CN1271793C - Apparatus and method for computation process of natural language using hand held communication equipment - Google Patents
Apparatus and method for computation process of natural language using hand held communication equipment Download PDFInfo
- Publication number
- CN1271793C CN1271793C CNB031020682A CN03102068A CN1271793C CN 1271793 C CN1271793 C CN 1271793C CN B031020682 A CNB031020682 A CN B031020682A CN 03102068 A CN03102068 A CN 03102068A CN 1271793 C CN1271793 C CN 1271793C
- Authority
- CN
- China
- Prior art keywords
- mentioned
- natural language
- hand
- voice
- natural
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Images
Landscapes
- Machine Translation (AREA)
Abstract
The present invention relates to an apparatus for the computation processing of a natural language using hand held communication equipment, which is used for receiving natural speech input in the hand held communication equipment, inputting natural speech in the hand held communication equipment and transmitting a result response after the computation processing. The apparatus comprises an automatic speech-identification unit, a natural language understanding unit and an action and response unit, wherein the automatic speech-identification unit receives the natural speech input and produces an automatic speech-identification result after characteristic picking and identification; the natural language understanding unit receives the automatic speech-identification result and produces a natural language understanding result after understanding and analysis; the action and response unit receives the natural language understanding result and produces the result response through proper processing.
Description
Technical field
The invention relates to a kind of device of hand-hold communication appliance processing language, particularly relevant a kind of device and method with hand-hold communication appliance computing natural language.
Background technology
Along with the progress of mechanics of communication, the use of hand-hold communication appliance (handheld communication device) is also constantly popularized thereupon.The development of hand-hold communication appliance at present has two main trend, and the size of first hand-hold communication appliance is more and more littler, and it two is that the computing capability (computingpower) of hand-hold communication appliance is more and more stronger with ability to communicate (communication capability).In a foreseeable future, every computing function and communication function are integrated in the single hand-hold communication appliance, are inevitable developing direction.Therefore, the acoustic control function of controlling with voice (speech) just becomes a ring important in the hand-hold communication appliance technology.
The acoustic control function of existing hand-hold communication appliance is to be controlled to be the master with order (command), and meaning is that user's input command is in order to control function specific in the hand-hold communication appliance.For example, but user's input dial, transmit voice such as news in brief, shutdown to carry out dialing in the hand-hold communication appliance, to transmit functions such as news in brief and shutdown.But the hand-hold communication appliance of these acoustic controls, no matter be mobile phone or PDA(Personal Digital Assistant), its speech recognition technology roughly is the order speech data elder generation process pre-process (pre-processing) with input, capture characteristic parameter, and then compare with acoustic model or sound template (template) that training in advance is finished, the best comparison result that obtains at last is identification result.
Foregoing speech recognition technology does not relate to the technology that the meaning of one's words is understood (understanding), when being not the control command of fixing as if the voice of importing, then there is no preferable processing method with existing technology.The language that the usual language form that uses of right general user is not order control, but natural language (natural language).Moreover because the function of personal digital assistant application program is more seen complexity, as travel schedule, address book, notepad etc., only utility command is controlled and is controlled these application programs and dislike not enough slightly, and can't cooperate the design of its man-machine interface fully.Therefore, hand-hold communication appliance could be in response to following actual technical development and user demand when having the ability of computing natural language.
Relevant technology is found in " JUPITER:A Telephone-Based ConversationInterface for Weather Information; " IEEE Trans.Speech and Audio Proc, 8 (1), 85-96,2000. and United States Patent (USP) US005749072 number, " Communicationsdevice responsive to spoken commands and methods of using same. "
Summary of the invention
In view of this, a purpose of the present invention is with hand-hold communication appliance computing natural language, the user can directly use the expression way of natural language, inform hand-hold communication appliance user's intention, and hand-hold communication appliance can through understanding and analyzing, be learnt user's intention by its calculating and disposal ability with the natural language that the user imported, according to user's intention of being learnt, utilize its ability to communicate to be carried out or finish again.For example, the user may import the input of natural language expressing modes such as " tonight remind me meet sb. at the airport to the airport ", " tell among me the high head part in mountain highway section whether traffic congestion ", " can rain tomorrow in the Taibei " at 8, hand-hold communication appliance is then understood the phonetic entry process and is analyzed, and carries out work such as prompting or inquiry.
Another object of the present invention is that the natural language processing unit is integrated in the hand-hold communication appliance.In other words, in single hand-hold communication appliance, just functions such as voice reception, speech recognition be can carry out, and required inquiry and communication function carried out by wireless telecommunications and network.Be different from present and receive voice, be sent to remote server and carry out speech recognition, again identification result is back to the function mode of hand-hold communication appliance, avoid simultaneously wasting frequency range because of data such as transmission feature parameters with hand-hold communication appliance.
For reaching above-mentioned all purposes, the present invention proposes a kind of device with hand-hold communication appliance computing natural language, be used to receive the nature phonetic entry in the hand-held communication apparatus, and natural-sounding is input in the hand-hold communication appliance, through spreading out of result's response after the computing, comprise the automatic identification unit of voice, natural language understanding unit and action and response unit.The natural-sounding input is meant the voice that general user is imported in the natural language expressing mode.
The automatic identification unit of voice, it places hand-hold communication appliance, in order to receiving the nature phonetic entry, and feature acquisition and identification is carried out in the natural-sounding input, produces the automatic identification result of voice.The automatic identification unit of voice comprises natural-sounding loader, phonetic feature acquisition device and speech recognition device.The natural-sounding loader is user's interface, in order to receive the nature phonetic entry.The phonetic feature acquisition device is coupled to the natural-sounding loader, in order to the phonetic feature of acquisition from the natural-sounding input of natural-sounding loader.The speech recognition device is coupled to the phonetic feature acquisition device, in order to reference language structural database and speech model database, and the phonetic feature of the natural-sounding input that identification phonetic feature acquisition device is captured, and produce the automatic identification result of voice.
The natural language understanding unit, it places hand-hold communication appliance, and is coupled to the automatic identification unit of voice, in order to the automatic identification result of reception voice, and with voice automatic identification result process understanding and analysis, produces the natural language understanding result.The natural language understanding unit comprises syntax analysis device, key word analysis device and semantic structure manager.The syntax analysis device in order to receive the automatic identification result of voice, is analyzed with reference to grammatical database and to the syntax of the automatic identification result of voice.The key word analysis device is coupled to the syntax analysis device, in order to receiving the automatic identification result of voice, and the keyword of the automatic identification result of voice is analyzed.The semantic structure manager is coupled to syntax analysis device and key word analysis device, in order to simultaneously with reference to syntax analysis device and key word analysis device for the analysis of the automatic identification result of voice, produce the natural language understanding result.
Action and response unit, it places hand-hold communication appliance, and is coupled to the natural language understanding unit, in order to receiving the natural language understanding result, and the natural language understanding result is suitably handled, and response bears results.Action comprises information manager, natural language generator and sound wave synthesizer with response unit.Information manager in order to receiving the natural language understanding result, and according to the natural language understanding result, is found out required semantic structure, and the mode that semantic structure can meaning of one's words framework (semantic frame) is expressed.The natural language generator is coupled to information manager, in order to the semantic structure of finding out according to information manager, forms the form of natural language.The sound wave synthesizer is coupled to the natural language generator, in order to the natural language that the natural language generator is formed, and synthetic sound wave, and the response that bears results.
Moreover the present invention proposes a kind of method with hand-hold communication appliance computing natural language, is used to receive the nature phonetic entry in the hand-hold communication appliance, and natural-sounding is input in the hand-hold communication appliance, through spreading out of result's response after the computing.The natural-sounding input is meant the voice that general user is imported in the natural language expressing mode.
At first, hand-hold communication appliance receives the nature phonetic entry, the phonetic feature of acquisition natural-sounding input, and reference language structural database and speech model database, the phonetic feature of the natural-sounding input that identification captured, and produce the automatic identification result of voice.
Then, hand-hold communication appliance is analyzed the syntax of the automatic identification result of voice with reference to grammatical database, and the keyword of the automatic identification result of voice is analyzed, and according to the analysis of the automatic identification result of voice, produces the natural language understanding result.
At last, hand-hold communication appliance is found out required semantic structure according to the natural language understanding result, and according to the semantic structure of being found out, forms the form of natural language, and with the synthetic sound wave of natural language, response bears results.
Description of drawings
Fig. 1 is the Organization Chart that shows hand-hold communication appliance and network among the disclosed embodiment;
Fig. 2 is the functional schematic that shows hand-hold communication appliance among the disclosed embodiment;
Fig. 3 shows functional block diagram of the present invention;
Fig. 4 shows flowchart of the present invention;
Fig. 5 is the schematic diagram that shows the natural language identification syntax of disclosed embodiment;
Fig. 6 is the schematic diagram that shows the syntax analysis of disclosed embodiment;
Fig. 7 is the schematic diagram that shows the natural language understanding result of disclosed embodiment;
Fig. 8 is the schematic diagram of semantic structure that shows the natural language form of disclosed embodiment;
Fig. 9 is the schematic diagram that shows the syntax analysis of disclosed embodiment;
Figure 10 is the schematic diagram of semantic structure that shows the natural language form of disclosed embodiment.
The reference numeral explanation:
100,102-hand-hold communication appliance; 104,106, the 108-webserver;
The 110-internet; The 200-hand-hold communication appliance;
The 202-display unit; The 204-CPU;
The 206-storage device; The 208-I/O device;
The 209-radio network interface; The 210-wireless network;
The input of 30-natural-sounding; The automatic identification unit of 40-voice;
50-natural language understanding unit; 60-action and response unit;
The 70-remote data base; 80-figure and literal display interface;
90-voice output interface; 402-natural-sounding loader;
404-phonetic feature acquisition device; 406-speech recognition device;
408-language construction database; 410-speech model database;
502-syntax analysis device; 504-key word analysis device;
506-semantic structure manager; 508-syntax database;
The 602-information manager; 604-natural language generator;
606-sound wave synthesizer.
Embodiment
Please refer to Fig. 1, Fig. 1 is the Organization Chart that shows hand-hold communication appliance and network among the disclosed embodiment.As shown in the figure, hand-hold communication appliance 100,102 has the wireless network communication ability, be connected by wireless network and internet 110, have the different server of function 104,106,108 on the internet 110, server 104,106,108 respectively has different Internet resources.Therefore, hand-hold communication appliance 100,102 can be by the every resource on wireless network inquiry or the use server 104,106,108.
Please refer to Fig. 2, Fig. 2 is the functional schematic that shows hand-hold communication appliance among the disclosed embodiment.As shown in the figure, hand-hold communication appliance 200 needs carry out communication by radio network interface 209 and wireless network 210, perhaps obtains every resource on the wireless network 210 by radio network interface 209.Hand-hold communication appliance 200 comprises display unit 202, CPU 204, storage device 206 and I/O device 208.Display unit 202 is selected for the user in order to type options perhaps to be provided in the display text.CPU 204 is in order to the computing speech data, and control display unit 202, storage device 206 and I/O device 208.Storage device 206, then is connected with it by wireless network 210 by CPU 204 when being large database as if required database in order to store voice deal with data or database.I/O device 208 is the voice output incoming interfaces for the user, and the user can import voice by I/O device 208, and hand-hold communication appliance 200 is also by I/O device 208 output voice.
Please refer to Fig. 3, Fig. 3 shows functional block diagram of the present invention.As shown in the figure, a kind of device with hand-hold communication appliance computing natural language, in order to the phonetic entry of reception nature, and, comprise the automatic identification unit 40 of voice, natural language understanding unit 50 and action and response unit 60 with spreading out of result's response after the computing of natural-sounding input process.
The automatic identification unit 40 of voice in order to reception nature phonetic entry 30, and is imported 30 with natural-sounding and is carried out feature acquisition and identification, produces the automatic identification result of voice.The automatic identification unit 40 of voice also comprises natural-sounding loader 402, phonetic feature acquisition device 404 and speech recognition device 406.Natural-sounding input 30 can be made a general reference the voice that general user is imported in the natural language expressing mode.
Natural-sounding loader 402 is to be user's interface, in order to receive nature phonetic entry 30.Phonetic feature acquisition device 404 is coupled to natural-sounding loader 402, in order to the phonetic feature of acquisition from the natural-sounding input of natural-sounding loader 402.Speech recognition device 406, be coupled to phonetic feature acquisition device 404, in order to be reference language structural database 408 and speech model database 410, the phonetic feature of the natural-sounding input that identification phonetic feature acquisition device 404 is captured, and produce the automatic identification result of voice.
Natural language understanding unit 50 is coupled to the automatic identification unit 40 of voice, in order to the automatic identification result of reception voice, and with voice automatic identification result process understanding and analysis, produces the natural language understanding result.Natural language understanding unit 50 also comprises syntax analysis device 502, key word analysis device 504 and semantic structure manager 506.
Action and response unit 60 are coupled to natural language understanding unit 50, in order to receiving the natural language understanding result, and the natural language understanding result are suitably handled, and response bears results.Action also comprises information manager 602, natural language generator 604 and sound wave synthesizer 606 with response unit 60.
Action also links with remote data base 70, figure and literal display interface 80 and voice output interface 90 with response unit 60, information manager 602 in action and the response unit 60 is during deal with data, when finding out required semantic structure is when needing the inquiry remote data base, just can link with remote data base 70, to obtain required data.After information manager 602 is found out required semantic structure, if need not be converted to voice output, but with alternate manner display result such as literal, figure or music, then can be by figure and literal display interface 80 displaying contents.
If information manager 602 is according to the natural language understanding result, the semantic structure of being found out need be converted to voice output, then be sent to natural language generator 604, to produce the semantic structure of natural language form, the natural language of natural language generator 604 being formed by sound wave synthesizer 606 again, synthetic sound wave, response bears results.Sound wave synthesizer 606 is connected with voice output interface 90, result's response that sound wave synthesizer 606 can utilize 90 outputs of voice output interface to be produced.But the natural language that natural language generator 604 is produced is literalization also, directly by figure and 80 outputs of literal display interface.
Moreover the present invention proposes a kind of method with hand-hold communication appliance computing natural language, in order to the phonetic entry of reception nature, and will spread out of result's response after the computing of natural-sounding input process.At first, receive nature phonetic entry (step S400), the voice that the general user of natural-sounding input general reference is imported in the natural language expressing mode.Then, reference language structural database and speech model database carry out feature acquisition and identification to natural-sounding input, and produce the automatic identification results of voice (step S402).
Then, the automatic identification result of voice through understanding and analyzing, is promptly analyzed the syntax and the keyword of the automatic identification result of voice, syntax analysis is with reference to grammatical database, produces natural language understanding result (step S404) according to analyzing.
At last, the natural language understanding result is handled (step S406), find out required semantic structure,, form the form of natural language, again natural language is synthesized sound wave with the response (step S408) that bears results according to the semantic structure of being found out.
For example, referring again to Fig. 3, natural-sounding as user's input is " I meet sb. at the airport 8 promptings tonight to the airport ", sound wave is through natural-sounding loader 402, for example behind the microphone module, the sound wave of input is converted into numeral sample (samples), the numeral sample of some constitutes framework (frame), these frameworks that partly overlap are passed through phonetic feature acquisition device 404 seriatim, to capture the characteristic parameter of sound wave, compare via speech recognition device 406 reference language structural databases 408 and speech model database 410 then, the sentence of finding out tool possibility is the automatic identification result of voice.
The automatic identification result of voice then enters natural language understanding unit 50 to understand and to analyze.At first, syntax analysis device 502 is analyzed according to the syntax of 508 pairs of automatic identification results of voice of grammatical database.The syntax in the syntax database 508 can be write definition in advance and be finished, as shown in Figure 5.Syntax analysis device 502 is parsed into structurized parsing tree (parsing tree) with sentence, as shown in Figure 6.If syntax analysis device 502 can successfully be parsed into structurized parsing tree with sentence, then semantic structure manager 506 can utilize structurized parsing tree, and this structurized parsing tree is expressed as structurized meaning of one's words framework (semantic frame).If syntax analysis device 502 can't successfully be parsed into structurized parsing tree with sentence, then utilize key word analysis device 504 that the keyword in the sentence is found out, utilize the antistop list of being found out to be shown meaning of one's words framework by semantic structure manager 506 again, as shown in Figure 7.Foregoing meaning of one's words framework is natural language understanding unit 50 through after understanding and analyzing, the natural language understanding result who is produced.
The natural language understanding result enters action and response unit 60 immediately, at first deliver to information manager 602, information manager 602 can assert that natural language understanding result as shown in Figure 7 belongs to prompting (remind), information manager 602 just can write down time and the content that need remind, as shown in Figure 8.When the time of required prompting arrives, information manager 602 can be in the content of figure and literal display interface 80 display alarms, maybe the content of reminding is delivered to natural language generator 604 and sound wave synthesizer 606, with synthetic result's response, this synthetic result should be " tonight I will arrive the airport meet sb. at the airport " at 8, plays results' response by voice output interface 90 at last.
If the natural language understanding result that information manager 602 is assert belongs to inquiry, for example, if the natural-sounding of user's input is " raining tomorrow in the Taibei ", sound wave is converted to numeral sample through natural-sounding loader 402, pass through phonetic feature acquisition device 404 again, to capture the characteristic parameter of sound wave, compare via speech recognition device 406 reference language structural databases 408 and speech model database 410 then, the sentence of finding out tool possibility is the automatic identification result of voice.
The automatic identification result of voice then enters natural language understanding unit 50 to understand and to analyze.Syntax analysis device 502 is analyzed according to the syntax of 508 pairs of automatic identification results of voice of grammatical database, produces structurized parsing tree as shown in Figure 9.Utilize structurized parsing tree by semantic structure manager 506 again, produce meaning of one's words framework, i.e. natural language understanding result as shown in figure 10.
Comprehensive speech, disclosed device and method, by the automatic identification unit of voice, natural language understanding unit and action and response unit, receive the voice that general user is imported in the natural language expressing mode, and, reach the purpose that the present invention desires to reach with spreading out of result's response after the computing of natural-sounding input process.Wherein, especially so that the natural language understanding unit is integrated in the single hand-hold communication appliance, in the voice processing technology of existing hand-hold communication appliance, real is outstanding integration mode, and has the quite distinguished effect of improving in natural language processing.
Claims (16)
1. device with hand-hold communication appliance computing natural language is used to receive natural-sounding input in the hand-held communication apparatus, and above-mentioned natural-sounding is input in the above-mentioned hand-hold communication appliance through spreading out of result response after the computing, comprising:
The automatic identification unit of one voice, it places above-mentioned hand-hold communication appliance, in order to receiving the input of above-mentioned natural-sounding, and feature acquisition and identification is carried out in above-mentioned natural-sounding input, produces the automatic identification result of voice;
The natural language understanding unit of forming by syntax analysis device, key word analysis device, semantic structure manager, it places above-mentioned hand-hold communication appliance, and be coupled to the automatic identification unit of above-mentioned voice, in order to receive the automatic identification result of above-mentioned voice, and, produce a natural language understanding result with automatic identification result process understanding of above-mentioned voice and analysis; And
One action and response unit, it places above-mentioned hand-hold communication appliance, and is coupled to above-mentioned natural language understanding unit, in order to receiving above-mentioned natural language understanding result, and above-mentioned natural language understanding result is handled, and produces the The above results response.
2. the device with hand-hold communication appliance computing natural language according to claim 1 is characterized in that: also comprise a radio network interface, it places above-mentioned hand-hold communication appliance, in order to the wireless network link communication.
3. the device with hand-hold communication appliance computing natural language according to claim 1 is characterized in that: the automatic identification unit of above-mentioned voice also comprises:
One natural-sounding loader is to be user's interface, in order to receive above-mentioned natural-sounding input;
One phonetic feature acquisition device, it is coupled to above-mentioned natural-sounding loader, in order to the phonetic feature of acquisition from the above-mentioned natural-sounding input of above-mentioned natural-sounding loader; And
One speech recognition device, it is coupled to above-mentioned phonetic feature acquisition device, the phonetic feature of the above-mentioned natural-sounding input that is captured in order to the above-mentioned phonetic feature acquisition device of identification, and produce the automatic identification result of above-mentioned voice.
4. the device with hand-hold communication appliance computing natural language according to claim 3, it is characterized in that: the phonetic feature that the above-mentioned natural-sounding that above-mentioned speech recognition device is captured in the above-mentioned phonetic feature acquisition device of identification is imported, and when producing the automatic identification result of above-mentioned voice, be with reference to a language construction database and a speech model database.
5. the device with hand-hold communication appliance computing natural language according to claim 1 is characterized in that: above-mentioned natural language understanding unit also comprises:
One syntax analysis device in order to receiving the automatic identification result of above-mentioned voice, and is analyzed the syntax of the automatic identification result of above-mentioned voice;
One key word analysis device, it is coupled to above-mentioned syntax analysis device, in order to receiving the automatic identification result of above-mentioned voice, and the keyword of the automatic identification result of above-mentioned voice is analyzed; And
One semantic structure manager, it is coupled to above-mentioned syntax analysis device and above-mentioned key word analysis device, in order to above-mentioned syntax analysis device of reference and above-mentioned key word analysis device produce above-mentioned natural language understanding result for the analysis of the automatic identification result of above-mentioned voice simultaneously.
6. the device with hand-hold communication appliance computing natural language according to claim 5 is characterized in that: when above-mentioned syntax analysis device is analyzed the syntax of the automatic identification result of above-mentioned voice, be with reference to a grammatical database.
7. the device with hand-hold communication appliance computing natural language according to claim 1 is characterized in that: above-mentioned action and response unit also comprise:
One information manager in order to receiving above-mentioned natural language understanding result, and according to above-mentioned natural language understanding result, is found out required semantic structure;
One natural language generator, it is coupled to above-mentioned information manager, in order to the semantic structure of finding out according to above-mentioned information manager, forms the form of natural language; And
One sound wave synthesizer, it is coupled to above-mentioned natural language generator, in order to the natural language that above-mentioned natural language generator is formed, synthetic sound wave, and produce the The above results response.
8. the device with hand-hold communication appliance computing natural language according to claim 1, wherein above-mentioned natural-sounding input is meant the voice that general user is imported in the natural language expressing mode.
9. method with hand-hold communication appliance computing natural language, be used to receive natural-sounding input in the hand-held communication apparatus, and above-mentioned natural-sounding is input in the above-mentioned hand-hold communication appliance through spreading out of result response after the computing, may further comprise the steps:
Above-mentioned hand-hold communication appliance receives above-mentioned natural-sounding input, and feature acquisition and identification are carried out in above-mentioned natural-sounding input, produces the automatic identification result of voice;
Above-mentioned hand-hold communication appliance is debated the syntax of knowing the result automatically to above-mentioned voice and is analyzed, and the keyword of knowing the result is debated automatically in above-mentioned voice analyze, and debate the analysis of knowing the result automatically according to above-mentioned voice, produces above-mentioned natural language understanding result; And
Above-mentioned hand-hold communication appliance is handled above-mentioned natural language understanding result, produces the The above results response.
10. the method with hand-hold communication appliance computing natural language according to claim 9, above-mentioned hand-hold communication appliance also comprises a radio network interface, above-mentioned radio network interface in order to the wireless network link communication.
11. the method with hand-hold communication appliance computing natural language according to claim 9, wherein, the step of the automatic identification result of the above-mentioned voice of above-mentioned generation, further comprising the steps of:
Receive above-mentioned natural-sounding input;
Capture the phonetic feature of above-mentioned natural-sounding input; And
The phonetic feature of the above-mentioned natural-sounding input that identification captured, and produce the automatic identification result of above-mentioned voice.
12. the method with hand-hold communication appliance computing natural language according to claim 11, wherein, in the step of the phonetic feature of the above-mentioned natural-sounding input that above-mentioned identification captured, above-mentioned identification is with reference to a language construction database and a speech model database.
13. the method with hand-hold communication appliance computing natural language according to claim 9, wherein, the above-mentioned natural language understanding result's of above-mentioned generation step, further comprising the steps of:
The syntax to the automatic identification result of above-mentioned voice are analyzed;
Keyword to the automatic identification result of above-mentioned voice is analyzed; And
According to the analysis of the automatic identification result of above-mentioned voice, produce above-mentioned natural language understanding result.
14. the method with hand-hold communication appliance computing natural language according to claim 13, wherein, in the above-mentioned step that the syntax of the automatic identification result of above-mentioned voice are analyzed, above-mentioned analysis is with reference to a grammatical database.
15. the method with hand-hold communication appliance computing natural language according to claim 9, wherein, the step that above-mentioned generation The above results is responded is further comprising the steps of:
According to above-mentioned natural language understanding result, find out required semantic structure;
According to the above-mentioned semantic structure of finding out, form the form of natural language; And
With the above-mentioned natural language of forming, synthetic sound wave, and produce the The above results response.
16. the method with hand-hold communication appliance computing natural language according to claim 9, wherein, above-mentioned natural-sounding input is meant the voice that general user is imported in the natural language expressing mode.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB031020682A CN1271793C (en) | 2003-01-29 | 2003-01-29 | Apparatus and method for computation process of natural language using hand held communication equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB031020682A CN1271793C (en) | 2003-01-29 | 2003-01-29 | Apparatus and method for computation process of natural language using hand held communication equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1521953A CN1521953A (en) | 2004-08-18 |
| CN1271793C true CN1271793C (en) | 2006-08-23 |
Family
ID=34281583
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB031020682A Expired - Fee Related CN1271793C (en) | 2003-01-29 | 2003-01-29 | Apparatus and method for computation process of natural language using hand held communication equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1271793C (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1801323B (en) * | 2005-01-06 | 2011-03-16 | 台达电子工业股份有限公司 | Load optimization method for speech/speaker recognition system |
| CN102629246B (en) * | 2012-02-10 | 2017-06-27 | 百纳（武汉）信息技术有限公司 | Recognize the server and browser voice command identification method of browser voice command |
| CN106782521A (en) * | 2017-03-22 | 2017-05-31 | 海南职业技术学院 | A kind of speech recognition system |
-
2003
- 2003-01-29 CN CNB031020682A patent/CN1271793C/en not_active Expired - Fee Related
Also Published As
| Publication number | Publication date |
|---|---|
| CN1521953A (en) | 2004-08-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1788266A (en) | Translation system | |
| CN1160700C (en) | System and method for providing network coordinated conversational services | |
| KR101027548B1 (en) | Voice browser dialog enabler for a communication system | |
| CN106409283B (en) | Man-machine mixed interaction system and method based on audio | |
| CN101576901B (en) | Method for generating search request and mobile communication equipment | |
| CN1770770A (en) | Method and system of enabling intelligent and lightweight speech to text transcription through distributed environment | |
| CN1639707A (en) | System and method for concurrent multimodal communication | |
| US20060235694A1 (en) | Integrating conversational speech into Web browsers | |
| WO2007008798A3 (en) | System and method for searching for network-based content in a multi-modal system using spoken keywords | |
| CN1752975A (en) | Method and system for voice-enabled autofill | |
| CN1711586A (en) | Speech recognition dictionary creation device and speech recognition device | |
| CN1826632A (en) | System and method for combined frequency-domain and time-domain pitch extraction for speech signals | |
| CN1538383A (en) | Distributed speech recognition for mobile communication devices | |
| CN1408111A (en) | Method and apparatus for processing input speech signal during presentation output audio signal | |
| CN1639681A (en) | System and method for concurrent multimodal communication using concurrent multimodal tags | |
| JPH10133847A (en) | Mobile terminal system for voice recognition, database search, and resource access communications | |
| CN110992955A (en) | Voice operation method, device, equipment and storage medium of intelligent equipment | |
| CN106504748A (en) | A kind of sound control method and device | |
| CN1829254A (en) | Mobile terminal and text-to-speech method of same | |
| CN103076893A (en) | Method and equipment for realizing voice input | |
| CN1271793C (en) | Apparatus and method for computation process of natural language using hand held communication equipment | |
| CN1427394A (en) | Speech sound browsing network | |
| US20040143436A1 (en) | Apparatus and method of processing natural language speech data | |
| CN1560834A (en) | Method of grammar describing and identification analyse of colloquial identification understanding | |
| CN102957711A (en) | Method and system for realizing website address location on television set by voice |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20060823 Termination date: 20180129 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |