherwise S520 continues to execute S510.
Specifically, receiving voice input by user by sound identification module, sound identification module is to language input by user
Sound carries out the identification and analysis of voice, syntax and semantics, obtains the type of service of user demand, that is, obtains user and use self-service gold
Melt the type of service that equipment is wanted to handle, wherein type of service includes transferred account service type, withdrawal type of service and deposit industry
Service type etc..The type of service for the user demand that sound identification module will identify that is sent to processor, and processor judges whether
Receive the demand type of service.
S520, judge whether to get business processing flow corresponding with the demand type of service, if so, executing
Otherwise S530 continues to execute S520, wherein the business processing flow includes:Whole corresponding with the demand type of service
Logical relation between business processing step and different business processing links.
Specifically, if processor receives demand type of service, obtain at business corresponding with demand type of service
Flow is managed, type of service and pending business information determine business processing flow jointly according to demand, wherein pending business
Information includes the corresponding all business processing steps of the manageable whole type of service of system and different business processing links
Between logical relation, pending business information can store in the processor, can also be stored in task control module, phase
Ying Di, business processing flow can be directly obtained by processor, and processing is sent to after can also being obtained by task control module
Device, the utility model are without limitation.Wherein, business processing flow includes the corresponding whole business processings of required type of service
Logical relation between link and different business processing links.Processor judges whether to get demand type of service correspondence
Business processing flow.
S530, judge whether that getting service model corresponding with the business processing flow integrates as a result, if so, holding
Otherwise row S540 continues to execute S530, wherein the service model integrates result and includes:Clothes corresponding with business processing step
The logical relation being engaged between model, and different service models.
Specifically, if processor gets the corresponding business processing flow of demand type of service, business processing is obtained
The corresponding service model of flow integrates result.Service model integrates result and can be directly obtained by processor, can also pass through clothes
Business module is sent to processor after obtaining, correspondingly, service model can store in the processor, can also be pre-stored in service
In module, the utility model is without limitation.According to reflecting for service model and the business processing step in business processing flow
Penetrate relationship, according between different service models logical relation and/or business processing step between logical relation, to service mould
Type carries out integrated operation, obtains service model and integrates result.Processor judges whether to get corresponding with business processing flow
Service model integrates result.
The business processing step and/or interactive information input by user that S540, basis are presently in, in the service mould
Type, which integrates, determines destination service model in result.
S550, control display module show and the business processing interface of the destination service Model Matching, and by the mesh
Mark service model is sent to speech production module, so that speech production module output is corresponding with the destination service model
Interactive voice.
Specifically, if processor get the corresponding service model of business processing flow it is integrated as a result, if obtain it is current
Residing business processing step, and according to the business processing step and/or interactive information input by user being presently in, taking
Destination service model is determined in model integrated result of being engaged in, wherein interactive information input by user can be voice input by user
Interactive information, or the interactive information that user is inputted by showing the business handling interface that screen provides, the utility model
It is without limitation.Interactive information is sent to speech production module by processor, and speech production module generates and destination service mould
The matched interactive voice of type, and interactive voice is exported to user.At the same time, it is corresponding to obtain destination service model for processor
Business handling interface, business handling interface can directly be provided by processor, can also be by the communication module of system by target
Service model is sent to far-end server, the business handling interface that far-end server is returned based on destination service model, this practicality
It is novel without limitation.The business handling interface that display module video-stream processor provides.
The utility model embodiment five provides a kind of method that the information interaction system based on speech recognition is applicable in, real
Good human-computer interaction function is showed, can accurately identify the demand of user, and provide the service of business handling to the user, carry
The intelligence degree of financial self-service equipment is risen, meanwhile, it is capable to meet user by carrying out speech exchange with financial self-service equipment
The demand of transacting business, effectively the user experience is improved.
Embodiment six
Fig. 6 is a kind of structural representation of information interactive device based on speech recognition in the utility model embodiment six
Figure, the device are applied to through the information interaction system based on speech recognition come the case where carrying out voice transacting business, the device
The mode that software and/or hardware may be used is realized, the processor of the information interaction system based on speech recognition is generally integrated in
In.As shown in fig. 6, device includes:Type of service acquisition module 610, business processing flow acquisition module 620, service model collection
At result acquisition module 630 and destination service model determining module 640.
Type of service acquisition module 610, the demand type of service of the user for obtaining sound identification module transmission;
Business processing flow acquisition module 620, for obtaining business processing flow corresponding with the demand type of service,
Wherein, the business processing flow includes:Whole business processing step corresponding with the demand type of service, and it is not of the same trade or business
Logical relation between processing links of being engaged in;
Service model integrates result acquisition module 630, for obtaining service model corresponding with the business processing flow
Integrated result, wherein the service model integrates result and includes:Service model corresponding with business processing step, and it is different
Logical relation between service model;
Destination service model determining module 640, for according to the business processing step being presently in and/or user's input
Interactive information, destination service model is determined in the service model integrates result, control display module show and the mesh
The matched business processing interface of service model is marked, and the destination service model is sent to speech production module, so that described
Speech production module exports interactive voice corresponding with the destination service model.
In the present embodiment, processor obtains sound identification module by type of service acquisition module and is based on language input by user
The demand type of service of user acquired in sound, wherein type of service includes transferred account service type, withdrawal type of service and deposits
Money type of service etc..Processor obtains business processing stream corresponding with demand type of service by business processing flow acquisition module
Journey, type of service and pending business information determine business processing flow jointly according to demand, wherein pending business information
Including between the corresponding all business processing steps of the manageable whole type of service of system and different business processing links
Logical relation, pending business information can store in the processor, can also be stored in task control module, therefore,
Business processing flow can be directly obtained by processor, and processor is sent to after can also being obtained by task control module,
In, business processing flow include the corresponding whole business processing steps of required type of service and different business processing links it
Between logical relation.Processor integrates result acquisition module by service model and obtains service mould corresponding with business processing flow
Type is integrated as a result, service model can store in the processor, can also be pre-stored in service module, therefore, service model
Integrated result can be directly obtained by processor, and processor is sent to after can also being obtained by service module.According to different clothes
The logical relation between logical relation and/or business processing step between business model, carries out integrated operation to service model, obtains
Service model is taken to integrate result.Processor is believed by obtaining the business processing step being presently in and/or interaction input by user
Breath determines destination service model, and offer and target by destination service model determining module in service model integrates result
The corresponding business handling interface of service model, control display module shows the business processing interface, and destination service model is sent out
It send to speech production module, so that the speech production module exports corresponding with destination service model interactive voice to user,
Realize the exchange and interdynamic with user.
The utility model embodiment six provides a kind of information interactive device based on speech recognition, and user is in business handling
In the process, it can engage in the dialogue with financial self-service equipment, by handling for interactive voice finishing service, particularly with partly there is hand
The user of portion's disease disclosure satisfy that its demand by financial self-service equipment transacting business, improve the intelligence of financial self-service equipment
Degree can be changed, improve user experience.
It is arbitrary that the information interactive device based on speech recognition that the utility model embodiment provides can perform the utility model
The information interacting method based on speech recognition that embodiment provides, has the corresponding function module of execution method and advantageous effect.
Note that above are only the preferred embodiment and institute's application technology principle of the utility model.Those skilled in the art's meeting
Understand, the utility model is not limited to specific embodiment described here, can carry out for a person skilled in the art various bright
Aobvious variation is readjusted and is substituted without departing from the scope of protection of the utility model.Therefore, although passing through above example
The utility model is described in further detail, but the utility model is not limited only to above example, is not departing from
Can also include other more equivalent embodiments in the case that the utility model is conceived, and the scope of the utility model is by appended
Right determine.
Claims (8)
1. a kind of information interaction system based on speech recognition, which is characterized in that including:Sound identification module, processor, voice
Generation module and display module, wherein
The sound identification module is based on the voice, identifies the demand service class of user for receiving voice input by user
Type, and the demand type of service of the user is sent to the processor;
The processor is used for the demand type of service of the user;It obtains at business corresponding with the demand type of service
Manage flow;
The speech production module, for engaging in the dialogue with user;
The display module, for providing business processing interface to the user.
2. system according to claim 1, which is characterized in that the sound identification module includes:Speech pattern recognition list
Member, linguistic interpretation unit and dialogue stream control unit, wherein
The speech pattern recognition unit obtains the voice input by user and pattern for receiving voice input by user
The matching result of model, and the matching result is sent to the linguistic interpretation unit, wherein the pattern model is according to receipts
Surface layer, grammer layer and the keyword of the business operation term of collection and build;
The linguistic interpretation unit for executing encoding operation to the matching result, and carries out language analysis to coding result,
After being verified to analysis result, the analysis result is sent to the dialogue stream control unit, wherein the language point
Analysis includes sentence structure analysis and semantic analysis;
The dialogue stream control unit, the demand type of service for identifying user according to the analysis result.
3. system according to claim 1, which is characterized in that the system also includes:
Task control module, the task control module are used for if it is determined that receiving the demand type of service of the user, root
It is determined at the corresponding business of the demand type of service according to the demand type of service and pre-stored pending business information
Flow is managed, sends the business processing flow to the processor, wherein pending business information includes in all types of service
Logical relation between the corresponding whole business processing steps of each type of service and different business processing links;
The processor is additionally operable to:The demand type of service of the user is sent to the task control module.
4. system according to claim 1, which is characterized in that the system also includes:
Service module, the service module are used for according to the business processing flow, and integrated operation is carried out to the service model,
Service model is integrated into result and is sent to the processor;
The processor is additionally operable to:The business processing flow is sent to the server, and receives server and is based on the industry
The service model that process flow of being engaged in returns integrates result.
5. system according to claim 4, which is characterized in that the service module includes:
Mapping relations acquiring unit is closed for obtaining the mapping between pre-stored service model and the business processing flow
System;
Model extraction unit is corresponded to for according to the mapping relations, extracting business processing step in the business processing flow
Service model;
Model integrated unit is used for according to the logical relation between different service models, to extraction, with the business processing stream
The matched service model of journey carries out integrated operation, obtains service model and integrates result.
6. system according to claim 2, which is characterized in that the system also includes:
Communication module, the communication module are connected with server, the use for receiving the speech pattern recognition unit
The voice of family input is sent to the server, and the user that server is calculated based on the voice input by user is defeated
The matching result of the voice and pattern model that enter is back to the speech pattern recognition unit;
The speech pattern recognition unit is additionally operable to:The matching result is received by communication module.
7. system according to claim 6, which is characterized in that the communication module also particularly useful for:
Destination service model is sent to the server, and the server is based on the destination service model, is determined
Business processing interface is back to the processor;
The processor is additionally operable to:The business processing interface is received by the communication module.
8. according to any systems of claim 1-7, which is characterized in that the system is applied to financial self-service equipment.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201721156798.4U CN207818190U (en) | 2017-09-11 | 2017-09-11 | A kind of information interaction system based on speech recognition |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201721156798.4U CN207818190U (en) | 2017-09-11 | 2017-09-11 | A kind of information interaction system based on speech recognition |
Publications (1)
Publication Number | Publication Date |
---|---|
CN207818190U true CN207818190U (en) | 2018-09-04 |
Family
ID=63335877
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201721156798.4U Expired - Fee Related CN207818190U (en) | 2017-09-11 | 2017-09-11 | A kind of information interaction system based on speech recognition |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN207818190U (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110211311A (en) * | 2019-07-09 | 2019-09-06 | 中国工商银行股份有限公司 | Voice guidance method, apparatus and system for self-service transacting business |
-
2017
- 2017-09-11 CN CN201721156798.4U patent/CN207818190U/en not_active Expired - Fee Related
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110211311A (en) * | 2019-07-09 | 2019-09-06 | 中国工商银行股份有限公司 | Voice guidance method, apparatus and system for self-service transacting business |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107424601A (en) | A kind of information interaction system based on speech recognition, method and its device | |
CN111488433B (en) | Artificial intelligence interactive system suitable for bank and capable of improving field experience | |
CN107665708B (en) | Intelligent voice interaction method and system | |
CN108089843B (en) | Intelligent bank enterprise-level demand management system | |
JP6859501B2 (en) | A method of starting a business by voice in communication software and a device corresponding to this | |
CN111179935B (en) | Voice quality inspection method and device | |
CN111098312A (en) | Window government affairs service robot | |
CN113688221B (en) | Model-based conversation recommendation method, device, computer equipment and storage medium | |
CN111651571A (en) | Man-machine cooperation based session realization method, device, equipment and storage medium | |
CN106528613A (en) | Intelligent question-answer (Q&A) method and device | |
CN112732871A (en) | Multi-label classification method for acquiring client intention label by robot | |
CN111159375A (en) | Text processing method and device | |
CN110704618A (en) | Method and device for determining standard problem corresponding to dialogue data | |
CN111383138B (en) | Restaurant data processing method, device, computer equipment and storage medium | |
CN207818190U (en) | A kind of information interaction system based on speech recognition | |
JP4143541B2 (en) | Method and system for non-intrusive verification of speakers using behavior models | |
US11386902B2 (en) | System for generation and maintenance of verified data records | |
CN111290677B (en) | Self-service equipment navigation method and navigation system thereof | |
CN115602160A (en) | Service handling method and device based on voice recognition and electronic equipment | |
KR20180042763A (en) | Chatting type financial robot and method for providing financial service using the same | |
CN116883888A (en) | Bank counter service problem tracing system and method based on multi-mode feature fusion | |
CN115309882A (en) | Interactive information generation method, system and storage medium based on multi-modal characteristics | |
CN115688758A (en) | Statement intention identification method and device and storage medium | |
TWM555521U (en) | Hospitality service system | |
CN112598840A (en) | Passing equipment control method and device based on face recognition and voice interaction, machine readable medium and equipment |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20180904 Termination date: 20210911 |
|
CF01 | Termination of patent right due to non-payment of annual fee |