CN1291336C - Multimode conversational system based on event hierarchy expression method - Google Patents
Multimode conversational system based on event hierarchy expression method Download PDFInfo
- Publication number
- CN1291336C CN1291336C CN 03101671 CN03101671A CN1291336C CN 1291336 C CN1291336 C CN 1291336C CN 03101671 CN03101671 CN 03101671 CN 03101671 A CN03101671 A CN 03101671A CN 1291336 C CN1291336 C CN 1291336C
- Authority
- CN
- China
- Prior art keywords
- meaning
- system based
- incident
- user
- conversational system
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Lifetime
Links
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention relates to a multimode dialog system based on an event hierarchy expression method, and the multimode dialog system is composed of an interface die set, an interpretation die set and a dialog management die set, wherein the interface die set has a multimode input function and an output function. After the system receives multimode input information from an input unit, the input information is identified and interpreted by the interpretation die set, and the input information is converted into a semanteme frame of a common format; then, the dialog state is updated by the dialog management die set to carry out theme inference, the dialog management die set is connected with servers of a back-end database server, a network resource server, a network resource server, a knowledge processing server, or an information searching server, etc. to obtain relevant information required by a user; finally, multimedia output is generated by an information dispatching unit, a language generation unit and a speech synthesis unit of the interpretation die set to be interacted with the user.
Description
Technical field
The present invention refers to a kind of multi-mode conversational system based on the incident hierarchical representation especially about a kind of multi-mode conversational system.
Background technology
Known dialogue management technology, most dialog model based on the list formula, and build the voice extendability SGML putting and manage easily and come (voice extensiblemarkup language to use according to industrywide standard, VXML) write the voice dialogue flow process, and form a list formula dialog model.With the voice seat reservation system is example, when the user desires booking tickets, must be after linking to the voice seat reservation system, use telephone key-press or phonetic entry according to the guiding of its set flow process after, can finish the ticket booking program.
Yet military order user's dialogue behavior is very restricted thus, and system response is stiff, only can be according to the in addition operation of set flow process, thereby lack practicality and use elasticity; Again, known speech dialogue system is only supported the phonetic entry of single-mode usually, and can not integrate for example input format such as literal input or GUI, can't satisfy the user demand of user's diversification; And known speech dialogue system is not supported multi-threaded switching, for example the user inquires that at the beginning " Taibei " arrive the rail fare in " platform south ", when confirming to book tickets, system will inquire again that the user desires rising up to the place of booking tickets, increases puzzlement on foot and prolongs the activity duration afterwards.In addition, known meter formula in daytime conversational system uses artificial intelligence's program language to carry out the logical deduction computing, and this mode is difficult for expressing the object data structure in the conversation object, also is difficult for the computing of handling procedure, is not very good.Hence one can see that, and known dialog management system still has many shortcomings and gives improved necessity.
Summary of the invention
Fundamental purpose of the present invention uses the incident hierarchical representation to describe dialogue management to see through logical relation and object, so that can express the conversation process of plan formula easily for a kind of multi-mode conversational system based on the incident hierarchical representation is being provided.
Another object of the present invention is used descriptive language with expression object data structure, and is carried out procedural computing, with the dialogue management of convenient plan formula for a kind of multi-mode conversational system based on the incident hierarchical representation is provided.
A further object of the present invention can be striden the data of theme and inherit, and handle multi-threaded switching problem, so that to present the dialogue behavior of wisdom for a kind of multi-mode conversational system based on the incident hierarchical representation is provided.
Another purpose of the present invention is for providing a kind of multi-mode conversational system based on the incident hierarchical representation, so that can accept voice, literal, and the input pattern of list button figures interface, to form friendly multi-mode user's interface.
For reaching above-mentioned purpose, the multi-mode conversational system based on the incident hierarchical representation proposed by the invention mainly comprises an interface module, a decipher module and a dialogue management module.Wherein, the interface module is in order to linking up interaction with the user, and it includes one in order to receiving an input information, input block, an and output unit; Decipher module identification input information, and be converted into meaning of one's words news frame, have in this meaning of one's words news frame keyword (key), numerical value (value), with confidence value (confidence); The dialogue management module then includes a subject description XML file, an and processor, its subject description XML file stores a plurality of conversation subjects and corresponding interactive program thereof, and and or figure (and-or graph) express the logical relation between each conversation process, each interactive program has at least one action command, this action command be selected from following one of them: in order to inquire crucial information, in order to confirm crucial information, and in order to upgrade crucial information, processor is fechtable meaning of one's words news frame then, and use a program language to find out corresponding conversation subject in the XML file to describe from theme according to the keyword of meaning of one's words news frame, and will be converted to an action strategy corresponding to each action command that interactive program had of this conversation subject in regular turn to export output unit to, so that carry out interaction with the user.
Wherein, this dialogue management module includes a historical summary storehouse, changes keyword in the salty previous meaning of one's words news frame in order to store this user in previous input information.
Wherein, a dialogue theme of correspondence found out in the keyword that this processor is interrogated frame according to the present meaning of one's words in this subject description XML file, and find out the keyword of this previous meaning of one's words news frame with in the extremely present meaning of one's words news of the succession frame in this historical summary storehouse certainly.
It uses the extendability SGML to describe these conversation subjects.
Wherein, this program language is the JavaScript language.
Wherein, this input block is a voice device, and this input information is a voice data, and this decipher module has a speech recognition unit, in order to this voice data of identification, re-uses natural language understanding technology this voice data is converted to meaning of one's words news frame.
Wherein, this input information is a written historical materials, and this decipher module uses natural language understanding technology this literal data is converted to meaning of one's words news frame.
Wherein, this input block is a GUI, and this GUI includes at least one option and clicks for the user, and this decipher module captures input information that option that this user clicks forms to be converted to meaning of one's words news frame.
Wherein, this output unit is exported a written historical materials.
Wherein, this output unit is exported a voice data.
Wherein, this output unit is exported a chart menu.
Description of drawings
For further specifying technology contents of the present invention, below in conjunction with embodiment and accompanying drawing the present invention is done a detailed description, wherein:
Fig. 1 is the functional block diagram of the embodiment of the invention.
Fig. 2 is the synoptic diagram of embodiment of the invention interface module.
Fig. 3 be among the present invention the XML tag language description and or figure.
Fig. 4 is the synoptic diagram of the interactive program of first embodiment of the invention ticket booking theme.
Fig. 5 is the synoptic diagram that second embodiment of the invention is upgraded the interactive program of ticket booking theme.
Fig. 6 is the synoptic diagram that third embodiment of the invention is switched conversation subject.
Embodiment
Fig. 1 is please consulted in relevant preferred embodiment of the present invention earlier, the multi-mode conversational system that shows present embodiment by interface module 1, decipher module 2, and dialogue management module 3 formed.Wherein, interface module 1 is in order to link up interaction with the user, it has multimodal input block 11 in order to receive by the input information that the user imported, for example the voice messaging of being imported by audio communication interface 111 and computer telephone interface 112, Word message, the input information that GUI 114 list that provides is provided and the image information of being imported by video camera 115 imported by literal interface 113 ... Deng; Interface module 1 also has multi-mode output unit 12 in order to output chart and menu 121, literal 122, people's face and phonetic synthesis 123 formed moving daytime and voice ... Deng.Therefore, when the user uses the multi-mode conversational system of present embodiment, will carry out interaction with interface module 1 as shown in Figure 2, it is an example with the train seat reservation system, since system combination the human face detection technology, therefore video camera 115 can detect the people near and begin dialogue, certainly the user also can use the natural spoken language conversational mode to order train ticket to see through audio communication interface 111, use the GUI 114 and the system mutual of touch-control screen, or use literal interface 113 to import the request of desire inquiry admission fee; And form the procuratorial image of a traffic information service by people's face synthesizes the design that reaches phonetic synthesis output unit 123, can allow the user feel talking with true man.
12 of decipher modules can be with multimodal input information in addition after the identification decipher, be converted to the meaning of one's words news frame of co expression mode, have in this meaning of one's words news frame " keyword (key) ", " numerical value (value) ", with " confidence value (confidence) ", with as the input of achievement dialogue management module afterwards; For example " keyword " is tomorrow, if be 91/12/17 today, what then " numerical value " just can be corresponding is calculated as 91/12/18: " confidence value " then has other value according to different input patterns, if for example 90 expressions of the confidence value of speech recognition identification accuracy is higher, if 50 expression identification accuracy are lower, and the input of literal or GUI is comparatively reliable usually, so the confidence value is 100.Different input patterns, other pre-treatment program is arranged, for example if input information is a voice messaging, then need see through speech recognition unit 21 and analyze information in the voice messaging, seeing through natural language understanding unit 22 is again resolved, to form meaning of one's words news frame:, be converted to meaning of one's words news frame after can directly using natural language understanding unit 22 to analyze if input information is a Word message; If input information is provided by GUI 114 or video camera 115, then need see through and be converted to meaning of one's words news frame after information interpretation unit 23 is analyzed.Because above-mentioned speech recognition unit 21, natural language understanding unit 22, and the function of information interpretation unit 23 be to reach by known technology, so do not give unnecessary details the detailed motion flow of above-mentioned each unit at this.In addition, decipher module 12 still includes an information dispatch unit 24, a language generation unit 25, reaches a phonetic synthesis unit 26, in order to produce multimedia output information to be sent to output unit 12.
Dialogue management module 3 upgrades dialogue state with dialogue management mechanism, carries out the theme inference, and links the subject description XML file 32 and the historical summary storehouse 33 of rear end, is moved to produce the required interactive program of user.Wherein, subject description XML file 32 stores a plurality of conversation subjects, for example book tickets theme, date theme, admission fee theme, and weather theme etc., and use the logical relation and the object of extendability SGML (XML) to describe the incident level, and each conversation subject has all defined an interactive program, wherein having at least one action command can be in order to inquiry, confirm or upgrade crucial information, and the theme of for example booking tickets just must be to " getting a lift the date ", " rising up to the station ", " car type train number ", reach crucial information such as " ticket kind numbers " carries out above-mentioned action; The incident level reaches the logical organization of different conversation object, available with or the mode of figure (and-or graph) express, see also Fig. 3; Among the present invention be with the XML tag language description with or the structure of figure, comprise " or (or) mark ", " with (and) mark ", " action (action) mark ", " condition (cond) mark ", " auxiliary (help) mark " and " description program (script) mark " etc.; For example book tickets theme, date theme, admission fee theme, and each theme such as weather theme between or relation (or), as long as just there is a theme selected, can carry out follow-up conversation process; " getting a lift the date " under the ticket booking theme, " rise up to station ", " car type train number ", and " ticket kind number " then be relation with (and), represent just can finish after these crucial information all must be learnt the theme of ticket booking; " action marker " describe each child node the corresponding action of carrying out, for example inquire crucial information; " condition flag " then in order to judge whether each child node has reached target, whether for example all learn, if condition is not reached, then carries out " action marker " defined program by rising up to the station of ticket booking; The processor 31 of present embodiment dialogue management module 3 uses the JavaScript language to represent the object data structure, carry out the computing of theme inference with the keyword of interrogating in the frame for the meaning of one's words, for example when having a keyword " admission fee " in the meaning of one's words news frame, processor 31 can be described from theme and find out the admission fee theme in the XML file 32 and carry out corresponding interactive program; Historical summary storehouse 33 then can store the user and change keyword in the salty previous meaning of one's words news frame in previous input information, and its concrete function will be in describing in detail hereinafter.
Below the interactive situation that the multi-mode conversational system is produced with the user is described under different situations for three embodiment.
First embodiment: in order to resolve and to carry out conversation subject:
The user sees through the voice messaging of computer telephone interface 112 inputs " I will arrive Kaohsiung tomorrow ", therefore see through the analysis of speech recognition unit 21 and natural language understanding unit 22, to be converted to the meaning of one's words news frame that includes keyword " Kaohsiung " to input information, and calculate via the processor 31 of dialogue management module 3 again and infer that this meaning of one's words news frame is corresponding to the ticket booking theme.See also the book tickets interactive program of theme 321 of Fig. 4, ticket booking theme 321 definition of present embodiment have five action commands to carry out interaction to export output unit 12 to the user to be converted to the action strategy in regular turn, promptly inquire, confirm, and upgrade " selection ", " option date ", " selection vehicle shift ", " ticket kind number ", and crucial information such as " affirmation ", treat that affirmation at last is errorless then can finish the ticket booking program afterwards up to the station.Since the meaning of one's words of changing out according to input information news frame include the keyword in " Kaohsiung " and speech recognition the confidence mark if the confidence mark greater than a default hurdle value, then this keyword can directly be substituted by crucial information, in this example, " Kaohsiung " promptly is substituted by the crucial information at " up to the station ", thus, as long as system is inquiry " rise station " information again, again repeat queries " up to standing " information once.
Be noted that because input block 11 can receive multimodal input information, so the dialog management system of present embodiment can use different affirmations and corrigendum mechanism to calculate the confidence mark of keyword and crucial information at the input of different mode.When for example using GUI 114 formed lists to import, because the user is the button that selects that clicks on the set menu, so the confidence mark is full marks; Or the literal information of using literal interface 113 to be imported, have high confidence mark equally; Usually the confidence mark of the voice messaging of being imported with audio communication interface 111 or computer telephone interface 112 is lower, if when having higher confidence mark, system also can adopt this keyword as crucial information automatically, and need not inquire that the user is to obtain crucial information certainly.
Second embodiment: in order to upgrade the interactive program of dialogue news frame:
See also Fig. 5, also based on ticket booking theme 321, its interactive flow process is all identical with first embodiment, only when rising after station, date, vehicle shift, ticket kind number have all been inquired and finished, when carrying out the affirmation of ticket kind number, the user has revised the date, therefore must re-execute the interactive program of the theme 321 of once booking tickets.Because the crucial information after the first forward direction affirmation that the user inquires is stored in the historical summary storehouse 33, so when carrying out interactive program again, system will capture confirmed crucial information and not need to inquire again the user, promptly " rise " and reach " ticket kind number " up to the station, " date " and relative " vehicle shift " then must be chosen again, and forms the action command shown in ticket booking theme 321 '.Wherein, the stored crucial information in historical summary storehouse 33 will and suspend theme and bring in constant renewal in along with the up-to-date dialogue state of each conversation subject.
The 3rd embodiment: in order to switch conversation subject:
See also Fig. 6, also be illustrated with ticket booking theme 321, when the user when booking tickets the interactive program of theme 321, for example when proceeding to the interactive program of option date, user's first inquiry system date today of possibility, therefore system will automatically switch to the problem that the user answered in date theme 322, and after this interactive program is finished, the tall and erect class information that the ticket booking theme 321 that system can in time automatically switch back original is desired to take to continue the inquiry user.And in the interactive program of selecting vehicle shift, the user has inquired the admission fee information, so system will switch to admission fee theme 323 again, and after initiatively switching back ticket booking theme 321, the user has inquired weather conditions again, therefore system will carry out the weather conditions of interactive program to inform that the user is correlated with of weather theme 324 equally, switches back ticket booking theme 321 again.Each above-mentioned theme switches, and all be the theme comparatively relevant with ticket booking, so system can the former ticket booking theme of unconditional switchback.If other are during with ticket booking more incoherent theme such as road condition query, flight inquiring, the former ticket booking theme of switchback again after system then can confirm earlier.
Among above-mentioned the 3rd embodiment, the interactive program of user and system can use following natural dialogue situation to be simulated, and wherein, system end is the multi-mode conversational system of present embodiment:
1. system end: welcome, may I ask you and want what service?
2. user: I want the improve oneself number of the Taibei to Kaohsiung.
Which day do 3. system end: you want set out?
4. user: What day is tomorrow?
5. system end: be Saturday tomorrow.
6. user:
7. system end: may I ask you and want when set out tomorrow?
Be 8. user: how many admission fees?
9. system end: the Taibei is 845 yuan to the number of the improving oneself ticket in Kaohsiung.
10. user: taro light number?
11. system end: the Taibei is 651 yuan to the number of the improving oneself ticket in Kaohsiung.
12. user:
Will 13. system end: you order taro light number?
(14. the user presses the button of " correctly ")
15. system end: the train number that please select you to want.
16. user: I want to look into the weather in Kaohsiung.
17. system end: tomorrow Kaohsiung weather cloudy thunder shower when cloudy, the highest temperature 32 degree, lowest temperature 27 degree.
18. user: that Taibei?
19. system end: tomorrow the Taibei the cloudy afternoon thunder shower of weather, the highest temperature 34 degree, lowest temperature 26 degree.
20. user:
21. system end: the train number that please select you to want.
(22. the user clicks taro light number 17 train numbers in the list, and the Taibei is to Kaohsiung)
23. system end: you to book any ticket how many?
24. user: I want two of adult tickets.
25. system end: you will buy is to cover two of light number 17 train adult tickets with mat tomorrow from the Taibei to Kaohsiung, comes to 1302 yuan.If determine, please brush your credit card.
(26. the user presses " determining " button)
27. system end: please take out your ticket, thank you, and welcome you to patronize once more.
Wherein, the 4th row, eighth row, and the dialogue of the 16th row represent that the user switches to date theme 322, admission fee theme 323, and weather theme 324 by ticket booking theme 321 respectively; The 7th row, the 13rd row, and the dialogue of the 21st row then represent originally the booked tickets program of theme 321 of the proper and initiative switchback of system end.
According to above-mentioned explanation, show that the multi-mode conversational system based on the incident hierarchical representation proposed by the invention is but that multi-threaded switching problem is handled on wisdom ground, presenting the dialogue behavior of wisdom: and multimodal user's interface has comprised voice, literal, and the interface of list button figuresization, help the user with the most natural, mode and system carry out interaction easily; In addition, the description of dialogue management of the present invention is based on logical relation and object, it use to extend SGML and defines the logical relation that conversation object and object are asked, and represent the object data structure and carry out procedural computing with descriptive language (script language), be easy to be transplanted to different language, having and highly use elasticity, is a much progress in fact.
The foregoing description only is to give an example for convenience of description, and the interest field that the present invention advocated should be as the criterion so that claim is described certainly, but not only limits to the foregoing description.
Claims (11)
1. the multi-mode conversational system based on the incident hierarchical representation is characterized in that, mainly comprises:
One interface module, in order to linking up interaction with the user, it includes one in order to the input block that receives an input information, an and output unit;
One decipher module, this input information of identification also are converted into meaning of one's words news frame, this meaning of one's words news frame have keyword, numerical value, with the confidence value; And
One dialogue management module, it includes:
One subject description XML file, store a plurality of conversation subjects and corresponding interactive program thereof, each interactive program has at least one action command, this action command be selected from following one of them: in order to inquire crucial information, in order to confirm crucial information, and in order to upgrade crucial information; And
One processor, capture this meaning of one's words news frame, and use a program language in this subject description XML file, to find out a corresponding dialogue theme according to the keyword of this meaning of one's words news frame, and will be converted to an action strategy corresponding to each action command that interactive program had of this conversation subject in regular turn to export this output unit to, so that carry out interaction with the user.
2. the multi-mode conversational system based on the incident hierarchical representation as claimed in claim 1, it is characterized in that, wherein, this dialogue management module includes a historical summary storehouse, in order to store the keyword of this user in the previous meaning of one's words news frame that previous input information converted to.
3. the multi-mode conversational system based on the incident hierarchical representation as claimed in claim 2, it is characterized in that, wherein, a dialogue theme of correspondence found out in the keyword that this processor is interrogated frame according to the present meaning of one's words in this subject description XML file, and find out the keyword of this previous meaning of one's words news frame with in the extremely present meaning of one's words news of the succession frame in this historical summary storehouse certainly.
4. the multi-mode conversational system based on the incident hierarchical representation as claimed in claim 1 is characterized in that, it uses the extendability SGML to describe these conversation subjects.
5. the multi-mode conversational system based on the incident hierarchical representation as claimed in claim 1 is characterized in that, wherein, this program language is the JavaScript language.
6. the multi-mode conversational system based on the incident hierarchical representation as claimed in claim 1, it is characterized in that, wherein, this input block is a voice device, this input information is a voice data, and this decipher module has a speech recognition unit, in order to this voice data of identification, re-uses natural language understanding technology this voice data is converted to meaning of one's words news frame.
7. the multi-mode conversational system based on the incident hierarchical representation as claimed in claim 1 is characterized in that, wherein, this input information is a written historical materials, and this decipher module uses natural language understanding technology this literal data is converted to meaning of one's words news frame.
8. the multi-mode conversational system based on the incident hierarchical representation as claimed in claim 1, it is characterized in that, wherein, this input block is a GUI, this GUI includes at least one option and clicks for the user, and this decipher module captures input information that option that this user clicks forms to be converted to meaning of one's words news frame.
9. the multi-mode conversational system based on the incident hierarchical representation as claimed in claim 1 is characterized in that wherein, this output unit is exported a written historical materials.
10. the multi-mode conversational system based on the incident hierarchical representation as claimed in claim 1 is characterized in that, wherein, this output unit is exported a voice data.
11. the multi-mode conversational system based on the incident hierarchical representation as claimed in claim 1 is characterized in that, wherein, this output unit is exported a chart menu.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 03101671 CN1291336C (en) | 2003-01-14 | 2003-01-14 | Multimode conversational system based on event hierarchy expression method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 03101671 CN1291336C (en) | 2003-01-14 | 2003-01-14 | Multimode conversational system based on event hierarchy expression method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1517901A CN1517901A (en) | 2004-08-04 |
| CN1291336C true CN1291336C (en) | 2006-12-20 |
Family
ID=34281442
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 03101671 Expired - Lifetime CN1291336C (en) | 2003-01-14 | 2003-01-14 | Multimode conversational system based on event hierarchy expression method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1291336C (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109213877A (en) * | 2017-06-29 | 2019-01-15 | 永丰商业银行股份有限公司 | A kind of financial services system based on intelligent sound human-computer interaction equipment |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7970730B2 (en) * | 2005-01-27 | 2011-06-28 | Microsoft Corporation | Efficient data access via runtime type inference |
| CN106503156B (en) * | 2016-10-24 | 2019-09-03 | 北京百度网讯科技有限公司 | Man-machine interaction method and device based on artificial intelligence |
-
2003
- 2003-01-14 CN CN 03101671 patent/CN1291336C/en not_active Expired - Lifetime
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109213877A (en) * | 2017-06-29 | 2019-01-15 | 永丰商业银行股份有限公司 | A kind of financial services system based on intelligent sound human-computer interaction equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1517901A (en) | 2004-08-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20160259779A1 (en) | Evidence-Based Natural Language Input Recognition | |
| CN109522083B (en) | Page intelligent response interaction system and method | |
| CN111368538B (en) | Voice interaction method, system, terminal and computer readable storage medium | |
| CN105808200B (en) | Intelligent automation assistant | |
| CN100338558C (en) | Systems and methods for inputting graphical data into a graphical input field | |
| CN102439661A (en) | Service oriented speech recognition for in-vehicle automated interaction | |
| CN110263324A (en) | Text handling method, model training method and device | |
| US20190120649A1 (en) | Dialogue system, vehicle including the dialogue system, and accident information processing method | |
| CN101292282A (en) | Mobile systems and methods of supporting natural language human-machine interactions | |
| WO2008113063A1 (en) | Speech-centric multimodal user interface design in mobile technology | |
| CN1639707A (en) | System and method for concurrent multimodal communication | |
| CN104123936A (en) | Method for automatic training of a dialogue system, dialogue system, and control device for vehicle | |
| CN1639696A (en) | System and method for concurrent multimodal communication session persistence | |
| CN103000173A (en) | Voice interaction method and device | |
| CN102428440A (en) | System and method for multi-modal input synchronization and disambiguation | |
| US20210201169A1 (en) | Systems and methods for collecting and processing data for insurance-related tasks | |
| CN1647023A (en) | Voice-controlled data entry | |
| CN110720098B (en) | Adaptive interface in voice activated networks | |
| CN111098312A (en) | Window government affairs service robot | |
| CN201869323U (en) | Stb | |
| US20130325483A1 (en) | Dialogue models for vehicle occupants | |
| CN114186108A (en) | Multimode man-machine interaction system oriented to electric power material service scene | |
| CN112148874A (en) | Intention identification method and system capable of automatically adding potential intention of user | |
| CN109670020A (en) | A kind of voice interactive method, system and device | |
| CN1291336C (en) | Multimode conversational system based on event hierarchy expression method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CX01 | Expiry of patent term | ||
| CX01 | Expiry of patent term |
Granted publication date: 20061220 |