CN109213877A - A kind of financial services system based on intelligent sound human-computer interaction equipment - Google Patents
A kind of financial services system based on intelligent sound human-computer interaction equipment Download PDFInfo
- Publication number
- CN109213877A CN109213877A CN201710513344.6A CN201710513344A CN109213877A CN 109213877 A CN109213877 A CN 109213877A CN 201710513344 A CN201710513344 A CN 201710513344A CN 109213877 A CN109213877 A CN 109213877A
- Authority
- CN
- China
- Prior art keywords
- financial
- unit
- voice
- output
- word string
- 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.)
- Pending
Links
- 230000003993 interaction Effects 0.000 title claims abstract description 17
- 238000013500 data storage Methods 0.000 claims abstract description 16
- 230000008520 organization Effects 0.000 claims description 48
- 238000012545 processing Methods 0.000 claims description 21
- 238000000034 method Methods 0.000 claims description 18
- 238000003058 natural language processing Methods 0.000 claims description 17
- 238000005516 engineering process Methods 0.000 claims description 12
- 238000002887 multiple sequence alignment Methods 0.000 claims description 10
- PCHJSUWPFVWCPO-UHFFFAOYSA-N gold Chemical compound [Au] PCHJSUWPFVWCPO-UHFFFAOYSA-N 0.000 claims description 9
- 239000010931 gold Substances 0.000 claims description 9
- 229910052737 gold Inorganic materials 0.000 claims description 9
- 238000007726 management method Methods 0.000 claims description 8
- 238000011160 research Methods 0.000 claims description 8
- 239000000463 material Substances 0.000 claims description 7
- 230000000694 effects Effects 0.000 claims description 4
- 230000001755 vocal effect Effects 0.000 claims description 3
- 238000012546 transfer Methods 0.000 claims description 2
- 230000005540 biological transmission Effects 0.000 claims 1
- 230000004044 response Effects 0.000 abstract description 6
- 230000008569 process Effects 0.000 description 9
- 238000013473 artificial intelligence Methods 0.000 description 6
- 238000006243 chemical reaction Methods 0.000 description 6
- 238000012790 confirmation Methods 0.000 description 5
- 241000209094 Oryza Species 0.000 description 4
- 235000007164 Oryza sativa Nutrition 0.000 description 4
- 235000009566 rice Nutrition 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 241000252794 Sphinx Species 0.000 description 1
- 241000209140 Triticum Species 0.000 description 1
- 235000021307 Triticum Nutrition 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 238000005138 cryopreservation Methods 0.000 description 1
- 238000006073 displacement reaction Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000002844 melting Methods 0.000 description 1
- 230000008018 melting Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005096 rolling process Methods 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L13/00—Speech synthesis; Text to speech systems
- G10L13/02—Methods for producing synthetic speech; Speech synthesisers
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/26—Speech to text systems
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L17/00—Speaker identification or verification techniques
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Business, Economics & Management (AREA)
- Computational Linguistics (AREA)
- Economics (AREA)
- Development Economics (AREA)
- Finance (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- Technology Law (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Accounting & Taxation (AREA)
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
Abstract
The present invention is a kind of financial services system based on intelligent sound human-computer interaction equipment, includes: intelligent apparatus, including voice output/input equipment, inputs voice, and output voice to provide user；Cloud network to couple in a wired or wireless manner with the intelligent apparatus, and transmits the data such as input voice, output voice；Server, it is coupled to the cloud network, to receive the input voice of the intelligent apparatus, and the input order of the financial service to financial service corresponding to the financial services system is converted to, and the financial service input order that the financial services system is completed is converted into the output voice to intelligent apparatus response；Data storage device with storage such as complex finance data record and Account Profile of swiping the card, and is coupled to the cloud network.
Description
Technical field
The present invention relates to a kind of to input through voice to carry out the system and method for Financial Management, particularly relates to through voice
Input equipment carries out voice input, to carry out human-computer interaction with intelligent apparatus, to manage the System and method for of required financial service.
Background technique
Artificial intelligence (Artificial Intelligence, AI), refers to that the system manually manufactured is showed
Intelligence.Usual artificial intelligence refers to by common computer or supercomputer realization, for certain situation have logic or
The intelligence of step judgement.The research of artificial intelligence can be divided into several technical problems, and branch field is concentrated mainly on solution tool
Body problem, one of them is, how using a variety of different algorithm problem analyses to complete specific objective.
Such as Alpha's go (AlphaGo) is the artificial intelligence go developed by London Google DeepMind
Program.In October, 2015, it becomes first without allowing son, the computer of go occupation chess scholar can be defeated on No. 19 chessboards
Go program.In March, 2016, in a five-set game match, AlphaGo defeats top duty in first three office and final set
Industry chess scholar Li Shi Tol becomes first and does not defeat the computer maintenance and servine program of the professional nine sections of chess scholars of go by son is allowed.This card
Artificial intelligence is illustrated in the potentiality of logic judgment and automated job, so that application will above have a possibility that unlimited in future, seems
Lockheed Corp. in the Alis carried in F-35A/B/C opportunity of combat, the intelligent personal assistant Cortana of Microsoft,
Or the built-in Alexa etc. of Echo wisdom speaker that Amazon Company produces.
Speech recognition (Speech recognition) technology, also referred to as automatic speech recognition (Automatic
Speech Recognition, ASR) or speech-to-text (Speech To Text, STT), target be with computer from
It is dynamic that the voice content of the mankind is converted into corresponding text.And in contrast, be text-to-speech (Text to Speech,
TTS technology), the target of the technology are then that the input of any text is changed into voice output, and user does not need for specific character
Content pre-recorded speech, as long as word content is inputted computer, computer can be automatically processed including date, time, phone, gold
The correct pronunciation of the inferior number of volume, item.The application of speech recognition technology includes phonetic dialing, the document retrieval of voice text, simple
Dictation data input etc., and the application of voice output is then inconvenient for the object of dysphotia or eyes, in such as work or drives
In driving etc..
The technology of speech recognition, most great breakthrough are the applications of hidden markov model (Hidden Markov Model),
Related mathematical reasoning is proposed from Baum, and by the research of Rabiner et al., the Li Kaifu of Carnegie Mellon University is finally realized
First large vocabulary speech recognition system Sphinx based on hidden markov model.Hereafter strictly speaking speech recognition technology is simultaneously
It is not detached from the frame of hidden markov model.It is illustrated with United States Patent (USP) US8315871B2, which proposes with aforementioned hidden
Markov model carries out a kind of rolling (Rope-Jumping) algorithm of text-to-speech, the mode which carries out
For unordered parameters,acoustic is modified to generate stable sound ray frequency spectrum, to be produced from through the restrictive condition set in advance
Right voice.
But the main means to be linked up through voice as people and computer or intelligent apparatus, it realizes such as science fiction
Piece seems starry sky adventure note, the mankind naturally with the audio command computer of itself or intelligent apparatus after, computer it is automatic and
Accurately with speech answering or some function is carried out, rather than be only merely that the voice of the mankind is changed into text to note down, then
After needing that aforementioned voice is further turned Text region, with the voice command of the various algorithm analysis mankind, and being converted to can
For the order of computer interpretation, begin that the follow-up process of the order can be continued to execute.
Natural language processing algorithm is that one of the technology of discrimination power is promoted as Such analysis speech-to-text at present, should
Algorithm combines Da Mailiu-Levenstein distance (Damerau-Levenshtein distance), Dynamic Programming
The algorithm of (Dynamic programming), Multiple Sequence Alignment (Multiple sequence alignment) etc..Its
In illustrated with Da Mailiu-Levenstein distance, the main purpose of the algorithm, to be converted to a lot of command statement
It simplifies, and can be corresponded to the command statement for being established in data bank in advance the most, to execute required movement.
Its practice is the distance of any two texts a, b in a long string command to be calculated, in a finite element sequence
It calculates needed for an order most briefly, can most give expression to the sentence of user's meaning, and during the calculating, it contains following
Four further act according to distance calculated, are respectively inserted into, delete, one synonymous word of displacement, with two texts
Word location exchange (such as: I and you, you and I；Paper and pen, pen and paper.Above-mentioned two groups of examples do not influence the sequence exchange of word
Context).In addition to this, which can also accomplish correcting more than 80% mosaic mistake.Through combine aforementioned voice identification and
Natural language processing finally can be realized the purpose that human-computer interaction is carried out with voice.
However, although having occurred many digital assistant products for carrying out human-computer interaction with voice on the market, still still meeting
Have very high probability by command statement be misinterpreted as another it is imperatival happen, be with voice with digital assistants if wanting
The main means of human-computer interaction, to be operated to current financial products, such as stock, warrant, futures, right to choose, error
Risk it is still too big, especially legal issue level involved in financial products involves very wide or even its money number easily
Tens of or even millions of take-offs can not allow have any mistake during operation really, also therefore each big bank or gold now
Melt the voice operating service that mechanism is released, is also only in the phone or before operating terminal machine, the voice pre-recorded with one section mentions
Showing user, how this operates.Such as user squeezes into bank with phone, the voice service used, the language which will hear in phone
Sound are as follows: " stock transaction please by 1, warrant places an order please by 2, futures place an order please by 3 ... attendant service please by 9 " content.
Although above-mentioned can be such that financial service automates to a certain degree really with pre-recorded speech prompt operation, such side
The shortcomings that formula, is nonelastic using upper very standard for it, also can not be with the side of the standard if user has a question in operation
Formula obtains the more information of the operation, therefore, there is still a need for a kind of correct interpretation that can have height for voice command, and
And the user demand of voice response automatic discrimination can be crossed when necessary to provide the digital assistants of necessary information, and be intended to reach this mesh
Mark, then need a kind of system, can just current digital assistants technology be improved.
Summary of the invention
The shortcomings that in view of aforementioned known techniques, the present invention propose a kind of finance based on intelligent sound human-computer interaction equipment
Service system is optimized the identification result of speech recognition technology, to reach naturally with voice and user session,
And the order to user can be reached from dialogue or be required to accurately understand, reach better effect, and user can be provided
The target of relevent information.
In order to achieve the purpose that take off, the first embodiment proposed according to the present invention, it includes: intelligent apparatus, including
One voice output/input equipment, to provide, user inputs voice and output financial services system one exports voice；Cloud
Network to couple in a wired or wireless manner with aforementioned intelligent device, and transmits the moneys such as above-mentioned input voice, output voice
Material；Server is coupled to the cloud network, to receive intelligent apparatus input voice, and is converted to the financial services system institute
The financial service order of corresponding financial service, and be converted to what the financial services system was completed pair to financial service order
At least one output voice of intelligent apparatus response；Data storage device, with store Account Profile, record of swiping the card, profit and loss on investment,
The finance data such as market trend, research report and quofed rice of commodities, and it is coupled to above-mentioned cloud network, by above-mentioned data according to demand
It is transferred to server.
The second embodiment proposed according to the present invention, server further include language and characters processing module, wherein the processing
Module includes user interface administrative unit, to manage the knowledge of operation workflow, user's vocal print in above-mentioned language and characters processing module
Not, with input voice and output voice connecting interface；Speech-to-text unit is coupled to user interface administrative unit, will
Input voice is converted to input text word string；Text-to-speech unit is coupled to user interface administrative unit, by financial service system
It unites after completed financial service order, generated output character word string is converted to output voice；And sound/character turns
Corpus library unit is changed, with store voice data, and written historical materials corresponding to the voice data, and it is coupled to speech-to-text list
Member and text-to-speech unit.
The second embodiment proposed according to the present invention, server further include user session administrative unit, to manage finance
The operation workflow of server in service system, is coupled to user interface administrative unit, with input/output " input text word string "
Or " output character word string ".
The second embodiment proposed according to the present invention further includes financial relative conversations meaning of one's words processing module in server,
The module contains Financial organization corpus library unit, to store the vocabulary and term of various Financial organizations, and is coupled to user session
Administrative unit.
One aspect of second embodiment according to the present invention, financial relative conversations meaning of one's words processing module further include natural language
Processing unit is coupled to user session administrative unit and Financial organization corpus library unit, when the user session administrative unit calculates
When input text word length is greater than a default value out, then with the natural language processing unit according in conjunction with Da Mailiu-Lai Wensi
The various algorithms such as smooth distance, Dynamic Programming and Multiple Sequence Alignment compare various gold stored in Financial organization corpus library unit
Melt specialized vocabulary and term, input word string is converted into financial service input order.
Another aspect according to a second embodiment of the present invention, it is single that Financial organization corpus library unit is coupled to user session management
Member, when the user session administrative unit calculates input text word length less than a default value, then with Financial organization corpus
Input word string is converted to financial service input order by stored various Financial organization vocabulary and term in library unit.
The financial relative conversations meaning of one's words processing module of the second embodiment proposed according to the present invention further includes financial Information
Service unit is coupled to user session administrative unit and data storage device, is ordered with being inputted according to financial service above-mentioned, knot
Financial data stored in data storage device is closed, with financial service needed for carrying out user.
Above-mentioned financial service may include financial commodity transaction, Account Status inquiry, transaction records are inquired, Market News are looked into
Inquiry or financial product quotation screens etc., the above is only citings for applicable financial service, however it is not limited to which above-mentioned financial service is answered
With.
An aspect according to the present invention, financial Information service unit generate finance clothes after completing above-mentioned financial service
It is engaged in output report, and is transmitted to user session administrative unit, when to calculate financial service defeated for the user session administrative unit
When reporting that contained report text word length is greater than a default value out, then reached with aforementioned natural language processing unit according to combination
The various algorithms such as Mai Liu-Levenstein distance, Dynamic Programming and Multiple Sequence Alignment, compare institute in Financial organization corpus library unit
Report text word string is converted to output character word string, is transferred to user session by the various Financial organization vocabulary and term of storage
Administrative unit, user session administrative unit by above-mentioned output character word string, are transferred to user interface administrative unit again.
Another aspect according to the present invention, financial Information service unit generate finance after completing above-mentioned financial service
Output report is serviced, and is transmitted to user session administrative unit, when the user session administrative unit calculates financial service
When report text word length is less than a default value contained by output report, then with stored in Financial organization corpus library unit
Report text word string is converted to output character word string, it is single to be transferred to user session management by various Financial organization vocabulary and term
Member, user session administrative unit by above-mentioned output character word string, are transferred to user interface administrative unit again.
The above system is to illustrate the purpose of the present invention, technological means and its attainable effect, in related fields
Be familiar with this technology people can via following embodiment demonstration with schema explanation and claim become apparent from it is bright
The present invention.
Detailed description of the invention
Fig. 1 is to show system architecture of the invention.
Fig. 2 is to show system thin portion framework of the invention.
Fig. 3 is to show application situation process of the invention: general finance question and answer (not being related to account profile).
Fig. 4 is to show application situation process of the invention: personalized service (is related to account profile).
Symbol description
100: system architecture
101A: user A 101B: user B 102A: intelligent apparatus A
102B: intelligent apparatus B 103: cloud network 104: server
105: data storage device
200: system thin portion framework
201: user 202: intelligent apparatus 203: voice output/input equipment
204: server 205: language and characters processing module
206: sound/character conversion and material library unit 207: speech-to-text unit
208: text-to-speech unit 209: user interface administrative unit
210: financial relative conversations meaning of one's words processing module 211: Financial organization corpus library unit
212: natural language processing unit 213: user session administrative unit
214: financial Information service unit 215: data storage device
300: applying situation process: general finance question and answer (not being related to account profile)
301: user 302: intelligent apparatus 303: server
S1-S12: situation process step is applied
400: apply situation process: personalized service (is related to account profile)
401: user 402: intelligent apparatus 403: server
S13-S38: situation process step is applied
Specific embodiment
Illustrate embodiments of the present invention by particular specific embodiment below, the personage for being familiar with this technology can be by this
The revealed content of specification understands effect and its advantage of the invention easily.And the present invention also can be specific real by other
It applies example to be used and implemented, every details that this specification is illustrated can also be applied based on different demands, and not departed from
A variety of different modifications or change are carried out under spirit of the invention.
The present invention will be described with preferred embodiment and viewpoint, and structure of the invention is explained by such narration system, only to
Illustrate rather than to limit claim of the invention.Therefore, in addition to the preferred embodiment in specification, the present invention is also
It can be rendered in other embodiments extensively.
First embodiment according to the present invention, system architecture of the invention (100), as shown in the first figure, it includes: at least
One intelligent apparatus (102A, 102B), to provide an at least user (101A, 101B) input at least one input voice, and output gold
Melt service system at least one and exports voice；Cloud network (103), with aforementioned intelligent device (102A, 102B) with wired or nothing
The mode of line couples, and transmits the data such as above-mentioned input voice, output voice；Server (104) is coupled to above-mentioned cloud net
Network (103), to handle the input voice of intelligent apparatus (102A, 102B) and be converted to corresponding to financial services system (100)
The financial service order of financial service, and the financial service input order that financial services system is completed is converted to intelligent dress
Set the output voice of (102A, 102B) response；Data storage device (105), to store account profile, record of swiping the card, investment damage
The finance data such as benefit, market trend, research report and quofed rice of commodities, and above-mentioned cloud network (103) are coupled to, by above-mentioned data
Server (104) are transferred to according to demand.
The second embodiment proposed according to the present invention, as shown in the second figure: server (204) further includes at language and characters
It manages module (205), wherein the processing module includes user interface administrative unit (209), to manage above-mentioned language and characters processing mould
The identification of operation workflow, user's vocal print in block (205), the connecting interface with input voice and output voice；Speech-to-text
Unit (207) is coupled to user interface administrative unit (209), to convert input speech into input text word string；Text turns language
Sound unit (208) is coupled to user interface administrative unit (209), after the financial service order that financial services system is completed,
Generated output character word string, is converted to output voice；And sound/character conversion corpus library unit (206), to store language
Sound data, and written historical materials corresponding to the voice data, and it is coupled to speech-to-text unit (207) and text-to-speech list
First (208).
The second embodiment proposed according to the present invention, as shown in the second figure: server (204) further includes user session pipe
It manages unit (213) and is coupled to user interface administrative unit to manage the operation workflow of server in financial services system (204)
(209), with input/output " input text word string " or " output character word string ".
The second embodiment proposed according to the present invention, as shown in the second figure: financial correlation is further included in server (204)
Talk with meaning of one's words processing module (210), which contains Financial organization corpus library unit (211), to store various Financial organizations
Vocabulary and term, and it is coupled to user session administrative unit (213).
One aspect of second embodiment according to the present invention, as shown in the second figure: financial relative conversations meaning of one's words processing module
(210) natural language processing unit (212) are further included, user session administrative unit (213) and Financial organization corpus are coupled to
Unit (211), with processing input text word string, with generate output character word string, when the user session administrative unit (213) calculate
When input text word length is greater than a default value out, then with the natural language processing unit (212) according in conjunction with Da Mailiu-Lay
The various algorithms such as Wen Sitan distance, Dynamic Programming and Multiple Sequence Alignment, compare and are stored up in Financial organization corpus library unit (211)
Input word string is converted to financial service input order by the various Financial organization vocabulary and term deposited.
Another aspect according to a second embodiment of the present invention, as shown in the second figure: Financial organization corpus library unit (211) coupling
User session administrative unit (213) are connected to, are less than when the user session administrative unit (213) calculates input text word length
When one default value, then with various Financial organization vocabulary and term stored in Financial organization corpus library unit (211), it will input
Word string is converted to financial service input order.
The second embodiment proposed according to the present invention, financial relative conversations meaning of one's words processing module as shown in the second figure
(210) financial Information service unit (214) are further included, user session administrative unit (213) and data storage device are coupled to
(215), it is ordered with being inputted according to financial service above-mentioned, the finance data in conjunction with stored by data storage device (215), with
Financial service needed for carrying out user may include financial commodity transaction, Account Status inquiry, transaction records inquiry, market money
News inquiry or financial product quotation screens etc., the above is only citings for applicable financial service, however it is not limited to above-mentioned financial service
Using.
An aspect according to a second embodiment of the present invention, as shown in the second figure: financial Information service unit (214) is in completion
After above-mentioned financial service, that is, financial service output report is generated, and be transmitted to user session administrative unit (213), when this
User session administrative unit (213) calculates report text word length contained by financial service output report greater than a default value
When, then with aforementioned natural language processing unit (212) according to combination Da Mailiu-Levenstein distance, Dynamic Programming and multisequencing
Algorithms more various than equity, compare various Financial organization vocabulary and term stored in Financial organization corpus library unit (211),
Report text word string is converted into output character word string, is transferred to user session administrative unit (213), user session administrative unit
(213) above-mentioned output character word string is transferred to user interface administrative unit (213) again.
Another aspect according to a second embodiment of the present invention, as shown in the second figure: financial Information service unit (214) is in complete
After above-mentioned financial service, that is, financial service output report is generated, and be transmitted to user session administrative unit (213), when
The user session administrative unit (213) calculates report text word length contained by financial service output report less than a default
When value, then with various Financial organization vocabulary and term stored in Financial organization corpus library unit (211), text word will be reported
String is converted to output character word string, is transferred to user session administrative unit (213), user session administrative unit (213) again will be upper
Output character word string is stated, user interface administrative unit (213) are transferred to.
According to a second embodiment of the present invention, as shown in the second figure: user interface administrative unit (209) receives output character
After word string, i.e., with speech-to-text unit (207), is converted according to sound/character and expect output stored in library unit (206)
Output character word string is converted to output voice by voice data corresponding to text word string, will be exported through cloud network (103)
Voice transfer plays to user (201) then by voice output/input equipment (203) to intelligent apparatus (202).
According to a third embodiment of the present invention, as shown in the third figure, financial services system proposed by the invention is disclosed
One of application aspect: " applying situation process: general finance question and answer (not being related to account profile) " (300).The application
The situation that situation is applicable in is that user (301) is in order to inquire certain financial Information, such as market trend, research report or commodity report
Valence, but it is not related to account profile itself.
The application situation process of 3rd embodiment is that user (301) penetrate intelligent apparatus (302) in S1 step with voice
Mode input voice, after which is received in S2 step by voice output/input equipment (203),
It is transferred to the speech recognition in user interface administrative unit (209) progress S3 step through cloud network (103), wherein voice
Turn text unit (207) and sound/character conversion corpus library unit (206), be coupled to user interface administrative unit (209),
Using hidden markov model or other algorithms, to convert input speech into input text word string in S4 step.
Input text be then transferred in user session administrative unit (213), with carry out in S5 step text meaning interpretation/
The step of confirmation dialogue is intended to.It carries out the judgement of text meaning in the S5 the step of to be intended to confirmation dialogue, the method root specifically carried out
According to an aspect of the invention, to be greater than a default when the user session administrative unit (213) calculates input text word length
When value, then with natural language processing unit (212) according to combination Da Mailiu-Levenstein distance, Dynamic Programming and multisequencing ratio
The various algorithms of equity, compare various Financial organization vocabulary stored in Financial organization corpus library unit (211) in S6 step
It, will be defeated with the data such as market trend, research report or quofed rice of commodities stored in term and data storage device (215)
Enter word string and is converted to financial service input order.
The step of text meaning interpretation/confirmation dialogue in S5 step is intended to, the method specifically carried out is according to the present invention another
One aspect, when the user session administrative unit (213) calculates input text word length less than a default value, then with finance
Institute in stored various Financial organization vocabulary and term and data storage device (215) in professional corpus library unit (211)
Input word string is converted to financial service input order, then by the data such as market trend, research report or the quofed rice of commodities of storage
Financial service input order is transferred to financial Information service unit (214), to carry out step S7, financial Information service unit
(214) according to financial service input order to carry out corresponding financial service.
The content that financial Information service unit (214) is ordered according to financial service input in step S8, in the above-mentioned gold of completion
Corresponding financial service output report is generated after melting service, in step S9, an aspect according to the present invention is transmitted to
User session administrative unit (213), when the user session administrative unit (213) calculates report contained by financial service output report
When accusing text word length greater than a default value, then with aforementioned natural language processing unit (212) according in conjunction with Da Mailiu-Lai Wen
The various algorithms such as Si Tan distance, Dynamic Programming and Multiple Sequence Alignment compare stored in Financial organization corpus library unit (211)
Various Financial organization vocabulary and term, will report text word string be converted to output character word string, be transferred to user session management
Unit (213), user session administrative unit (213) by above-mentioned output character word string, are transferred to user interface administrative unit again
(213)。
In step S9, another aspect according to the present invention, financial Information service unit (214) is in the above-mentioned finance clothes of completion
After business, that is, financial service output report is generated, and be transmitted to user session administrative unit (213), when the user session pipe
When reason unit (213) calculates that report text word length is less than a default value contained by financial service output report, then with gold
Melt various Financial organization vocabulary and term stored in professional corpus library unit (211), report text word string is converted to defeated
Text word string out is transferred to user session administrative unit (213), and user session administrative unit (213) is again by above-mentioned output character
Word string is transferred to user interface administrative unit (213), carries out step S10.
In step S10, after user interface administrative unit (209) receives output character word string, i.e., with speech-to-text
Unit (207) is converted according to sound/character and expects voice corresponding to output character word string stored in library unit (206)
Output character word string is converted to output voice by data, is transmitted to intelligent apparatus (202) through cloud network (103), then by
By voice output/input equipment (203), user (201) are played to.
According to a fourth embodiment of the present invention, as shown in the 4th figure, financial services system proposed by the invention is disclosed
One of application aspect: " applying situation process: individualized (being related to account profile) " (400).This is fitted using situation
With the case where be that user (401) is in order to carry out certain financial transaction or financial operation, Account Profile inquiry, record inquiry of swiping the card,
Financial products, which place an order etc., is related to the operation behavior of account profile itself.
The application situation process of fourth embodiment is that user (401) penetrate intelligent apparatus (302) in S13 step with voice
Mode input one first input voice, this first input voice in S14 step by voice output/input equipment (203) institute
After receiving, user interface administrative unit (209) are transferred to through cloud network (103) and carry out user (401) in S15 step
Application on Voiceprint Recognition, after Application on Voiceprint Recognition passes through, in S16 step compare user (401) personal information, and in S17 in
User interface administrative unit (213) leads to the problem of the core body problem word string of preset at least user's core body, and (such as: your birthday is
When, also can elasticity setting compared with secret the problem of, to improve safety.), user interface administrative unit (209) are transmitted to, then
Step S18 is carried out, after user interface administrative unit (209) receives core body problem word string, i.e., with speech-to-text unit
(207), it converts according to sound/character and expects voice data corresponding to output character word string stored in library unit (206),
Core body problem word string is converted into the first output voice, is transmitted to intelligent apparatus (202) through cloud network (103), then by
Step S19 plays to user (401) by voice output/input equipment (203), by the first output voice.
User (401) listens to the first output voice in step S20 at this time, and according to the core body problem of the first output voice
Content is answered, and the content of answer inputs the second input voice as shown in step S21, which walks in S22
After being received in rapid by voice output/input equipment (203), user interface management is transferred to through cloud network (103)
Unit (209) carries out the speech recognition in S23 step, and wherein speech-to-text unit (207) and sound/character convert corpus
Library unit (206) is coupled to user interface administrative unit (209), using hidden markov model or other algorithm, in
The second input voice is converted into the first input text word string in S24 step, and whether just to confirm core body data in step S25
Really, if the confirmation of core body data is errorless, step S26, clothes needed for financial Information service unit (214) generates at least one at this time are carried out
One inquiry word string of business.
Wherein in step S26, the inquiry of the required service can take for a pre-set problem or financial Information
Business unit is customized according to the personal account data of user (401), the data such as record of swiping the card in data storage device (215)
Problem, the inquiry word string are transferred to user session administrative unit (213).
An aspect according to the present invention, when the user session administrative unit (213) calculates the inquiry word length greater than one
When default value, then with aforementioned natural language processing unit (212) according to combine Da Mailiu-Levenstein distance, Dynamic Programming and
The various algorithms such as Multiple Sequence Alignment compare various Financial organization vocabulary stored in Financial organization corpus library unit (211)
With term, inquiry word string is converted into inquiry text output word string, is transmitted to user interface administrative unit (209).
Another aspect according to the present invention is less than when the user session administrative unit (213) calculates inquiry word length
When one default value, then with various Financial organization vocabulary and term stored in Financial organization corpus library unit (211), it will inquire
Word string is converted to inquiry text output word string, is transmitted to user interface administrative unit (209).
In step S27, after user interface administrative unit (209) receives inquiry text output word string, i.e., turned with voice
Text unit (207), it is right according to sound/character conversion and inquiry text output word string institute stored in material library unit (206)
Inquiry text output word string is converted to the second output voice, is transmitted to intelligence through cloud network (103) by the voice data answered
Second output voice is played to user then by step S27 by voice output/input equipment (203) by device (202)
(401)。
When user (401) is after step S28 hears the second output voice, in step S29, i.e., with its institute of spoken responses
The service to be carried out, to input third input voice.The third inputs voice in S30 step by voice output/input equipment
(203) after receiving, user interface administrative unit (209) is transferred to through cloud network (103) and are carried out in S31 step
Speech recognition, wherein speech-to-text unit (207) and sound/character conversion corpus library unit (206), is coupled to user Jie
Face administrative unit (209), using hidden markov model or other algorithms, to be converted input speech into S32 step
Second input text word string.
Second input text word string is then transferred in user session administrative unit (213), to carry out in S33 step
The step of text meaning interpretation/confirmation dialogue is intended to, an aspect according to the present invention, to be counted when the user session administrative unit (213)
When calculating the second input text word length greater than a default value, then wheat is reached according to combination with natural language processing unit (212)
- Levenstein distance, the various algorithms such as Dynamic Programming and Multiple Sequence Alignment are stayed, is compared Financial organization corpus library unit (211)
In stored various Financial organization vocabulary and term, the second input text word string is converted into financial service input order, is connect
By financial service input order be transferred to financial Information service unit (214).
Another aspect according to the present invention, when the user session administrative unit (213) calculates the second input text word string
When length is less than a default value, then with various Financial organization vocabulary and art stored in Financial organization corpus library unit (211)
Second input text word string is converted to financial service input order, financial service input order is then transferred to finance by language
Information service unit (214).
Then, step S34 is executed, financial Information service unit (214) is inputted according to financial service and ordered, connection data storage
In cryopreservation device, the Account Profile of user (401) carries out corresponding personalized financial service in step S35.
Then, after completing above-mentioned financial service, in step S36, corresponding financial service output report is generated, it will
It is transferred to user session administrative unit (213), an aspect according to the present invention, when the user session administrative unit (213) is counted
When calculating report text word length contained by financial service output report greater than a default value, then with aforementioned natural language processing
Unit (212) compares gold according to the various algorithms such as Da Mailiu-Levenstein distance, Dynamic Programming and Multiple Sequence Alignment are combined
Melt in professional corpus library unit (211) stored various Financial organization vocabulary and term, report text word string is converted to the
Three output character word strings are transferred to user session administrative unit (213), and user session administrative unit (213) is again by above-mentioned third
Output character word string is transferred to user interface administrative unit (213).
In step S36, another aspect according to the present invention, financial Information service unit (214) is in the above-mentioned finance of completion
After service, that is, financial service output report is generated, and be transmitted to user session administrative unit (213), when the user session
When administrative unit (213) calculates that report text word length is less than a default value contained by financial service output report, then with
Stored various Financial organization vocabulary and term in Financial organization corpus library unit (211) are converted to report text word string
Third output character word string is transferred to user session administrative unit (213), and user session administrative unit (213) is again by above-mentioned
Three output character word strings are transferred to user interface administrative unit (213), to carry out the voice response of step S37.
In step S37, after user interface administrative unit (209) receives third output character word string, i.e., turned with voice
Text unit (207), it is right according to third output character word string institute stored in sound/character conversion corpus library unit (206)
Third output character word string is converted to third output voice, is transmitted to intelligence through cloud network (103) by the voice data answered
Device (202) executes step S38, plays to user (401) then by voice output/input equipment (203).
Claims (9)
1. a kind of financial services system based on intelligent sound human-computer interaction equipment, characterized by comprising:
One server is coupled to a cloud network, and wherein the cloud network can be coupled with an external intelligent apparatus, wherein the intelligence
Device includes one voice output/input equipment, inputs voice to provide user, and export voice to the service
Device；Wherein the server can receive the input voice of the intelligent apparatus, and be converted to one corresponding to the financial services system
One financial service of financial service inputs order, and the financial service input order that can be completed the financial services system turns
It is changed to the output voice responded to the intelligent apparatus；And
One data storage device, to store Account Profile, record of swiping the card, profit and loss on investment, market trend, research report and commodity report
The complex finance data of valence at least one of them, and it is coupled to the cloud network, above-mentioned data is transferred to this according to demand
Server.
2. the financial services system based on intelligent sound human-computer interaction equipment as described in claims 1, which is characterized in that described
Server also includes:
One language and characters processing module, wherein the language and characters processing module further include:
One user interface administrative unit, with manage the identification of the operation workflow in the language and characters processing module, user's vocal print with
The connecting interface of the input voice and the output voice；
One speech-to-text unit, is coupled to the user interface administrative unit, which is converted to an input text
Word string；
One text-to-speech unit couples the user interface administrative unit, and it is defeated to complete the financial service in the financial services system
After entering order, generated output character word string is converted to an output voice；And
One sound/character converts corpus library unit, corresponding multiple with the complex voice data to store complex voice data
Several written historical materials, and couple the speech-to-text unit and the text and turn voice unit.
3. the financial services system based on intelligent sound human-computer interaction equipment as described in claims 1, which is characterized in that described
Server also includes:
One user session administrative unit is coupled to the user to manage the operation workflow of the server in the financial services system
Interface administrative unit to input the input text word string, or exports the output character word string.
4. the financial services system based on intelligent sound human-computer interaction equipment as described in claims 3, which is characterized in that described
After user interface administrative unit receives the output character word string, i.e., with the speech-to-text unit, turn according to the sound/character
Complex voice data corresponding to the output character word string stored in corpus library unit is changed, which is turned
Be changed to the output voice, through the cloud network by the output voice transfer to the intelligent apparatus, then by the voice output/
Input equipment, to play out.
5. the financial services system based on intelligent sound human-computer interaction equipment as described in claims 1, which is characterized in that described
Server also includes: a financial relative conversations meaning of one's words processing module, wherein the finance relative conversations meaning of one's words processing module further include:
One Financial organization corpus library unit to store the vocabulary and term of complex Financial organization, and is coupled to the user session
Administrative unit；And
One natural language processing unit is coupled to the Financial organization corpus library unit and the user session administrative unit, in conjunction with this
Stored the complex Financial organization vocabulary and term in Financial organization corpus library unit, to handle the input text word string,
With the generation output character word string.
6. the financial services system based on intelligent sound human-computer interaction equipment as described in claims 5, which is characterized in that described
Natural language processing unit is to be optimized the identification result of speech recognition technology, so that optimum results and Financial organization art
Language/vocabulary data bank reach preferably compare effect, processing natural language method combine Da Mailiu-Levenstein away from
From, Dynamic Programming and Multiple Sequence Alignment at least one algorithm.
7. the financial services system based on intelligent sound human-computer interaction equipment as described in claims 5, which is characterized in that described
Financial relative conversations meaning of one's words processing module, also include a financial Information service unit, be coupled to the user session administrative unit and
One data storage device is ordered with being inputted according to the financial service, in conjunction with complex gold stored in the data storage device
Financing material, to carry out the financial service needed for user.
8. the financial services system based on intelligent sound human-computer interaction equipment as described in claims 7, which is characterized in that when described
Financial Information service unit complete the financial service after, that is, generate a financial service output report, and be transmitted to the use
Family dialogue management unit, when to calculate the financial service contained by the financial service output report defeated for the user session administrative unit
It reports out, when a contained report text word length is greater than a default value, then with the natural language processing unit, in conjunction with the gold
Melt complex Financial organization vocabulary and term stored in professional corpus library unit, it is defeated that this report text word string is converted to this
Text word string out is transferred to the user session administrative unit, and the user session administrative unit is again by the output character word string, transmission
Give the user interface administrative unit.
9. the financial services system based on intelligent sound human-computer interaction equipment as described in claims 5, which is characterized in that when described
Financial Information service unit complete the financial service after, that is, generate a financial service output report, and be transmitted to the use
Family dialogue management unit, when to calculate the financial service contained by the financial service output report defeated for the user session administrative unit
It reports out, when a contained report text word length is less than a default value, then to be stored up in the Financial organization corpus library unit
This report text word string is converted to the output character word string, is transferred to the use by the complex Financial organization vocabulary and term deposited
Family dialogue management unit, the user session administrative unit are transferred to the user interface administrative unit again by the output character word string.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710513344.6A CN109213877A (en) | 2017-06-29 | 2017-06-29 | A kind of financial services system based on intelligent sound human-computer interaction equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710513344.6A CN109213877A (en) | 2017-06-29 | 2017-06-29 | A kind of financial services system based on intelligent sound human-computer interaction equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109213877A true CN109213877A (en) | 2019-01-15 |
Family
ID=64960498
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710513344.6A Pending CN109213877A (en) | 2017-06-29 | 2017-06-29 | A kind of financial services system based on intelligent sound human-computer interaction equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109213877A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112150233A (en) * | 2019-06-26 | 2020-12-29 | 三竹资讯股份有限公司 | Device and method for controlling financial quotation of television application program by voice |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1291336C (en) * | 2003-01-14 | 2006-12-20 | 财团法人资讯工业策进会 | Multimode conversational system based on event hierarchy expression method |
| CN101017428A (en) * | 2006-12-22 | 2007-08-15 | 广东电子工业研究院有限公司 | Embedded voice interaction device and interaction method thereof |
| CN103760984A (en) * | 2014-01-24 | 2014-04-30 | 成都万先自动化科技有限责任公司 | Man-machine conversation system |
| CN104168270A (en) * | 2014-07-31 | 2014-11-26 | 腾讯科技（深圳）有限公司 | Identity verifying method, server, client side and system |
| CN105224278A (en) * | 2015-08-21 | 2016-01-06 | 百度在线网络技术（北京）有限公司 | Interactive voice service processing method and device |
| CN105810200A (en) * | 2016-02-04 | 2016-07-27 | 深圳前海勇艺达机器人有限公司 | Man-machine dialogue apparatus and method based on voiceprint identification |
| CN105895101A (en) * | 2016-06-08 | 2016-08-24 | 国网上海市电力公司 | Speech processing equipment and processing method for power intelligent auxiliary service system |
| CN106023993A (en) * | 2016-07-29 | 2016-10-12 | 西安旭天电子科技有限公司 | Robot control system based on natural language and control method thereof |
| US20160351206A1 (en) * | 2015-05-26 | 2016-12-01 | Speaktoit, Inc. | Dialog system with automatic reactivation of speech acquiring mode |
| CN106550156A (en) * | 2017-01-23 | 2017-03-29 | 苏州咖啦魔哆信息技术有限公司 | A kind of artificial intelligence's customer service system and its implementation based on speech recognition |
-
2017
- 2017-06-29 CN CN201710513344.6A patent/CN109213877A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1291336C (en) * | 2003-01-14 | 2006-12-20 | 财团法人资讯工业策进会 | Multimode conversational system based on event hierarchy expression method |
| CN101017428A (en) * | 2006-12-22 | 2007-08-15 | 广东电子工业研究院有限公司 | Embedded voice interaction device and interaction method thereof |
| CN103760984A (en) * | 2014-01-24 | 2014-04-30 | 成都万先自动化科技有限责任公司 | Man-machine conversation system |
| CN104168270A (en) * | 2014-07-31 | 2014-11-26 | 腾讯科技（深圳）有限公司 | Identity verifying method, server, client side and system |
| US20160351206A1 (en) * | 2015-05-26 | 2016-12-01 | Speaktoit, Inc. | Dialog system with automatic reactivation of speech acquiring mode |
| CN105224278A (en) * | 2015-08-21 | 2016-01-06 | 百度在线网络技术（北京）有限公司 | Interactive voice service processing method and device |
| CN105810200A (en) * | 2016-02-04 | 2016-07-27 | 深圳前海勇艺达机器人有限公司 | Man-machine dialogue apparatus and method based on voiceprint identification |
| CN105895101A (en) * | 2016-06-08 | 2016-08-24 | 国网上海市电力公司 | Speech processing equipment and processing method for power intelligent auxiliary service system |
| CN106023993A (en) * | 2016-07-29 | 2016-10-12 | 西安旭天电子科技有限公司 | Robot control system based on natural language and control method thereof |
| CN106550156A (en) * | 2017-01-23 | 2017-03-29 | 苏州咖啦魔哆信息技术有限公司 | A kind of artificial intelligence's customer service system and its implementation based on speech recognition |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112150233A (en) * | 2019-06-26 | 2020-12-29 | 三竹资讯股份有限公司 | Device and method for controlling financial quotation of television application program by voice |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US12032920B2 (en) | Direct speech-to-speech translation via machine learning | |
| US9753914B2 (en) | Natural expression processing method, processing and response method, device, and system | |
| WO2021160191A1 (en) | Human-in-the-loop voice communication system and method | |
| TWI712975B (en) | A voice-activated system for financial information services | |
| CN110019688A (en) | The method that robot is trained | |
| KR20200092499A (en) | Method and apparatus for counseling support using interactive artificial intelligence technology | |
| CN110046232A (en) | Natural expression processing method, response method, equipment and the system of natural intelligence | |
| CN110059168A (en) | The method that man-machine interactive system based on natural intelligence is trained | |
| Scotti et al. | A modular data-driven architecture for empathetic conversational agents | |
| CN110059166A (en) | Natural expression processing method, response method, equipment and the system of natural intelligence | |
| US11960852B2 (en) | Robust direct speech-to-speech translation | |
| EP3573050B1 (en) | Computing platform and method for modifying voice data | |
| JP3667615B2 (en) | Spoken dialogue method and system | |
| CN110008317A (en) | Natural expression processing method, response method, equipment and the system of natural intelligence | |
| CN110059167A (en) | Natural expression processing method, response method, equipment and the system of natural intelligence | |
| CN109213877A (en) | A kind of financial services system based on intelligent sound human-computer interaction equipment | |
| KR20180121120A (en) | A machine learning based voice ordering system that can combine voice, text, visual interfaces to purchase products through mobile divices | |
| CN111651582B (en) | Method and system for simulating user speaking | |
| JPH10143485A (en) | Communication method and data processing system | |
| Garg et al. | Automation and presentation of word document using speech recognition | |
| Shirai et al. | Dialogue AI for financial services | |
| Houidhek et al. | Dnn-based speech synthesis for arabic: modelling and evaluation | |
| Zhang et al. | Emotional speech synthesis based on DNN and PAD emotional state model | |
| Mehta et al. | Fake it to make it: Using synthetic data to remedy the data shortage in joint multimodal speech-and-gesture synthesis | |
| TW201923621A (en) | Conversational financial service device and system comprising a transceiver module, a control module and an emotion recognition module |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20190115 |
|
| WD01 | Invention patent application deemed withdrawn after publication |