CN108021559A - Natural language understanding system and lexical analysis method - Google Patents
Natural language understanding system and lexical analysis method Download PDFInfo
- Publication number
- CN108021559A CN108021559A CN201810112700.8A CN201810112700A CN108021559A CN 108021559 A CN108021559 A CN 108021559A CN 201810112700 A CN201810112700 A CN 201810112700A CN 108021559 A CN108021559 A CN 108021559A
- Authority
- CN
- China
- Prior art keywords
- module
- lexical analysis
- expression formula
- natural language
- language understanding
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/253—Grammatical analysis; Style critique
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Machine Translation (AREA)
- Devices For Executing Special Programs (AREA)
Abstract
The present invention proposes that a kind of natural language understanding system includes lexical analysis module and processor.Lexical analysis module includes expression formula module.Expression formula module includes common segment and non-universal part, wherein non-universal part includes parameter list.Processor is to perform lexical analysis module, so that the common segment of lexical analysis module combination expression formula module and non-universal part.Lexical analysis module provides multiple actual parameters to the common segment of expression formula module according to the parameter list of the non-universal part of expression formula module, to produce syntax rule data.In addition, a kind of lexical analysis method is also suggested.
Description
Technical field
The present invention relates to a kind of natural language understanding technology, and more particularly to a kind of natural language understanding system and the meaning of one's words
Analysis method.
Background technology
In the application field of intelligent sound identification, natural language understanding (Natural Language
Understanding, NLU) technology is important key technology at present.Natural language understanding technology can be responsible for extracting user's sentence
In key message, and the intention of user can be judged, to perform subsequent treatment corresponding to the intention of user.However, based on
In the practical application of the lexical analysis method of rule, natural language understanding system needs to separately design correspondence for substantial amounts of language material
Syntax rule.But since there are many similar word structures in natural language so that in corresponding syntax rule
There are the content of many redundancies, causes increase to compile the workload of syntax rule, also increases the difficulty of maintenance and modification.There is mirror
In this, the solution of several exemplary embodiments will be presented below.
The content of the invention
The present invention provides a kind of natural language understanding system and lexical analysis method, can be produced according to expression formula module
More syntax rule data, effectively to save the workload of natural language understanding system compiling syntax rule.
The natural language understanding system of the present invention includes lexical analysis module and processor.Lexical analysis module includes table
Up to formula module.Expression formula module includes common segment and non-universal part.Non-universal part includes parameter list.Processor is used
To perform lexical analysis module, so that the common segment of lexical analysis module combination expression formula module and non-universal part.Language
Meaning analysis module according to expression formula module non-universal part parameter list come provide multiple actual parameters at least within it
One to expression formula module common segment, to produce an at least syntax rule data.
A kind of lexical analysis method of the present invention is suitable for natural language understanding system.Natural language understanding system includes language
Meaning analysis module.Lexical analysis method comprises the following steps.Lexical analysis module is performed, so that lexical analysis module combines expression
The common segment of formula module and non-universal part, wherein non-universal part includes parameter list.According to the non-of expression formula module
The parameter list of common segment come provide multiple actual parameters at least one to expression formula module common segment, with production
A raw at least syntax rule data.Corpus data is compared via an at least syntax rule data by lexical analysis module, with
Obtain the semantics information corresponding to corpus data.
It can be compiled based on above-mentioned, of the invention natural language understanding system and lexical analysis method in expression formula module
Common segment and the form of non-universal part, so as to be produced by combining the common segment of expression formula module with non-universal part
Raw more syntax rule data.Therefore, natural language understanding system of the invention and lexical analysis method can be saved effectively certainly
The workload of right language understanding system compiling syntax rule, and can be easy to safeguard.
To make the foregoing features and advantages of the present invention clearer and more comprehensible, special embodiment below, and it is detailed to coordinate attached drawing to make
Carefully it is described as follows.
Brief description of the drawings
Fig. 1 illustrates the schematic diagram of the natural language understanding system of one embodiment of the invention.
Fig. 2 illustrates the schematic diagram of the lexical analysis module of one embodiment of the invention.
Fig. 3 illustrates the flow chart of the lexical analysis method of one embodiment of the invention.
Embodiment
, below can actually evidence as the present invention especially exemplified by embodiment in order to allow present disclosure easily to be understood
With the example of implementation.In addition, in place of all possibility, element/component/step of identical label is used in drawings and the embodiments,
Represent same or like component.
Fig. 1 illustrates the schematic diagram of the natural language understanding system of one embodiment of the invention.In the present embodiment, natural language
Speech understand (Natural Language Understanding, NLU) system 100 include processor 110, memory 120 and
Input unit 130.Memory 120 stores lexical analysis module 121.Processor 110 couples memory 120 and input unit
130.In the present embodiment, lexical analysis module 121 can be stored in advance in another storage device, when processor 110 is held
During row lexical analysis module 121, processor 110 reads this another storage device, and lexical analysis module 121 is loaded into storage
Device 120.This another storage device can be for example hard disk (Hard Disk Drive, HDD), solid state hard disc (Solid State
Disk, SSD) or the suchlike device such as CD (Optical Disc, OD), the present invention is not limited thereto.In this implementation
In example, processor 110 performs lexical analysis module 121 to produce syntax rule data.
In the present embodiment, processor 110 be, for example, central processing unit (Central Processing Unit,
CPU), the microprocessor of system single chip (System on Chip, SOC) or other programmable general services or specific use
Device (Microprocessor), digital signal processor (Digital Signal Processor, DSP), programmable controller,
Special application integrated circuit (Application Specific Integrated Circuits, ASIC), programmable logic dress
Put the combination of (Programmable Logic Device, PLD), other similar process devices or these devices.It is noticeable
It is that in the present embodiment, processor 110 includes compiler (Compiler), and wherein compiler can be used to perform each reality of the present invention
Apply the various natural languages and program compiling work.
In the present embodiment, memory 120 is, for example, dynamic random access memory (Dynamic Random Access
Memory, DRAM), flash memory (Flash memory) or non-volatile random access memory (Non-Volatile Random
Access Memory, NVRAM) etc..In the present embodiment, memory 120 is storing the number described in various embodiments of the present invention
According to, parameter and program module, and processor 110 can perform these data, parameter and program module, to realize the present invention
System and method described in each embodiment.
In the present embodiment, input unit 130 can be used to receive the order parameter of user's output.Input unit 130 is for example
It is the group of keyboard (Keyboard), voice receiver (Voice receiver), microphone (Microphone) or these devices
Close etc..Input unit 130 can be used to receive voice messaging or order parameter of user's offer etc., and the present invention is not limited thereto.
In the present embodiment, input unit 130 can provide data or parameter to processor according to the input mode of above-mentioned various types
110。
Fig. 2 illustrates the schematic diagram of the lexical analysis module of one embodiment of the invention.With reference to figure 1 and Fig. 2.In this implementation
In example, lexical analysis module 121 includes expression formula module 122.In the present embodiment, lexical analysis module 121 is according to context
Grammars (Context free Grammars, CFGs) carry out construction expression formula module 122.Also, lexical analysis module 121
The language of expression formula module 122 is formulated to extend Ba Kesi-Nao that normal forms (Extended Backus-Naur Form, EBNF)
Method rule.In the present embodiment, table of the compiling in advance of expression formula module 122 with common segment 123 and non-universal part 124
Up to formula module 122.In the present embodiment, when processor 110 performs lexical analysis module 121, lexical analysis module 121 combines
The common segment 123 of expression formula module 122 and non-universal part 124, to produce multiple syntax rule data 130_1,130_2
~130_N, wherein N are the positive integer more than 1.
Specifically, in the present embodiment, the common segment 123 of expression formula module 122 includes grammar templates, and this language
Method template includes the production module of parametrization.Non-universal part 124 includes parameter list, and parameter list includes multiple realities
Border parameter (actual parameter).In the present embodiment, when processor 110 performs lexical analysis module 121 with reference to table
Up to formula module 122 common segment 123 and non-universal part 124 when, lexical analysis module 121 is by this among parameter list
A little actual parameters are transferred in the production module of parametrization, to produce syntax rule data 130_1,130_2~130_N.
In the present embodiment, the production module of the parametrization of grammar templates may include formal parameter (formal
parameter).That is, when lexical analysis module 121 to combine the common segment 123 of expression formula module 122 and non-through
During with part 124, these actual parameters among parameter list can be transferred to the production mould form ginseng in the block of parametrization
Number.For example, the expression formula content of expression formula module 122 can be such as following grammer.<Folded word (verb)>:$ (verb) $
(verb);<Do something (verb, predicate)>:I thinks<Folded word ($ (verb))><recent>$(predicate);<
recent>:Nearest | recent | new;<Grammer 1>:<Do something (seeing, film)>;<Grammer 2>:<Something is done (to listen, prevalence
Song)>.
In above-mentioned grammer, "<Folded word (verb)>:$ (verb) $ (verb) " and "<Do something (verb,
predicate)>:I thinks<Folded word ($ (verb))><recent>$ (predicate) " is the grammar templates of common segment 123
The production of parametrization.“<Folded word (verb)>" there is a formal parameter " (verb) ".“<Do something (verb,
predicate)>" there is two formal parameters " (verb) " and " (predicate) ".In above-mentioned grammer, "<Grammer 1>”
And "<Grammer 2>" be expression formula module 122 non-universal part 124, and corresponding to it is above-mentioned "<Do something (verb,
predicate)>”.“<Grammer 1>" there is two actual parameters " seeing " and " film ".“<Grammer 2>" there are two actual ginsengs
Number " listening " and " popular song ".In this example, "<Grammer 1>" and "<Grammer 2>" there is verb (verb) reality respectively
Parameter and predicate (predicate) actual parameter.
That is, lexical analysis module 121 can produce two according to the expression formula content of above-mentioned expression formula module 122
Syntax rule.According to "<Grammer 1>", one of them of the two syntax rules is " I wants to have a look at<recent>Film;<
recent>:Nearest | recent | new ".According to "<Grammer 2>", the other in which of the two syntax rules is " I wants to listen
Just listen<recent>Popular song;<recent>:Nearest | recent | new ".In other words, 121 nothing of lexical analysis module
Palpus additional compilation corresponds to independent two expression formula of two syntax rules.In this example, 121 need of lexical analysis module
One expression formula module 122 with common segment 123 of compiling, with by the parameter of the grammar templates among common segment 123
The production of change, you can obtain above-mentioned two syntax rule.Therefore, the natural language understanding system 100 of the present embodiment can be effective
Save the workload of compiling syntax rule.
In one embodiment, the production module of the parametrization of grammar templates may also comprise location parameter (positional
parameter).That is, when lexical analysis module 121 to combine the common segment 123 of expression formula module 122 and non-through
During with part 124, these actual parameters among parameter list can be transferred to the production mould position ginseng in the block of parametrization
Number.For example, above-mentioned "<Folded word>" and "<Do something>" for expression formula module 122 common segment 123 grammer mould
The production of the parametrization of plate.In this example, "<Folded word>" syntactic representation formula can be "<Folded word>:_ 1 one _ 1 ", and "<
Do something>" syntactic representation formula can be "<Do something>:I thinks<Folded word><recent>_ 2 ", wherein " _ 1 " and " _ 2 " can divide
First parameter and second parameter in corresponding parameter list.Similarly in above-mentioned formal parameter, lexical analysis module 121
An expression formula module 122 with common segment 123 need to be compiled, with by the ginseng of the grammar templates among common segment 123
The production of numberization, you can obtain above-mentioned two syntax rule.
Referring again to Fig. 1 and Fig. 2, in the present embodiment, natural language understanding system 100 can connect by input unit 130
Corpus data input by user is received, and compares corpus data, to obtain the semantics information corresponding to corpus data.In this implementation
In example, lexical analysis module 121 can be according to the common segment 123 of above-described embodiment combination expression formula module 122 and non-universal
The mode of part 124 produces these syntax rule data 130_1,130_2~130_N.In the present embodiment, lexical analysis mould
The mode that block 121 produces these syntax rule data 130_1,130_2~130_N can be as indicated in the pretreatment rank in C language
Grand (macro) expansion mode that section is carried out.The deployable above-mentioned grammar templates of processor 110, and record to memory 120
In, to produce these syntax rule data 130_1,130_2~130_N at the same time via expression formula module 122.That is, deposit
Reservoir 120 can record these syntax rule data 130_ of the expression formula module 122 after expansion with more storage space
1st, 130_2~130_N, and compare the language material that user provides by these syntax rule data 130_1,130_2~130_N
Data, to obtain the semantics information corresponding to corpus data.Therefore, when natural language understanding system 100 carries out natural language reason
During solution operation, natural language understanding system 100 has efficient comparison effect.
However, in one embodiment, lexical analysis module 121 produce these syntax rule data 130_1,130_2~
The mode of 130_N also can be as indicated in function call (function call) mode in C language.That is, processor
110 retain above-mentioned grammar templates in memory 120, and these actual parameters dynamic in non-universal part 124 is replaced
Shift in the production module of the parametrization of the grammar templates in common segment 123, to be produced one by one via expression formula module 122
These raw syntax rule data 130_1,130_2~130_N.In other words, in these actual parameter dynamic replacements to parametrization
Production module during, often replace and be once then compared immediately.Therefore, when natural language understanding system 100 carries out
When natural language understanding operates, natural language understanding system 100 can effectively save 120 space of memory.
Fig. 3 illustrates the flow chart of the lexical analysis method of one embodiment of the invention.Referring to figs. 1 to Fig. 3, the present embodiment
Lexical analysis method can at least be suitable for the natural language understanding system 100 and lexical analysis module of Fig. 1 and Fig. 3 embodiments
121.In step S310, natural language understanding system 100 performs lexical analysis module 121, so that 121 knot of lexical analysis module
Common segment 123 and the non-universal part 124 of expression formula module 122 are closed, wherein non-universal part 124 includes parameter list.
In step s 320, natural language understanding system 100 is come according to the parameter list of the non-universal part 124 of expression formula module 122
Multiple actual parameters are provided to the common segment 123 of expression formula module 122, with produce syntax rule data 130_1,130_2~
130_N.In step S330, natural language understanding system 100 is by lexical analysis module 121 via syntax rule data 130_
1st, 130_2~130_N compares corpus data, to obtain the semantics information corresponding to corpus data.Therefore, the present embodiment meaning of one's words
Analysis method can effectively save the workload that natural language understanding system 100 compiles syntax rule.
In addition, the relevant apparatus feature and technology contents on the natural language understanding system 100 of the present embodiment can be according to
Enough teachings, suggestion are obtained according to the content of above-mentioned Fig. 1 and Fig. 2 embodiments and implements explanation, therefore is repeated no more.
In conclusion the natural language understanding system and lexical analysis method of the present invention can be in the general of expression formula module
In part, grammar templates are compiled in advance.When natural language understanding system executable expressions module, natural language understanding system can
Multiple actual parameters of the parameter list of non-universal part are provided in the grammar templates of common segment, to produce more grammers
Regular data.That is, the natural language understanding system and lexical analysis method of the present invention need not compile largely in advance
Syntax rule data, you can produce more syntax rule data according to grammar templates.Therefore, natural language understanding system of the invention
System and lexical analysis method can effectively save the workload of natural language understanding system compiling syntax rule, and can be easy to dimension
Shield.
Although the present invention is disclosed as above with embodiment, so it is not limited to the present invention, any people in the art
Member, without departing from the spirit and scope of the present invention, when can make a little change and retouch, therefore protection scope of the present invention is when regarding
Subject to appended claims institute defender.
【Symbol description】
100:Electronic device
110:Processor
120:Memory
121:Lexical analysis module
122:Expression formula module
123:Common segment
124:Non-universal part
130_1、130_2、130_N:Syntax rule data
S310、S320、S330:Step
Claims (15)
1. a kind of natural language understanding system, including:
Lexical analysis module, including expression formula module, and the expression formula module includes common segment and non-universal part,
Wherein described non-universal part includes parameter list;And
Processor, to perform the lexical analysis module, so that the lexical analysis module is with reference to the expression formula module
The common segment and the non-universal part, and the lexical analysis module is according to the described non-of the expression formula module
The parameter list of common segment come provide multiple actual parameters at least one to described in the expression formula module
Common segment, to produce an at least syntax rule data.
2. natural language understanding system as claimed in claim 1, wherein the lexical analysis module is according to context-free text
Method carrys out expression formula module described in construction.
3. natural language understanding system as claimed in claim 1, wherein the common segment of the expression formula module includes
Grammar templates, and the grammar templates include the production module of parametrization, wherein when the processor performs the meaning of one's words
When analysis module comes the common segment and the non-universal part with reference to the expression formula module, the lexical analysis mould
At least one of the multiple actual parameter is transferred in the production module of the parametrization by block, with described in producing extremely
Few syntax rule data.
4. natural language understanding system as claimed in claim 3, wherein the production module of the parametrization is joined including form
Number.
5. natural language understanding system as claimed in claim 3, wherein the production module of the parametrization is joined including position
Number.
6. natural language understanding system as claimed in claim 3, further includes:
Memory, couples the processor, and when the processor performs the lexical analysis module, the processor will
The lexical analysis module is loaded into the memory,
Wherein described lexical analysis module produces multiple syntax rule data, and compares via the multiple syntax rule data
To corpus data, to obtain the semantics information corresponding to the corpus data.
7. natural language understanding system as claimed in claim 6, wherein the grammar templates are unfolded in the processor, and remembers
Record is into the memory, to produce multiple syntax rule data at the same time via the expression formula module.
8. natural language understanding system as claimed in claim 6, wherein the processor retains the parameter of the grammar templates
Message form is in the memory, and by the ginseng in the multiple actual parameter dynamic replacement to the grammar templates
In the production module of numberization, to produce multiple syntax rule data one by one via the expression formula module.
9. a kind of lexical analysis method, suitable for natural language understanding system, and the natural language understanding system includes language
Meaning analysis module, wherein the lexical analysis method includes:
The lexical analysis module is performed, so that the common segment of the lexical analysis module combination expression formula module and non-through
With part, wherein the non-universal part includes parameter list;
Multiple actual parameters are provided according to the parameter list of the non-universal part of the expression formula module at least
One of them to the expression formula module the common segment, to produce an at least syntax rule data;And
Corpus data is compared via an at least syntax rule data by the lexical analysis module, is corresponded to obtaining
The semantics information of the corpus data.
10. lexical analysis method as claimed in claim 9, wherein the lexical analysis module is come according to context-free grammar
Expression formula module described in construction.
11. lexical analysis method as claimed in claim 9, wherein the common segment of the expression formula module includes grammer
Template, and the grammar templates include the production module of parametrization, wherein the lexical analysis module is performed, so that described
Lexical analysis module with reference to the expression formula module the common segment and the non-universal part the step of include:
, will when the common segment and the non-universal part of the lexical analysis module with reference to the expression formula module
At least one of the multiple actual parameter is transferred in the production module of the parametrization, with least one described in generation
Syntax rule data.
12. lexical analysis method as claimed in claim 11, wherein the production module of the parametrization includes formal parameter.
13. lexical analysis method as claimed in claim 11, wherein the production module of the parametrization includes location parameter.
14. lexical analysis method as claimed in claim 11, wherein the natural language understanding system further includes memory, and
And include described in producing the step of an at least syntax rule data:
The grammar templates are unfolded, and record into the memory, it is multiple to be produced at the same time via the expression formula module
Syntax rule data.
15. lexical analysis method as claimed in claim 11, wherein the natural language understanding system further includes memory, and
And include described in producing the step of an at least syntax rule data:
Retain the parameter information form of the grammar templates in the memory, and the multiple actual parameter dynamic is replaced
Shift in the production module of the parametrization in the grammar templates, it is multiple to be produced one by one via the expression formula module
Syntax rule data.
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810112700.8A CN108021559B (en) | 2018-02-05 | 2018-02-05 | Natural language understanding system and semantic analysis method |
TW107111597A TWI673705B (en) | 2018-02-05 | 2018-04-02 | Natural language understanding system and semantic analysis method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810112700.8A CN108021559B (en) | 2018-02-05 | 2018-02-05 | Natural language understanding system and semantic analysis method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108021559A true CN108021559A (en) | 2018-05-11 |
CN108021559B CN108021559B (en) | 2022-05-03 |
Family
ID=62075106
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810112700.8A Active CN108021559B (en) | 2018-02-05 | 2018-02-05 | Natural language understanding system and semantic analysis method |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN108021559B (en) |
TW (1) | TWI673705B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109543192A (en) * | 2018-11-30 | 2019-03-29 | 北京羽扇智信息科技有限公司 | Natural language analytic method, device, equipment and storage medium |
Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6675370B1 (en) * | 2000-02-02 | 2004-01-06 | International Business Machines Corporation | System and method for imbedding hyperlinked language grammar notation in a “literate” programming environment |
CN101763211A (en) * | 2009-04-30 | 2010-06-30 | 广东国笔科技股份有限公司 | System for analyzing semanteme in real time and controlling related operation |
CN101814065A (en) * | 2009-02-23 | 2010-08-25 | 富士通株式会社 | Syntactic analysis device and syntactic analysis method |
CN101894236A (en) * | 2010-07-28 | 2010-11-24 | 北京华夏信安科技有限公司 | Software homology detection method and device based on abstract syntax tree and semantic matching |
CN102262622A (en) * | 2010-05-31 | 2011-11-30 | 国际商业机器公司 | Document processing, template generating and conceptbase generating methods and devices |
CN102609462A (en) * | 2012-01-14 | 2012-07-25 | 杭州安恒信息技术有限公司 | Method for compressed storage of massive SQL (structured query language) by means of extracting SQL models |
CN102768682A (en) * | 2012-06-28 | 2012-11-07 | 用友软件股份有限公司 | Business rule processing device and business rule processing method |
CN103294666A (en) * | 2013-05-28 | 2013-09-11 | 百度在线网络技术(北京)有限公司 | Grammar compiling method, semantic parsing method and corresponding devices |
US20150242396A1 (en) * | 2014-02-21 | 2015-08-27 | Jun-Huai Su | Translating method for translating a natural-language description into a computer-language description |
CN106020803A (en) * | 2016-05-11 | 2016-10-12 | 深圳市麦斯杰网络有限公司 | Method and device for dynamic definition of monitored object template |
CN107340999A (en) * | 2017-01-09 | 2017-11-10 | 北京理工大学 | Software automation method and system and the method in structure natural language understanding storehouse |
Family Cites Families (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
SG142196A1 (en) * | 2006-11-02 | 2008-05-28 | Crimsonlogic Pte Ltd | System and method for processing language independent natural language statements |
CN103871402B (en) * | 2012-12-11 | 2017-10-10 | 北京百度网讯科技有限公司 | Language model training system, speech recognition system and correlation method |
JP5809381B1 (en) * | 2014-04-29 | 2015-11-10 | 楽天株式会社 | Natural language processing system, natural language processing method, and natural language processing program |
CN106326303B (en) * | 2015-06-30 | 2019-09-13 | 芋头科技(杭州)有限公司 | A kind of spoken semantic analysis system and method |
CN107229616B (en) * | 2016-03-25 | 2020-10-16 | 阿里巴巴集团控股有限公司 | Language identification method, device and system |
-
2018
- 2018-02-05 CN CN201810112700.8A patent/CN108021559B/en active Active
- 2018-04-02 TW TW107111597A patent/TWI673705B/en active
Patent Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6675370B1 (en) * | 2000-02-02 | 2004-01-06 | International Business Machines Corporation | System and method for imbedding hyperlinked language grammar notation in a “literate” programming environment |
CN101814065A (en) * | 2009-02-23 | 2010-08-25 | 富士通株式会社 | Syntactic analysis device and syntactic analysis method |
CN101763211A (en) * | 2009-04-30 | 2010-06-30 | 广东国笔科技股份有限公司 | System for analyzing semanteme in real time and controlling related operation |
CN102262622A (en) * | 2010-05-31 | 2011-11-30 | 国际商业机器公司 | Document processing, template generating and conceptbase generating methods and devices |
CN101894236A (en) * | 2010-07-28 | 2010-11-24 | 北京华夏信安科技有限公司 | Software homology detection method and device based on abstract syntax tree and semantic matching |
CN102609462A (en) * | 2012-01-14 | 2012-07-25 | 杭州安恒信息技术有限公司 | Method for compressed storage of massive SQL (structured query language) by means of extracting SQL models |
CN102768682A (en) * | 2012-06-28 | 2012-11-07 | 用友软件股份有限公司 | Business rule processing device and business rule processing method |
CN103294666A (en) * | 2013-05-28 | 2013-09-11 | 百度在线网络技术(北京)有限公司 | Grammar compiling method, semantic parsing method and corresponding devices |
US20150242396A1 (en) * | 2014-02-21 | 2015-08-27 | Jun-Huai Su | Translating method for translating a natural-language description into a computer-language description |
CN106020803A (en) * | 2016-05-11 | 2016-10-12 | 深圳市麦斯杰网络有限公司 | Method and device for dynamic definition of monitored object template |
CN107340999A (en) * | 2017-01-09 | 2017-11-10 | 北京理工大学 | Software automation method and system and the method in structure natural language understanding storehouse |
Non-Patent Citations (1)
Title |
---|
周燕: "嵌入式操作系统内核模型的代码自动生成及规则优化", 《中国优秀硕士学位论文全文数据库信息科技辑(月刊)》 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109543192A (en) * | 2018-11-30 | 2019-03-29 | 北京羽扇智信息科技有限公司 | Natural language analytic method, device, equipment and storage medium |
CN109543192B (en) * | 2018-11-30 | 2023-05-05 | 出门问问创新科技有限公司 | Natural language analysis method, device, equipment and storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN108021559B (en) | 2022-05-03 |
TWI673705B (en) | 2019-10-01 |
TW201935463A (en) | 2019-09-01 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US20110137653A1 (en) | System and method for restricting large language models | |
JP4901155B2 (en) | Method, medium and system for generating a grammar suitable for use by a speech recognizer | |
US20070083369A1 (en) | Generating words and names using N-grams of phonemes | |
WO2021223287A1 (en) | Focalgan-based short text automatic generation method, apparatus, and device, and storage medium | |
CN110494841A (en) | Context language translation | |
CN106354716A (en) | Method and device for converting text | |
CN102236573A (en) | Name binding extensibility for typed programming language | |
CN114503073A (en) | Automatic conversion of program written in procedural programming language into dataflow graph and related systems and methods | |
CN108255956A (en) | The method and system of dictionary are adaptively obtained based on historical data and machine learning | |
CN108021559A (en) | Natural language understanding system and lexical analysis method | |
US20070055492A1 (en) | Configurable grammar templates | |
KR20140082711A (en) | Efficient incremental modification of optimized finite-state transducers(fsts) for use in speech applications | |
US20030196195A1 (en) | Parsing technique to respect textual language syntax and dialects dynamically | |
WO2017159207A1 (en) | Processing execution device, method for controlling processing execution device, and control program | |
JP6644141B2 (en) | Response device, control method of response device, and control program | |
de Brock | On system sequence descriptions | |
JP2020531953A5 (en) | ||
JP6619932B2 (en) | Morphological analyzer and program | |
CN109065016B (en) | Speech synthesis method, speech synthesis device, electronic equipment and non-transient computer storage medium | |
US20180011833A1 (en) | Syntax analyzing device, learning device, machine translation device and storage medium | |
US7617089B2 (en) | Method and apparatus for compiling two-level morphology rules | |
JP2010145873A (en) | Text replacement device, text voice synthesizer, text replacement method, and text replacement program | |
US20220328045A1 (en) | Electronic apparatus and method for controlling electronic apparatus thereof | |
JP2013171214A (en) | Information processor and program | |
JPH11344998A (en) | Method, device for setting reading and metrical information and storage medium in which reading and metrical information setting program is stored |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |