CN106201424A - A kind of information interacting method, device and electronic equipment - Google Patents
A kind of information interacting method, device and electronic equipment Download PDFInfo
- Publication number
- CN106201424A CN106201424A CN201610538666.1A CN201610538666A CN106201424A CN 106201424 A CN106201424 A CN 106201424A CN 201610538666 A CN201610538666 A CN 201610538666A CN 106201424 A CN106201424 A CN 106201424A
- Authority
- CN
- China
- Prior art keywords
- content
- interaction content
- information
- user
- targeted customer
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/16—Sound input; Sound output
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/335—Filtering based on additional data, e.g. user or group profiles
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/08—Speech classification or search
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/08—Speech classification or search
- G10L15/14—Speech classification or search using statistical models, e.g. Hidden Markov Models [HMMs]
- G10L15/142—Hidden Markov Models [HMMs]
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/08—Speech classification or search
- G10L15/16—Speech classification or search using artificial neural networks
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/26—Speech to text systems
Abstract
Embodiments provide a kind of information interacting method, device and electronic equipment, relate to communication and artificial intelligence field.A kind of information interacting method, the method includes: obtain the overall interaction content of multiple targeted customer；Determine the specific interaction content relevant to each targeted customer；Based on a determination that described specific interaction content judge that the targeted customer relevant to described specific interaction content is the need of information alert；The most then search the information mated with described specific interaction content, and show described information at the prompting sight matched to described specific interaction content to the targeted customer relevant with described specific interaction content.The scheme provided by the application, it is possible to the interaction content of multiple targeted customers is analyzed, and information is provided targetedly.
Description
Technical field
The present embodiments relate to communication and artificial intelligence field, particularly relate to a kind of information interacting method, device and electricity
Subset.
Background technology
Along with semanteme resolves and the development of artificial intelligence technology, man-machine interaction (Human-Computer Interaction,
Write a Chinese character in simplified form HCI) in actual life, obtained increasing application, the mode of man-machine interaction include people and electronic equipment (mobile phone,
TV, computer, intelligent robot etc.) between use certain conversational language, with certain interactive mode, for completing to determine task
People and electronic equipment between information exchanging process.Along with the difference of user's request, man-machine interaction is in increasing field
Suffer from wide application.One is partner robot than more typical example application, i.e. partner robot can be with multiple
Mode receives the input (these inputs include but not limited to phonetic entry, word input etc.) of user, and partner robot analysis is used
The input content at family, by searching for the data base of Local or Remote, provides corresponding answer.Such as user wonders the sky of today
Gas how, as long as partner robot is said " Pekinese's weather today ", partner robot will receive the voice content of user, logical
Cross the weather data storehouse of retrieval association, in the way of voice or word, show user's Weather information to user.
Inventor finds during realizing the present invention, and in prior art, interactive process is the most single
User actively initiates, and user needs to initiate to put question to targetedly to the electronic equipment with human-computer interaction function, Intelligent electric
Subset just can provide corresponding answer, for the electronic equipment with human-computer interaction function, this interactive mode by
Move.And in the middle of actually used, in addition to can directly answering the enquirement of user, user more wishes have man-machine interaction
The electronic equipment of function can by a kind of positive in the way of obtain the demand information of user, and according to the demand information master obtained
Dynamic carries out man-machine interaction.
Summary of the invention
The embodiment of the present invention provides a kind of information interacting method, device and electronic equipment, in order to solve use in prior art
Family is during man-machine interaction, and interactive mode is single, the shortcoming that can not meet the interaction demand of complex interaction environment, it is achieved that
Intelligent under complex interaction environment of man-machine interaction.
First aspect, the embodiment of the present invention provides a kind of information interacting method, and the method includes:
Obtain the overall interaction content of multiple targeted customer；
Determine the specific interaction content relevant to each targeted customer；
Based on a determination that described specific interaction content judge the targeted customer relevant to described specific interaction content whether need
Want information alert；
The most then search the information mated with described specific interaction content, and with described specific interaction content phase
The prompting sight of coupling shows described information to the targeted customer relevant to described specific interaction content.
A kind of specific implementation according to embodiments of the present invention, the multiple targeted customer's of described acquisition is overall mutual interior
Hold, including:
Obtain the interactive voice content of more than one user, based on described interactive voice content, determine that described entirety is mutual
Content.
A kind of specific implementation according to embodiments of the present invention, described based on described interactive voice content, determine described
Overall interaction content, including:
Obtain the context characteristic corresponding with described interactive voice content；
Obtain the context characteristic data base being associated with described context characteristic；
Described interactive voice content is carried out speech recognition, and by the content after speech recognition and described context characteristic data
Context characteristic pattern in storehouse is mated, and obtains described overall interaction content.
A kind of specific implementation according to embodiments of the present invention, described based on described interactive voice content, determine described
Overall interaction content, also includes:
Extract the vocal print characteristic information in described interactive voice content；
Obtain the user feature database being associated with described vocal print characteristic information；
Described interactive voice content is carried out speech recognition, and by the content after speech recognition and described user characteristic data
Content in storehouse is mated, and obtains described overall interaction content.
A kind of specific implementation according to embodiments of the present invention, described determines the specific friendship relevant to each targeted customer
Content mutually, including:
Extract the vocal print characteristic information in described interactive voice content；
The each targeted customer relevant to described interactive voice content is determined according to described vocal print characteristic information；
Described overall interaction content is split with described each targeted customer, obtains and described each targeted customer's phase
The specific interaction content closed.
A kind of specific implementation according to embodiments of the present invention, described by described overall interaction content and described each mesh
Mark user splits, and obtains the specific interaction content relevant to described each targeted customer, including:
Based on described vocal print characteristic information, described overall interaction content is split；
Obtain the user feature database corresponding with described vocal print characteristic information；
The content obtained after being split by described overall interaction content is mated with described user feature database, obtain and
The specific interaction content that described each targeted customer is correlated with.
A kind of specific implementation according to embodiments of the present invention, the overall interaction content of the plurality of targeted customer, bag
Include:
Other users of active user and more than one are carried out: in voice call content, mail interaction content, short message interacting
More than one in appearance, talk interaction content, instant messaging word interaction content.
A kind of specific implementation according to embodiments of the present invention, described with carrying that described specific interaction content matches
Show that sight shows described information to the targeted customer relevant to described specific interaction content, including:
Determine the prompting scene mated with described interaction content；
Judge the matching degree of the prompting scene whether current scene mate with described interaction content；
When described matching degree is more than predetermined threshold value, show described information in default information alert mode to user.
A kind of specific implementation according to embodiments of the present invention, described with carrying that described specific interaction content matches
Show that sight, after the targeted customer relevant to described specific interaction content shows described information, also includes:
Receive the feedback information of the targeted customer input relevant to described specific interaction content；And
Determine whether to proceed information alert based on described feedback information.
Second aspect, the embodiment of the present invention also provides for a kind of information interactive device, and this device includes:
Acquisition module, for obtaining the overall interaction content of multiple targeted customer；
Determine module, for determining the specific interaction content relevant to each targeted customer；
Judge module, for based on a determination that described specific interaction content judge the mesh relevant to described specific interaction content
Mark user is the need of information alert；
Reminding module, for when the targeted customer relevant to described specific interaction content needs information alert, search and
The information of described specific interaction content coupling, and at the prompting sight matched with described specific interaction content to described
The targeted customer that specific interaction content is relevant shows described information.
A kind of specific implementation according to embodiments of the present invention, described acquisition module is additionally operable to:
Obtain the interactive voice content of more than one user, based on described interactive voice content, determine that described entirety is mutual
Content.
A kind of specific implementation according to embodiments of the present invention, described based on described interactive voice content, determine described
Overall interaction content, including:
Obtain the context characteristic corresponding with described interactive voice content；
Obtain the context characteristic data base being associated with described context characteristic；
Described interactive voice content is carried out speech recognition, and by the content after speech recognition and described context characteristic data
Context characteristic pattern in storehouse is mated, and obtains described overall interaction content.
A kind of specific implementation according to embodiments of the present invention, described based on described interactive voice content, determine described
Overall interaction content, also includes:
Extract the vocal print characteristic information in described interactive voice content；
Obtain the user feature database being associated with described vocal print characteristic information；
Described interactive voice content is carried out speech recognition, and by the content after speech recognition and described user characteristic data
Content in storehouse is mated, and obtains described overall interaction content.
A kind of specific implementation according to embodiments of the present invention, described determines that module is additionally operable to:
Extract the vocal print characteristic information in described interactive voice content；
The each targeted customer relevant to described interactive voice content is determined according to described vocal print characteristic information；
Described overall interaction content is split with described each targeted customer, obtains and described each targeted customer's phase
The specific interaction content closed.
A kind of specific implementation according to embodiments of the present invention, described by described overall interaction content and described each mesh
Mark user splits, and obtains the specific interaction content relevant to described each targeted customer, including:
Based on described vocal print characteristic information, described overall interaction content is split；
Obtain the user feature database corresponding with described vocal print characteristic information；
The content obtained after being split by described overall interaction content is mated with described user feature database, obtain and
The specific interaction content that described each targeted customer is correlated with.
A kind of specific implementation according to embodiments of the present invention, the overall interaction content of the plurality of targeted customer, bag
Include:
Other users of active user and more than one are carried out: in voice call content, mail interaction content, short message interacting
More than one in appearance, talk interaction content, instant messaging word interaction content.
A kind of specific implementation according to embodiments of the present invention, described reminding module is additionally operable to:
Determine the prompting scene mated with described interaction content；
Judge the matching degree of the prompting scene whether current scene mate with described interaction content；
When described matching degree is more than predetermined threshold value, show described information in default information alert mode to user.
A kind of specific implementation according to embodiments of the present invention, described information interactive device also includes feedback module, institute
State feedback module for:
Use to the target relevant with described specific interaction content at the prompting sight matched to described specific interaction content
After described information is shown at family, receive the feedback information of the targeted customer input relevant to described specific interaction content；With
And
Determine whether to proceed information alert based on described feedback information.
The third aspect, the embodiment of the present invention provides a kind of electronic equipment, and described electronic equipment includes: housing, processor, deposit
Reservoir, circuit board and power circuit, wherein, circuit board is placed in the interior volume that housing surrounds, processor and memorizer and arranges
On circuit boards；Power circuit, powers for each circuit or the device for above-mentioned electronic equipment；Memorizer is used for storing can be held
Line program code；Processor runs and executable program code pair by reading the executable program code of storage in memorizer
The program answered, for performing aforementioned arbitrary described information interacting method
Information interacting method, device and the electronic equipment that the embodiment of the present invention provides, by carrying out entirety interaction content
Split targetedly, by the determination of effective interaction scenarios, it is possible to targetedly to specific user under specific scene
The information of carrying out is mutual, solves the shortcoming that interactive mode in prior art is the most single, cannot adapt to complex interaction environment, improves
Mutual efficiency.
Accompanying drawing explanation
In order to be illustrated more clearly that the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing
In having technology to describe, the required accompanying drawing used is briefly described, it should be apparent that, the accompanying drawing in describing below is this
Some bright embodiments, for those of ordinary skill in the art, on the premise of not paying creative work, it is also possible to root
Other accompanying drawing is obtained according to these accompanying drawings.
The schematic flow sheet of a kind of information interacting method that Fig. 1 provides for the embodiment of the present invention；
Fig. 2 obtains overall interaction content flow process schematic diagram for what the embodiment of the present invention provided based on interactive voice mode；
Fig. 3 carries out the signal of interactive voice content recognition flow process for the vocal print feature based on user that the embodiment of the present invention provides
Figure；
Fig. 4 provides a kind of vocal print characteristic information based on user to determine the spy relevant to targeted customer for the embodiment of the present invention
Determine the schematic flow sheet of interaction content；
Fig. 5 provides another kind of vocal print characteristic information based on user to determine relevant to targeted customer for the embodiment of the present invention
The schematic flow sheet of specific interaction content；
The schematic flow sheet showing information based on scene to user that Fig. 6 provides for the embodiment of the present invention；
The schematic flow sheet carrying out information alert based on user feedback content that Fig. 7 provides for the embodiment of the present invention；
The structural representation of a kind of information interactive device that Fig. 8 provides for the embodiment of the present invention；
The structural representation of the another kind of information interactive device that Fig. 9 provides for the embodiment of the present invention；
A kind of electronic devices structure schematic diagram that Figure 10 provides for the embodiment of the present invention.
Detailed description of the invention
For making the purpose of the embodiment of the present invention, technical scheme and advantage clearer, below in conjunction with the embodiment of the present invention
In accompanying drawing, the technical scheme in the embodiment of the present invention is clearly and completely described, it is clear that described embodiment is
The a part of embodiment of the present invention rather than whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art
The every other embodiment obtained under not making creative work premise, broadly falls into the scope of protection of the invention.
Interactive mode of the prior art is usually fixed the interactive mode of formula, and mutual initiator is typically all and uses
Family, and interactive device is typically the passive input content receiving user, and then the input to user is answered.As one
Example, user can inquire mobile phone assistant (Siri built-in in such as Iphone): " making a phone call to John ", at this time mobile phone helps
Hands will dial the telephone number of John, and this interactive mode is the most single, and more often, user wishes interactive object
(such as mobile phone assistant) can the behavioural information obtaining user actively, and then actively provide the user with information alert.
Fig. 1 is the schematic flow sheet of information interacting method in the embodiment of the present invention, and the method includes:
S101: obtain the overall interaction content of multiple targeted customer.
Before carrying out the overall interaction content of acquisition user, needing, overall interaction content is set and obtains program.Overall friendship
Content retrieval program can exist in a variety of forms mutually, and the information being such as arranged in subscriber computer (such as receives in mail
Hold) monitoring programme or be arranged in mobile phone cellphone information (note, message registration, the chat record etc. of mobile phone A pp) prison
Control journey or individually designed robot device that can monitor and record more than one user's chat content.
Can obtain the overall interaction content of targeted customer in several ways, a kind of mode is that user actively opens
Overall interaction content obtains operation switch, and now overall interaction content obtains program and will obtain user's (ratio within a period of time
Such as one week) overall interaction content, these interaction contents comprise the mailbox contents of a period of time content of user, user online
Chat record that search record, the shopping record of user, the message registration of user, user are carried out by JICQ, use
The content etc. made a phone call in family.
In addition to user actively opens the acquisition mode that overall interaction content obtains operation switch, it is also possible to use whole
Body interaction content obtains the way of program active obtaining, i.e. after user opens overall interaction content acquisition function, overall
Interaction content obtains the interaction content of program just moment monitoring user.
Concrete, overall interaction content obtains program and can be built in multiple equipment of user in the way of using software,
Overall interaction content on multiple equipment of user obtains program and uses same ID users carry out acquisition of information and collect.
Overall interaction content includes the overall interaction content of active user and one or more users, such as active user and
The content that one or more users are linked up by mail, or the language that active user sits on other colleagues in meeting room
Sound content, or the chat record that active user is in QQ group and between multiple QQ good friend.
S102: determine the specific interaction content relevant to each targeted customer.
Owing to there is multiple interactive object in overall interaction content, it is therefore desirable to the information in overall interaction content is carried out
Splitting, such as active user John have sent job content arrangement, Ada to his subordinate Ada and Bob in mail respectively
Job placement proposes respectively in mail different suggestions with Bob, and scheme is carried out by John suggestion based on Ada and Bob
Amendment, at this time, it may be necessary to determine what the concrete scheme content relevant to John be, in the concrete scheme relevant to Ada and Bob
Holding is what respectively, and these constitute, to the change of each user-dependent concrete scheme content, the specific friendship that each targeted customer is relevant
Content mutually.
Determine that the specific interaction content relevant to each targeted customer has various ways, for example.Mutual for mailbox
Content, can judge the interaction content of each user by the label of sender and recipient；For active user with one
Or the voice-enabled chat record between multiple user, the content inside voice-enabled chat record can be determined by the way of speech recognition
Belong to which voice-enabled chat object, and using the concrete chat content of each voice-enabled chat object as with this voice-enabled chat object phase
The specific interaction content closed.
S103: based on a determination that described specific interaction content judge the targeted customer relevant to described specific interaction content be
No need information alert.
By analyzing the information comprised in specific interaction content, it is possible to obtain information alert key element, user points out key element
It is to point out relevant a particular content to user, such as, the scheduling information of user.With traditional user's driving arrangement day
Unlike journey information, the method can carry out semantic analysis according to the information of content-data, and then obtain whether user needs
Want information.Such as user James mentions in specific interaction content that " weather will be pretty good tomorrow, and Steven, we want one
Rise and go to climb the mountain ", at this moment determine and likely to have an appointment with Steven in tomorrow for James, determine and need user James is entered
Row information alert, further, it is also possible to send information alert to Steven.
S104: the most then search the information mated with described specific interaction content, and with described specific alternately in
Hold the prompting sight matched and show described information to the targeted customer relevant to described specific interaction content.
After determining that user needs to carry out information alert, it is thus necessary to determine that the particular content of the information of user, Yi Ji
Under which kind of sight, user is carried out information alert.
The particular content of user's information, refers to the specifying information content pointing out with user key element to want to mate, such as uses
The user of family James points out key element to be " tomorrow goes to climb the mountain ", then include to the information mated with " tomorrow goes to climb the mountain ": inquiry is used
Family concrete departure time, recommend than place of climbing the mountain conveniently to user, advise the points for attention climbed the mountain to user, go to climb
The information such as the vehicles that mountain region point is feasible.
In addition to the particular content determining information, in addition it is also necessary to judge that prompting scene, prompting scene are directed to target
User sends the key elements such as information most suitable time, position.Such as user David about Mary 20:00 tonight exists
Rose hotel meets, the stroke needing 30 minutes to Rose hotel from the departure place of David, it is contemplated that 10 minutes in advance to Rose
Hotel, 17:20 when, has constituted information time prompting scene, can send out to user David under this scene
Send and show information.Based on same principle, it is also possible to send to Mary and show information.
Show that information has plurality of optional mode, can be opened up by modes such as mail, note, calendar promptings
Show, it is also possible to reminded by the alert device (display screen, partner robot etc.) of special setting.
By the scheme in the present embodiment, it is possible to actively extract the overall interaction content of user, based on interior alternately to entirety
The analysis held, determines concrete suggestion content, and sends information to user under suitably prompting scene, solve user
Need to be specifically designed for reminding item to carry out the problem being separately provided, more comprehensive user interactive environment letter can be obtained simultaneously
Breath, it is provided that more comprehensively information reminding.
As another embodiment, Fig. 2 illustrates the scheme obtaining overall interaction content based on interactive voice mode, should
Scheme comprises the steps:
S201: obtain the interactive voice content of more than one user.
Various ways can be used to obtain the interactive voice content of user, such as built-in interactive voice in the mobile phone of user
Content record program, by the talk record of interactive voice content record program record active user Yu one or more users,
Can also such as arrange and there is the intelligent electronic device of voice record function to record user by the way of software/hardware combines
Interactive voice content.
Alternatively, can be by the interactive voice content of user with specific phonetic matrix (MP3, WAV etc.)
Carry out locally stored, it is also possible to by interactive voice content uploading to server set in advance.
S202: obtain the context characteristic corresponding with described interactive voice content.
During carrying out voice record, can obtain current context characteristic, these context characteristic include but do not limit
In, the chat that the temporal information of interactive voice, positional information and behavioural habits or current chat content according to user determine
Scene.These chat scenarios include: operative scenario, home scenarios, dining scene, moving scene etc..
S203: obtain the context characteristic data base being associated with described context characteristic.
In context characteristic data base, storage has context characteristic, different context characteristic to constitute different contextual models,
This locality of electronic equipment or the server end storage with electronic equipment communication connection have sight property data base, when step S202 is true
After having determined to determine current context characteristic, by the data base mated with this context characteristic at local or server-side search,
It is defined as the context characteristic data base being associated with described context characteristic.
When context characteristic data base is when locally stored, context characteristic data base can be loaded directly into, when context characteristic number
According to warehouse compartment when the server end, then by locally downloading for context characteristic data base, and perform to load operation.
S204: described interactive voice content is carried out speech recognition, and the content after speech recognition is special with described sight
The context characteristic pattern levied in data base is mated, and obtains described overall interaction content.
Before carrying out speech recognition, it usually needs acoustical signal is carried out pretreatment, pretreatment refers in feature extraction
, first raw tone is processed before, partially remove the impact that noise brings with different speakers, make the signal after process more
The substitutive characteristics of voice can be reflected.It is for instance possible to use end-point detection and speech enhan-cement mode carry out pretreatment.End-point detection is
Refer to voice and non-speech audio period be made a distinction in voice signal, accurately determine out the starting point of voice signal.Warp
After crossing end-point detection, voice signal just can only be carried out by subsequent treatment, this degree of accuracy and recognition correct rate to improving model
Play an important role.The mainly elimination environment noise impact on voice of speech enhan-cement, such as uses the mode pair of Wiener filtering
Noise signal filters.
Speech recognition is the process of a kind of pattern recognition, by carrying out the voice of record one by one with preset reference pattern
Compare, and then obtain the result of speech recognition.Multiple voice recognizer can be used to complete speech recognition process, these voices
Recognizer includes: based on dynamic time warping (DTW) algorithm, vector quantization based on nonparametric model (VQ) method, based on
The method of the hidden Markov model (HMM) of parameter model, know based on voices such as artificial neural network (ANN) and support vector machine
Other method etc..
During speech recognition, need the probabilistic model using language model to calculate a sentence probability of occurrence.
It is mainly used in determining that the probability of which word sequence is bigger, or predicts that in the case of occurring in that several word the next one will
The content of the word occurred, carries out grammer, semantic analysis to training text data base, through obtaining language based on statistical model training
Speech model.And using different training text data bases, it identifies that the result obtained also can be different, the side in the embodiment of the present invention
Case is then to carry out Secondary Match by the language content after identifying with the content in context characteristic data base, finally gives accurately
Spend more preferable voice identification result.Wherein, during carrying out speech recognition, need to obtain spy from context characteristic data base
Determine context characteristic, the interaction content currently obtained is mated with context characteristic pattern, thus judge current exchange scenario.
By the way of context characteristic data base being mated with identification content, carry out speech recognition, improve voice and know
Other accuracy.
Fig. 3 illustrates vocal print feature based on user and carries out the content aware scheme of interactive voice, and the scheme shown in Fig. 3 can
To merge execution with the scheme shown in Fig. 2, it is also possible to as an independent scheme, individually perform, the program includes walking as follows
Rapid:
S301: extract the vocal print characteristic information in described interactive voice content.
The task of vocal print feature extraction is to extract and select to have the vocal print of speaker that separability is strong, the high spy of stability
The acoustics of property or language feature.Vocal print feature has the content of many aspects, goes out from the angle utilizing mathematical method to model
Sending out, vocal print identifies that the feature that model can use includes automatically: (1) acoustic features (cepstrum)；(2) lexical characteristics (speaker
Relevant word n-gram, phoneme n-gram)；(3) prosodic features (utilizing fundamental tone and energy " posture " that n-gram describes)；(4)
Languages, dialect and accent information；(5) channel information (using which kind of passage).
Interactive voice content is usually present the vocal print characteristic information of multiple user, this carry out vocal print feature extraction time
Wait, need to extract the vocal print characteristic information of multiple user.
S302: obtain the user feature database being associated with described vocal print characteristic information.
This locality of electronic equipment or with electronic equipment communication connection server end storage have user feature database, when
After step S302 determines the vocal print feature of current one or more users, by local or server-side search
With the data base of this vocal print characteristic matching, it is defined as the user feature database being associated with described vocal print feature.
When user feature database is when locally stored, user feature database can be loaded directly into, when user characteristics number
According to warehouse compartment when the server end, then user feature database is locally downloading, and perform to load operation.
S303: described interactive voice content is carried out speech recognition, and the content after speech recognition is special with described user
The content levied in data base is mated, and obtains described overall interaction content.
Before carrying out speech recognition, it usually needs acoustical signal is carried out pretreatment, pretreatment refers in feature extraction
, first raw tone is processed before, partially remove the impact that noise brings with different speakers, make the signal after process more
The substitutive characteristics of voice can be reflected.It is for instance possible to use end-point detection and speech enhan-cement mode carry out pretreatment.End-point detection is
Refer to voice and non-speech audio period be made a distinction in voice signal, accurately determine out the starting point of voice signal.Warp
After crossing end-point detection, voice signal just can only be carried out by subsequent treatment, this degree of accuracy and recognition correct rate to improving model
Play an important role.The mainly elimination environment noise impact on voice of speech enhan-cement, such as uses the mode pair of Wiener filtering
Noise signal filters.
Speech recognition is the process of a kind of pattern recognition, by carrying out the voice of record one by one with preset reference pattern
Compare, and then obtain the result of speech recognition.Multiple voice recognizer can be used to complete speech recognition process, these voices
Recognizer includes: based on dynamic time warping (DTW) algorithm, vector quantization based on nonparametric model (VQ) method, based on
The method of the hidden Markov model (HMM) of parameter model, know based on voices such as artificial neural network (ANN) and support vector machine
Other method etc..
During speech recognition, need the probabilistic model using language model to calculate a sentence probability of occurrence.
It is mainly used in determining that the probability of which word sequence is bigger, or predicts that in the case of occurring in that several word the next one will
The content of the word occurred, carries out grammer, semantic analysis to training text data base, through obtaining language based on statistical model training
Speech model.And using different training text data bases, it identifies that the result obtained also can be different, the side in the embodiment of the present invention
Case is then to carry out Secondary Match by the language content after identifying with the content in user feature database, finally gives accurately
Spend more preferable voice identification result.
By the way of user feature database being mated with identification content, carry out speech recognition, further increase
The accuracy of speech recognition.
Fig. 4-5 illustrates vocal print characteristic information based on user and determines the mistake of the specific interaction content relevant to targeted customer
Journey, including:
S401: extract the vocal print characteristic information in described interactive voice content.
The task of vocal print feature extraction is to extract and select to have the vocal print of speaker that separability is strong, the high spy of stability
The acoustics of property or language feature.Vocal print feature has the content of many aspects, goes out from the angle utilizing mathematical method to model
Sending out, vocal print identifies that the feature that model can use includes automatically: (1) acoustic features (cepstrum)；(2) lexical characteristics (speaker
Relevant word n-gram, phoneme n-gram)；(3) prosodic features (utilizing fundamental tone and energy " posture " that n-gram describes)；(4)
Languages, dialect and accent information；(5) channel information (using which kind of passage).
Interactive voice content is usually present the vocal print characteristic information of multiple user, this carry out vocal print feature extraction time
Wait, need to extract the vocal print characteristic information of multiple user.
S402: determine each target same family relevant to described interactive voice content according to described vocal print characteristic information.
By the way of pattern recognition or cluster, the information with identical vocal print feature is divided into a class, and determines each
User corresponding to class vocal print feature.For example, following recognition method can be used to be identified:
(1) template matching method: utilize dynamic time bending (DTW) with alignment training and test feature sequence, mainly use
Application (usually text inter-related task) in fixed phrases；
(2) arest neighbors method: retain all characteristic vectors during training, finds in trained vector each vector during identification
Nearest K, being identified accordingly, the storage of usual model is the biggest with the amount of similar calculating；
(3) neural net method: have a variety of form, such as multilayer perception, RBF (RBF) etc., can explicitly instruct
Practice to distinguish speaker and its background speaker；
(4) hidden Markov model (HMM) method: generally use the HMM of single state, or gauss hybrid models (GMM)；
(5) VQ clustering method (such as LBG): the method effect is pretty good, and algorithm complex is the highest, and HMM method coordinates
More can receive more preferable effect.
S403: described overall interaction content is split with described each targeted customer, obtains and described each target
User-dependent specific interaction content.
Particular user based on the identification in step S402 and the particular content of each particular user association, obtain with described
The specific interaction content that each targeted customer is correlated with.
Further, described overall interaction content is split by step S403 with described each targeted customer, obtain and
The specific interaction content that described each targeted customer is correlated with, also comprises the steps:
S4031: described overall interaction content is split based on described vocal print characteristic information.
S4032: obtain the user feature database corresponding with described vocal print characteristic information.
In user feature database, storage has the data of at least two type: one is user type data, user according to
The characteristic information (such as vocal print characteristic information) of user determines concrete user；Another type of data are the language of particular user
Property data base, after determining concrete user type, loads the voice feature data storehouse corresponding with this user, to specific
The chat content of user is modified.
S4033: the content obtained after being split by described overall interaction content is mated with described user feature database,
Obtain the specific interaction content relevant to described each targeted customer.
By the content of this embodiment, it is possible to overall interaction content is split by vocal print characteristic information based on user,
And decide specific aims user and the specific interaction content of this targeted customer association.
Fig. 6 illustrates the scheme showing information based on scene to user, and the program includes:
S601: determine the prompting scene mated with described interaction content.
In addition to the particular content determining information, in addition it is also necessary to judge that prompting scene, prompting scene are directed to target
User sends the key elements such as information most suitable time, position.Such as user David about Mary 20:00 tonight exists
Rose hotel meets, the stroke needing 30 minutes to Rose hotel from the departure place of David, it is contemplated that 10 minutes in advance to Rose
Hotel, 17:20 when, has constituted information time prompting scene, can send out to user David under this scene
Send and show information.Based on same principle, it is also possible to send to Mary and show information.
S602: judge the matching degree of the prompting scene whether current scene mate with described interaction content.
Not all scene is suitable for sending information to user, now it needs to be determined that current scene and mutual field
The matching degree of scape, such as user David intends to meet in Rose hotel with Mary 20:00 tonight, is determined by current field
Scape finds that David has arrived at Rose hotel, the most just there is no need again to remind David.
Concrete scene matching degree can use many algorithms to realize, such as based on credible matching process, and can
Letter property is divided into reliability, availability and ageing, it is established that the Integrated Model of Credibility Assessment, retouches the most in detail at this
State.
S603: when described matching degree is more than predetermined threshold value, show described prompting in default information alert mode to user
Information.
Show that information has plurality of optional mode, can be opened up by modes such as mail, note, calendar promptings
Show, it is also possible to reminded by the alert device (display screen, partner robot etc.) of special setting.
Fig. 7 gives the flow chart carrying out information alert based on user feedback content, and the program is except comprising corresponding to Fig. 1
Embodiment in content outside, also comprise the steps:
S105: receive the feedback information of the targeted customer input relevant to described specific interaction content.
After having shown information to user, also provide for the interface of user feedback, such as user and mention in talk
Crossing " intending recently to buy a mobile phone ", when there is mobile phone advertising campaign, sending information to user " has X-type number recently
Mobile phone, promotes at a reduced price, if need to buy one？", now user can be by voice or manually enter that " I does not likes
This mobile phone ", as the feedback information at same family.
S106: determine whether to proceed information alert based on described feedback information.
When, after the feedback information receiving user, the input information of user being carried out self study, understanding of the defeated of user
After entering information, it can be determined that the need of continuing in information, such as S105, user is specifically called out this that do not like recommendation
After portion's mobile phone, in the way of continuing with information, other kinds of mobile phone can be shown to user.
By the content of this embodiment, it is possible to feedback information based on user learns again, more to user and
Have more exhibition information targetedly.
Corresponding with the method for Fig. 1 embodiment, embodiments of the invention also disclose a kind of information interactive device, such as Fig. 8
Shown in, information interactive device includes:
Acquisition module 801, for obtaining the overall interaction content of multiple targeted customer.
Before carrying out the overall interaction content of acquisition user, needing, overall interaction content is set and obtains program.Overall friendship
Content retrieval program can exist in a variety of forms mutually, and the information being such as arranged in subscriber computer (such as receives in mail
Hold) monitoring programme or be arranged in mobile phone cellphone information (note, message registration, the chat record etc. of mobile phone A pp) prison
Control journey or individually designed robot device that can monitor and record more than one user's chat content.
Can obtain the overall interaction content of targeted customer in several ways, a kind of mode is that user actively opens
Overall interaction content obtains operation switch, and now overall interaction content obtains program and will obtain user's (ratio within a period of time
Such as one week) overall interaction content, these interaction contents comprise the mailbox contents of a period of time content of user, user online
Chat record that search record, the shopping record of user, the message registration of user, user are carried out by JICQ, use
The content etc. made a phone call in family.
In addition to user actively opens the acquisition mode that overall interaction content obtains operation switch, it is also possible to use whole
Body interaction content obtains the way of program active obtaining, i.e. after user opens overall interaction content acquisition function, overall
Interaction content obtains the interaction content of program just moment monitoring user.
Concrete, overall interaction content obtains program and can be built in multiple equipment of user in the way of using software,
Overall interaction content on multiple equipment of user obtains program and uses same ID users carry out acquisition of information and collect.
Overall interaction content includes the overall interaction content of active user and one or more users, such as active user and
The content that one or more users are linked up by mail, or the language that active user sits on other colleagues in meeting room
Sound content, or the chat record that active user is in QQ group and between multiple QQ good friend.
Determine module 802, for determining the specific interaction content relevant to each targeted customer.
Owing to there is multiple interactive object in overall interaction content, it is therefore desirable to the information in overall interaction content is carried out
Splitting, such as active user John have sent job content arrangement, Ada to his subordinate Ada and Bob in mail respectively
Job placement proposes respectively in mail different suggestions with Bob, and scheme is carried out by John suggestion based on Ada and Bob
Amendment, at this time, it may be necessary to determine what the concrete scheme content relevant to John be, in the concrete scheme relevant to Ada and Bob
Holding is what respectively, and these constitute, to the change of each user-dependent concrete scheme content, the specific friendship that each targeted customer is relevant
Content mutually.
Determine that the specific interaction content relevant to each targeted customer has various ways, for example.Mutual for mailbox
Content, can judge the interaction content of each user by the label of sender and recipient；For active user with one
Or the voice-enabled chat record between multiple user, the content inside voice-enabled chat record can be determined by the way of speech recognition
Belong to which voice-enabled chat object, and using the concrete chat content of each voice-enabled chat object as with this voice-enabled chat object phase
The specific interaction content closed.
Judge module 803, for based on a determination that described specific interaction content judge relevant to described specific interaction content
Targeted customer the need of information alert.
By analyzing the information comprised in specific interaction content, it is possible to obtain information alert key element, user points out key element
It is to point out relevant a particular content to user, such as, the scheduling information of user.With traditional user's driving arrangement day
Unlike journey information, the method can carry out semantic analysis according to the information of content-data, and then obtain whether user needs
Want information.Such as user James mentions in specific interaction content that " weather will be pretty good tomorrow, and Steven, we want one
Rise and go to climb the mountain ", at this moment determine and likely to have an appointment with Steven in tomorrow for James, determine and need user James is entered
Row information alert, further, it is also possible to send information alert to Steven.
Reminding module 804, for when the targeted customer relevant to described specific interaction content needs information alert, searches
The information mated with described specific interaction content, and the prompting sight matched with described specific interaction content to institute
The targeted customer stating specific interaction content relevant shows described information.
After determining that user needs to carry out information alert, it is thus necessary to determine that the particular content of the information of user, Yi Ji
Under which kind of sight, user is carried out information alert.
The particular content of user's information, refers to the specifying information content pointing out with user key element to want to mate, such as uses
The user of family James points out key element to be " tomorrow goes to climb the mountain ", then include to the information mated with " tomorrow goes to climb the mountain ": inquiry is used
Family concrete departure time, recommend than place of climbing the mountain conveniently to user, advise the points for attention climbed the mountain to user, go to climb
The information such as the vehicles that mountain region point is feasible.
In addition to the particular content determining information, in addition it is also necessary to judge that prompting scene, prompting scene are directed to target
User sends the key elements such as information most suitable time, position.Such as user David about Mary 20:00 tonight exists
Rose hotel meets, the stroke needing 30 minutes to Rose hotel from the departure place of David, it is contemplated that 10 minutes in advance to Rose
Hotel, 17:20 when, has constituted information time prompting scene, can send out to user David under this scene
Send and show information.Based on same principle, it is also possible to send to Mary and show information.
Show that information has plurality of optional mode, can be opened up by modes such as mail, note, calendar promptings
Show, it is also possible to reminded by the alert device (display screen, partner robot etc.) of special setting.
By the scheme in the present embodiment, it is possible to actively extract the overall interaction content of user, based on interior alternately to entirety
The analysis held, determines concrete suggestion content, and sends information to user under suitably prompting scene, solve user
Need to be specifically designed for reminding item to carry out the problem being separately provided, more comprehensive user interactive environment letter can be obtained simultaneously
Breath, it is provided that more comprehensively information reminding.
As another embodiment, seeing Fig. 9, information interactive device also includes feedback module 805, feedback module 805
For:
Use to the target relevant with described specific interaction content at the prompting sight matched to described specific interaction content
After described information is shown at family, receive the feedback information of the targeted customer input relevant to described specific interaction content；With
And
Determine whether to proceed information alert based on described feedback information.
By the scheme in the present embodiment, it is possible to actively extract the overall interaction content of user, based on interior alternately to entirety
The analysis held, determines concrete suggestion content, and sends information to user under suitably prompting scene, solve user
Need to be specifically designed for reminding item to carry out the problem being separately provided, more comprehensive user interactive environment letter can be obtained simultaneously
Breath, it is provided that more comprehensively information reminding.
In Fig. 8-9, in embodiment, content performed by modules is relative with the content that various method steps in Fig. 1-7 performs
Should, do not repeat them here.
The embodiment of the present invention also provides for a kind of electronic equipment, and described electronic equipment comprises the dress described in aforementioned any embodiment
Put.
Figure 10 is the structural representation of one embodiment of electronic equipment of the present invention, it is possible to achieve real shown in Fig. 1-7 of the present invention
Executing the flow process of example, as shown in Figure 10, above-mentioned electronic equipment may include that housing 1001, processor 1002, memorizer 1003, electricity
Road plate 1004 and power circuit 1005, wherein, circuit board 1004 is placed in the interior volume that housing 1001 surrounds, processor 1002
It is arranged on circuit board 1004 with memorizer 1003；Power circuit 1005, is used for each circuit for above-mentioned electronic equipment or device
Part is powered；Memorizer 1003 is used for storing executable program code；Processor 1002 is by reading storage in memorizer 1003
Executable program code runs the program corresponding with executable program code, for performing the letter described in aforementioned any embodiment
Breath exchange method.
Processor 1002 to concrete process and the processor 1002 of performing of above-mentioned steps by running executable program generation
The step that code performs further, may refer to the description of Fig. 1-7 illustrated embodiment of the present invention, does not repeats them here.
This electronic equipment exists in a variety of forms, includes but not limited to:
(1) mobile communication equipment: the feature of this kind equipment is to possess mobile communication function, and to provide speech, data
Communication is main target.This Terminal Type includes: smart mobile phone (such as iPhone), multimedia handset, functional mobile phone, and low
End mobile phone etc..
(2) super mobile personal computer equipment: this kind equipment belongs to the category of personal computer, has calculating and processes merit
Can, the most also possess mobile Internet access characteristic.This Terminal Type includes: PDA, MID and UMPC equipment etc., such as iPad.
(3) portable entertainment device: this kind equipment can show and play content of multimedia.This kind equipment includes: audio frequency,
Video player (such as iPod), handheld device, e-book, and intelligent toy and portable car-mounted navigator.
(4) server: providing the equipment of the service of calculating, the composition of server includes that processor, hard disk, internal memory, system are total
Lines etc., server is similar with general computer architecture, but owing to needing to provide highly reliable service, is therefore processing energy
The aspects such as power, stability, reliability, safety, extensibility, manageability require higher.
(5) other have the electronic equipment of data interaction function.
One of ordinary skill in the art will appreciate that all or part of flow process of realizing in above-described embodiment method/, be can
Completing instructing relevant hardware by computer program, software instruction data and instruction (software) are stored in respective
In storage device, this storage device shows as one or more computer-readable or usable storage medium.Storage medium includes not
With the memorizer of form, including such as semiconductor memory apparatus, dynamic or static random accesP memorizer (DRAMs or
SRAMs), erasable formula and programmable type read only memory (EEPROMs), Electrical Erasable formula and programmable read only memory
And flash memory (EEPROMs)；Disk includes Fixed disk, floppy disk and portable hard drive；Other magnetic mediums include tape；And
Optical medium, such as CD (CDs) or video disc (DVDs).It is pointed out that the instruction of above-mentioned software can be by a meter
Calculation machine is readable or computer-usable storage medium provides, or selectively, it is possible to can be used by multiple computer-readables or computer
Storage medium provides, and these storage mediums are distributed in a large scale system with multinode.Computer-readable or computer
Usable storage medium is considered the part of object (or object of product).The object of object or product refers to the list of any production
Individual or the most individual parts.
The above, the only detailed description of the invention of the present invention, but protection scope of the present invention is not limited thereto, and any
Those familiar with the art in the technical scope that the invention discloses, the change that can readily occur in or replacement, all answer
Contain within protection scope of the present invention.Therefore, protection scope of the present invention should be as the criterion with scope of the claims.
Claims (10)
1. an information interacting method, it is characterised in that the method includes:
Obtain the overall interaction content of multiple targeted customer；
Determine the specific interaction content relevant to each targeted customer；
Based on a determination that described specific interaction content judge that the targeted customer relevant to described specific interaction content is the need of letter
Breath prompting；
The most then search the information mated with described specific interaction content, and matching with described specific interaction content
Prompting sight show described information to the targeted customer relevant to described specific interaction content.
Method the most according to claim 1, it is characterised in that the overall interaction content of the multiple targeted customer of described acquisition,
Including:
Obtain the interactive voice content of more than one user, based on described interactive voice content, determine described overall interaction content.
Method the most according to claim 2, it is characterised in that described based on described interactive voice content, determines described whole
Body interaction content, including:
Obtain the context characteristic corresponding with described interactive voice content；
Obtain the context characteristic data base being associated with described context characteristic；
Described interactive voice content is carried out speech recognition, and by the content after speech recognition and described context characteristic data base
Context characteristic pattern mate, obtain described overall interaction content.
Method the most according to claim 3, it is characterised in that described based on described interactive voice content, determines described whole
Body interaction content, also includes:
Extract the vocal print characteristic information in described interactive voice content；
Obtain the user feature database being associated with described vocal print characteristic information；
Described interactive voice content is carried out speech recognition, and by the content after speech recognition and described user feature database
Content mate, obtain described overall interaction content.
Method the most according to claim 2, it is characterised in that described determine relevant to each targeted customer specific alternately
Content, including:
Extract the vocal print characteristic information in described interactive voice content；
The each targeted customer relevant to described interactive voice content is determined according to described vocal print characteristic information；
Described overall interaction content is split with described each targeted customer, obtains relevant to described each targeted customer
Specific interaction content.
Method the most according to claim 5, it is characterised in that described by described overall interaction content and described each target
User splits, and obtains the specific interaction content relevant to described each targeted customer, including:
Based on described vocal print characteristic information, described overall interaction content is split；
Obtain the user feature database corresponding with described vocal print characteristic information；
The content obtained after being split by described overall interaction content is mated with described user feature database, obtains with described
The specific interaction content that each targeted customer is correlated with.
Method the most according to claim 1, it is characterised in that the overall interaction content of the plurality of targeted customer, including:
Other users of active user and more than one are carried out: voice call content, mail interaction content, short message interacting content,
More than one in talk interaction content, instant messaging word interaction content.
Method the most according to claim 1, it is characterised in that described in the prompting matched with described specific interaction content
Sight shows described information to the targeted customer relevant to described specific interaction content, including:
Determine the prompting scene mated with described interaction content；
Judge the matching degree of the prompting scene whether current scene mate with described interaction content；
When described matching degree is more than predetermined threshold value, show described information in default information alert mode to user.
Method the most according to claim 1, it is characterised in that described in the prompting matched with described specific interaction content
Sight, after the targeted customer relevant to described specific interaction content shows described information, also includes:
Receive the feedback information of the targeted customer input relevant to described specific interaction content；And
Determine whether to proceed information alert based on described feedback information.
10. an information interactive device, it is characterised in that this device includes:
Acquisition module, for obtaining the overall interaction content of multiple targeted customer；
Determine module, for determining the specific interaction content relevant to each targeted customer；
Judge module, for based on a determination that described specific interaction content judge relevant to described specific interaction content target use
Family is the need of information alert；
Reminding module, for when the targeted customer relevant to described specific interaction content needs information alert, searches with described
The information of specific interaction content coupling, and at the prompting sight matched with described specific interaction content to specific with described
The targeted customer that interaction content is relevant shows described information.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610538666.1A CN106201424B (en) | 2016-07-08 | 2016-07-08 | A kind of information interacting method, device and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610538666.1A CN106201424B (en) | 2016-07-08 | 2016-07-08 | A kind of information interacting method, device and electronic equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106201424A true CN106201424A (en) | 2016-12-07 |
| CN106201424B CN106201424B (en) | 2019-10-01 |
Family
ID=57473997
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610538666.1A Active CN106201424B (en) | 2016-07-08 | 2016-07-08 | A kind of information interacting method, device and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106201424B (en) |
Cited By (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107146614A (en) * | 2017-04-10 | 2017-09-08 | 北京猎户星空科技有限公司 | A kind of audio signal processing method, device and electronic equipment |
| CN107583291A (en) * | 2017-09-29 | 2018-01-16 | 深圳希格玛和芯微电子有限公司 | Toy interaction method, apparatus and toy |
| CN107977072A (en) * | 2017-07-28 | 2018-05-01 | 北京物灵智能科技有限公司 | What a kind of robot used form method, forms expert system and electronic equipment |
| CN108399700A (en) * | 2018-01-31 | 2018-08-14 | 上海乐愚智能科技有限公司 | Theft preventing method and smart machine |
| CN108537564A (en) * | 2017-03-02 | 2018-09-14 | 九阳股份有限公司 | A kind of dining information method for pushing and home-services robot |
| CN108831456A (en) * | 2018-05-25 | 2018-11-16 | 深圳警翼智能科技股份有限公司 | It is a kind of by speech recognition to the method, apparatus and system of video marker |
| CN109325097A (en) * | 2018-07-13 | 2019-02-12 | 海信集团有限公司 | A kind of voice guide method and device, electronic equipment, storage medium |
| CN109524001A (en) * | 2018-12-28 | 2019-03-26 | 北京金山安全软件有限公司 | Information processing method and device and child wearable device |
| CN109754816A (en) * | 2017-11-01 | 2019-05-14 | 北京搜狗科技发展有限公司 | A kind of method and device of language data process |
| CN109767338A (en) * | 2018-11-30 | 2019-05-17 | 平安科技（深圳）有限公司 | Processing method, device, equipment and the readable storage medium storing program for executing of enterogastritis reimbursement process |
| CN110111793A (en) * | 2018-02-01 | 2019-08-09 | 腾讯科技（深圳）有限公司 | Processing method, device, storage medium and the electronic device of audio-frequency information |
| CN110265013A (en) * | 2019-06-20 | 2019-09-20 | 平安科技（深圳）有限公司 | The recognition methods of voice and device, computer equipment, storage medium |
| CN110824940A (en) * | 2019-11-07 | 2020-02-21 | 深圳市欧瑞博科技有限公司 | Method and device for controlling intelligent household equipment, electronic equipment and storage medium |
| CN110909142A (en) * | 2019-11-20 | 2020-03-24 | 腾讯科技（深圳）有限公司 | Question and sentence processing method and device of question-answer model, electronic equipment and storage medium |
| CN111124121A (en) * | 2019-12-24 | 2020-05-08 | 腾讯科技（深圳）有限公司 | Voice interaction information processing method and device, storage medium and computer equipment |
| CN111813281A (en) * | 2020-05-28 | 2020-10-23 | 维沃移动通信有限公司 | Information acquisition method, information output method, information acquisition device, information output device and electronic equipment |
| CN111914070A (en) * | 2019-05-09 | 2020-11-10 | 上海触乐信息科技有限公司 | Intelligent information prompting assistant system, information prompting method and terminal equipment |
| CN112929502A (en) * | 2021-02-05 | 2021-06-08 | 国家电网有限公司客户服务中心 | Voice recognition method and system based on electric power customer service |
| TWI763207B (en) * | 2020-12-25 | 2022-05-01 | 宏碁股份有限公司 | Method and apparatus for audio signal processing evaluation |
| CN114661899A (en) * | 2022-02-15 | 2022-06-24 | 北京结慧科技有限公司 | Task creating method and device, computer equipment and storage medium |
| CN114924666A (en) * | 2022-05-12 | 2022-08-19 | 上海云绅智能科技有限公司 | Interaction method and device for application scene, terminal equipment and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140248599A1 (en) * | 2005-01-28 | 2014-09-04 | Breakthrough Performance Tech, Llc | Systems and methods for computerized interactive training |
| CN104933028A (en) * | 2015-06-23 | 2015-09-23 | 百度在线网络技术（北京）有限公司 | Information pushing method and information pushing device |
| CN105389304A (en) * | 2015-10-27 | 2016-03-09 | 小米科技有限责任公司 | Event extraction method and apparatus |
-
2016
- 2016-07-08 CN CN201610538666.1A patent/CN106201424B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140248599A1 (en) * | 2005-01-28 | 2014-09-04 | Breakthrough Performance Tech, Llc | Systems and methods for computerized interactive training |
| CN104933028A (en) * | 2015-06-23 | 2015-09-23 | 百度在线网络技术（北京）有限公司 | Information pushing method and information pushing device |
| CN105389304A (en) * | 2015-10-27 | 2016-03-09 | 小米科技有限责任公司 | Event extraction method and apparatus |
Cited By (28)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108537564A (en) * | 2017-03-02 | 2018-09-14 | 九阳股份有限公司 | A kind of dining information method for pushing and home-services robot |
| CN107146614A (en) * | 2017-04-10 | 2017-09-08 | 北京猎户星空科技有限公司 | A kind of audio signal processing method, device and electronic equipment |
| CN107977072A (en) * | 2017-07-28 | 2018-05-01 | 北京物灵智能科技有限公司 | What a kind of robot used form method, forms expert system and electronic equipment |
| CN107977072B (en) * | 2017-07-28 | 2021-06-08 | 北京物灵智能科技有限公司 | Formation method for robot, formation expert system and electronic equipment |
| CN107583291A (en) * | 2017-09-29 | 2018-01-16 | 深圳希格玛和芯微电子有限公司 | Toy interaction method, apparatus and toy |
| CN107583291B (en) * | 2017-09-29 | 2023-05-02 | 深圳希格玛和芯微电子有限公司 | Toy interaction method and device and toy |
| CN109754816B (en) * | 2017-11-01 | 2021-04-16 | 北京搜狗科技发展有限公司 | Voice data processing method and device |
| CN109754816A (en) * | 2017-11-01 | 2019-05-14 | 北京搜狗科技发展有限公司 | A kind of method and device of language data process |
| CN108399700A (en) * | 2018-01-31 | 2018-08-14 | 上海乐愚智能科技有限公司 | Theft preventing method and smart machine |
| CN110111793A (en) * | 2018-02-01 | 2019-08-09 | 腾讯科技（深圳）有限公司 | Processing method, device, storage medium and the electronic device of audio-frequency information |
| CN108831456A (en) * | 2018-05-25 | 2018-11-16 | 深圳警翼智能科技股份有限公司 | It is a kind of by speech recognition to the method, apparatus and system of video marker |
| CN109325097B (en) * | 2018-07-13 | 2022-05-27 | 海信集团有限公司 | Voice guide method and device, electronic equipment and storage medium |
| CN109325097A (en) * | 2018-07-13 | 2019-02-12 | 海信集团有限公司 | A kind of voice guide method and device, electronic equipment, storage medium |
| CN109767338A (en) * | 2018-11-30 | 2019-05-17 | 平安科技（深圳）有限公司 | Processing method, device, equipment and the readable storage medium storing program for executing of enterogastritis reimbursement process |
| CN109524001A (en) * | 2018-12-28 | 2019-03-26 | 北京金山安全软件有限公司 | Information processing method and device and child wearable device |
| CN111914070A (en) * | 2019-05-09 | 2020-11-10 | 上海触乐信息科技有限公司 | Intelligent information prompting assistant system, information prompting method and terminal equipment |
| CN110265013A (en) * | 2019-06-20 | 2019-09-20 | 平安科技（深圳）有限公司 | The recognition methods of voice and device, computer equipment, storage medium |
| WO2020253064A1 (en) * | 2019-06-20 | 2020-12-24 | 平安科技（深圳）有限公司 | Speech recognition method and apparatus, and computer device and storage medium |
| CN110824940A (en) * | 2019-11-07 | 2020-02-21 | 深圳市欧瑞博科技有限公司 | Method and device for controlling intelligent household equipment, electronic equipment and storage medium |
| CN110909142A (en) * | 2019-11-20 | 2020-03-24 | 腾讯科技（深圳）有限公司 | Question and sentence processing method and device of question-answer model, electronic equipment and storage medium |
| CN110909142B (en) * | 2019-11-20 | 2023-03-31 | 腾讯科技（深圳）有限公司 | Question and sentence processing method and device of question-answer model, electronic equipment and storage medium |
| CN111124121A (en) * | 2019-12-24 | 2020-05-08 | 腾讯科技（深圳）有限公司 | Voice interaction information processing method and device, storage medium and computer equipment |
| CN111813281A (en) * | 2020-05-28 | 2020-10-23 | 维沃移动通信有限公司 | Information acquisition method, information output method, information acquisition device, information output device and electronic equipment |
| TWI763207B (en) * | 2020-12-25 | 2022-05-01 | 宏碁股份有限公司 | Method and apparatus for audio signal processing evaluation |
| US11636844B2 (en) | 2020-12-25 | 2023-04-25 | Acer Incorporated | Method and apparatus for audio signal processing evaluation |
| CN112929502A (en) * | 2021-02-05 | 2021-06-08 | 国家电网有限公司客户服务中心 | Voice recognition method and system based on electric power customer service |
| CN114661899A (en) * | 2022-02-15 | 2022-06-24 | 北京结慧科技有限公司 | Task creating method and device, computer equipment and storage medium |
| CN114924666A (en) * | 2022-05-12 | 2022-08-19 | 上海云绅智能科技有限公司 | Interaction method and device for application scene, terminal equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106201424B (en) | 2019-10-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106201424B (en) | A kind of information interacting method, device and electronic equipment | |
| JP7101322B2 (en) | Voice trigger for digital assistant | |
| US10176810B2 (en) | Using voice information to influence importance of search result categories | |
| US9905228B2 (en) | System and method of performing automatic speech recognition using local private data | |
| US9742912B2 (en) | Method and apparatus for predicting intent in IVR using natural language queries | |
| US9502032B2 (en) | Dynamically biasing language models | |
| US9502031B2 (en) | Method for supporting dynamic grammars in WFST-based ASR | |
| KR20180070684A (en) | Parameter collection and automatic dialog generation in dialog systems | |
| US9786281B1 (en) | Household agent learning | |
| KR20200007011A (en) | Intercom style communication using multiple computing devices | |
| CN107004410A (en) | Voice and connecting platform | |
| US9502029B1 (en) | Context-aware speech processing | |
| CN105320726A (en) | Reducing the need for manual start/end-pointing and trigger phrases | |
| KR101891492B1 (en) | Method and computer device for providing contextual natural language conversation by modifying plain response, and computer readable recording medium | |
| CN105009203A (en) | Methods and apparatus for detecting a voice command | |
| CN103078995A (en) | Customizable individualized response method and system used in mobile terminal | |
| CN103035240A (en) | Speech recognition repair using contextual information | |
| Neustein | Advances in speech recognition: mobile environments, call centers and clinics | |
| CN106341539A (en) | Automatic evidence obtaining method of malicious caller voiceprint, apparatus and mobile terminal thereof | |
| KR102120751B1 (en) | Method and computer readable recording medium for providing answers based on hybrid hierarchical conversation flow model with conversation management model using machine learning | |
| CN112567718A (en) | Electronic device for performing tasks including calls in response to user speech and method of operation | |
| CN112513978A (en) | Hot word identification and passive assistance | |
| US11381675B2 (en) | Command based interactive system and a method thereof | |
| US20220161131A1 (en) | Systems and devices for controlling network applications | |
| US11798538B1 (en) | Answer prediction in a speech processing system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |