CA1235817A - Method and apparatus for natural language processing - Google Patents
Method and apparatus for natural language processingInfo
- Publication number
- CA1235817A CA1235817A CA000494041A CA494041A CA1235817A CA 1235817 A CA1235817 A CA 1235817A CA 000494041 A CA000494041 A CA 000494041A CA 494041 A CA494041 A CA 494041A CA 1235817 A CA1235817 A CA 1235817A
- Authority
- CA
- Canada
- Prior art keywords
- vocabularies
- dictionary
- semantic
- concept
- vocabulary
- 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
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/237—Lexical tools
- G06F40/242—Dictionaries
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/205—Parsing
- G06F40/211—Syntactic parsing, e.g. based on context-free grammar [CFG] or unification grammars
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/268—Morphological analysis
-
- 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
- G06F40/40—Processing or translation of natural language
- G06F40/42—Data-driven translation
- G06F40/47—Machine-assisted translation, e.g. using translation memory
Abstract
ABSTRACT OF THE DISCLOSURE
A natural language including Japanese, Korean, Chinese, etc., is suitably processed when an undefined word which is not registered in a dictionary appears in an input character train. When the undefined word appears, semantic analysis means decides a semantic concept of the undefined word by using a case dictionary, and further by using a vocabulary hierarchy dictionary which registers vocabularies in a hierarchy structure, the semantic analysis means extracts a group of vocabu-laries (synonyms) corresponding to a lower concept of the semantic concept of the undefined word. The natural language processing can be achieved practically and efficiently with the dictionaries of a limited storage capacity.
A natural language including Japanese, Korean, Chinese, etc., is suitably processed when an undefined word which is not registered in a dictionary appears in an input character train. When the undefined word appears, semantic analysis means decides a semantic concept of the undefined word by using a case dictionary, and further by using a vocabulary hierarchy dictionary which registers vocabularies in a hierarchy structure, the semantic analysis means extracts a group of vocabu-laries (synonyms) corresponding to a lower concept of the semantic concept of the undefined word. The natural language processing can be achieved practically and efficiently with the dictionaries of a limited storage capacity.
Description
~l~35~
1 BACKGROUND OE~ THE INVENTION
The present invention relates to a method and apparatus for processing a natural language such as Japanese, korean, chinese, etc., and in particular, to S a method and apparatus for natural language processing suitable for processing sentences which contain vocabu-laries not registered in a dictonary (hereinafter referred to as undefined words).
In performing sophisticated processing of a natural lanyuage, generally, syntax analysis of a sentence is carried out, and based on tha analysis, composition and editing o~ a new sentence are performed. However, in order to achieve the natural language processing, all vocabularies possible to compose a sentence are not always registered in a dictionary entirely, and thus even when a large dictionary is made available, undefined words which are not registered in the dictionary fre-quently appear in the sentence in many cases.
A technique to solve this problem is disclosed in Japanese patent unexamined publication Nos. 58-175074 (1983), 58-175075 (1983) and 58-175076. In the tech-nique disclosed in these publications, for an undefined word occurring as a result of failure in consulting a dictionary, the user is requested to input a synonym included in the words which have been registered and .,,` ~
~:35~3~7 l existing in the dictionary, and depending on this user's response, information for syntax analysis is added.
However~ in this technique, it is necessary for the user to memorize beforehand synonyms, etc., registered in the dictionary, or it is necessary for the system to output a list of all vocabularies regist~red in the dictionary 50 that the user himself extracts a synonym or the like from the list. Accordingly, in the case of a sophisticated processing, the number of registered words increases, and the burden on the user increases remarkably, and thus it has been difficult to carry out the natural language processing in a simple manner and effectively.
In view of the problems mentioned above, an object of the present invention is to provide a method and apparatus for natural language processing which is capable of performiny a sophisticated and practical na~ural language processing in that even when an un-defined word appears in a sentence to be processed, the user is pr~vided with information of a synonym from the apparatus to enable the processing with a dictionary of a limited storage capacityO
SUMMARY OF THE INVENTION
In the present invention, when an undefined word appears in an input character train of a natural language, an uppex concept of the undefined word is obtained by semantic analysis means which determines a 35~3~7 1 semantic concept of the vocabulary by using a case dictionary, and then by using a vocabulary hierarchy dictionary which registers vocabularies in a hierarchy stxucture formed according to the upper concept, a group of vocabularies corresponding to a lower concept are extracted. Further, it is also a featur~ of khe inven-tion in which the extracted group of vocabularies are displayed on a display device and the user is enabled to select a synonym from the group.
Another feature of the invention resides in that when an undefined word appears in an input character train of a natural language, an upper concept of the unde~ined word is obtained by semantic analysis means which determines a semantlc concept of a vocabulary by using a case dictionary, and then by using a vocabulary hierarchy dictionary which registers vocabularies in a hierarchy structure formed according to the upper concept, a group of vocabularies corresponding to a lower concept are extracted, and a synon~m is selected by inference means which selects from the extracted group of vocabularies of the lower concept a vocabulary by inference, of which vocabulary is in conformity with the intention of the inputted character train, and then the undefined word is replaced by the selected synonym.
BRIEF DESCRIPTION OF T~IE DRAWINGS
Fig. 1 is a block diagram of an overall arrange-a~7 1 ment in accordance with an embodiment of the present invention;
Figs. 2A to 2C show an example of an input sentence;
Fig. 3 is a flowchart for ~xplaining a proces-5 sing procedure of morphemic analysis means;
Fig. 4 is a flowchart for explaining a proces-sing procedure of syntax ana~ysis means, Fig. 5 shows an example of a vocabulary hierarchy dictionary;
Fig. 6 is a flowchart for explaining a proces-sing procedure of semantic analysis means;
Fig. 7 shows an example of a syntax tree;
Fig. 8 shows an example of display of a synonym;
Fig. 9 is a block diagram of an overall arrange-ment in accordance with another embodiment of the present invention;
Figs. lOA and lOB show respectively an example of a picture screen status of an information terminal and an example of representation of a knowledge base (status memory);
Fig~ 11 shows an example o~ a knowledge base (function memory); and Fig. 12 is a flowchart for explaining a processing procedure in accordance with another embodi-ment of the present invention.
-- 4 ~
i8~7 An embodiment of the present invention will be described with reference to the drawings.
Fig. 1 shows an overall arrangement of the invention, that is, shows an intelligent guidance syst~m of the question-answer type in which the user inputs by means of an input device 1 (e.g., a keyboard, voice recognition device, hand-wrikten character input device~
etc.) a natural language (e.g., Japanese) into a proces-sing unit 2, and the processing unit 2 executes processingas regards the natural language, and provides the user with inormation by displaying the contents and results of the processing on a display device 3. Here, the processing unit 2 is composed of a keyword dictionary 8, a con-junction relation5hip dictionary 9, an inflectional ending dictionary 10, and morphemi~ analysis means 4 for dividing an input sentence into vocabularies by use of in~rmation of these dictionaries 8 to 10; a grammatical rule dictionary 11, and syntax analysis means 5 for obtaining a syntax of the input sentence by use of the dictionary 11; a case dictionary 12, a vocabulary hierarchy dictionary 13, and semantic analysis means 5 for extracting a meaning of the input sentence by use of these dictionaries 12 and 13; a sentence pattern dictionary 14 containing several patterns of answer sentences, and sentence producing means 7 for producing a sentence by use of this dictionary 14.
Next, individual processings in the processing l unit 2 will be described in detail by way of an example in which as shown in Fig. 2A~ a sentence in kana charac-ters (Japanese syllabary) in solid writings:
(mi gi shi ta no ba sho ni ku u ha ku i ki wo !wish to draw up a blank area on the sa ku se i shi ta i) lower right place ]
is inputted as an input sentence.
~1) Morphemic Analysis First, in the morphenic analysis, by using the ke~word dictionary 8, aonjunction relationship dictionary 9, and inflectional ending dictionary 10, it is found whether or not a keyword exists in the character train o~ the input sentence, and a conjunction relationship is checked, and then the input sentence is divided into vocabularies. Here, in Tables l, 2, and 3, examples of the keyword dictionary 8, conjunction relationship dictionary 9, and inflectional ending dictionary lO are respectively shown.
~2~317 ,~, ____ ~ __ .,~ ~ ,~ ,~ ~ ~r er ,~ ,~ ,~ ~r . . .
'C1 ~ ~ ~n ~ c~ c~L ~ a:~ ~L C~L
o ~
; ~o . .
g ~
~ ~ t~ ~1 ~) el' el'r~ ~ r_l ~ ~ ~ ~
n ~ ~ ~ i~3 ~3 ~ ~ ~ ~
O O
_~ _ _ ~ ~ ~ a) Q) ~ ~I ~ 'U 'U 'U
~,~ ~ h ~rl aJ U 1: ~ ~ h ~ ~ ~i h . . .
U ~1 1~ ~ ~1~ O 1~ ~ (~ O O O ~d ,~ O ,~ ~ ~
~ r~ ~
~ ~ ,~ ~, .,., a) a) ~ ~ ~C Ul o~ U~
h 1~ S l ~ h ~ ~ 1~ td P~ ~h U ~-~ n~ U U U
~ - ----- --.----.
~l -~ ~
,_1 ~ ih5 ~< rl E~ ~ _ ~.~
~ ~Q~ _~ ~ ~
C~r~l r~ ~1 _ _ m~So ~ _ *~ _~
d ~ ~ ~ ~0~ ~ ~ ~ ~ 3 _ _ ~_ _ _ ~._ _ _ , _ __ _ __ _ _ ~a G) ~1 ~ ~ ~ u~ ~O 1` 00 a s~ O O O O O O O O O
O O O O O O O O O
~a o o o o o o o o o ~3~ 7 Table 2 Conjunction Relationship Dictionary ~ _ ___ _ _ beginning 1 2 3 4 5 6 sentence ~1 ~
O - ... ... ... -O l _ _ __ _ . _ ~ _ _. _ _ _ ending of sentence _ _ _ _ S8~L7 _ _ _ _ ~ o h O ~ ~ ~ . . .
~1 _, _ ,~ ~
:~ __--~
o u _ ~n _ ~ In ~ .,1 ~0 ca t~.C ...
s~
,~
~ .
h O ~ O
.,~
... .. .
1 In the keyword dictionary 8, there are registered with addresses, keywords, parts of spaech, and conjunction conditions, and in particular, in the conjunction conditions, there are registered with forward and backward conjunction relationships of each keyword.
Further, in the conjunction relationship dictionary 9, there are registered with correspondences between the conjunction relationships of each keyword mentioned above. For example, when a vocabulary having a forward conjunction condition of ~3 and another vocabulary having a backward conjunction condition of ~6 are pos~
sible to he connected with each other, then this xelation-ship is registered as "1". The inflectional ending dictionary 10 is us~d to register beforehand the endings o each keyword corresponding to conjugations of the keyword since the keyword conjugates when it is a verb.
In this respect, in the keyword dlctionary 8, when a keyword ~e.g., t~ 7 ~ ~ ) ; (sa ku se i tsu ru));
~to draw up]) is the word which conjugates, the backward conjunction relationship is registered as ~1 so that the inflectional ending dictionary 10 can be used in looking up immediately.
~ ith reference to Fig. 3, the processing procedure of the morphemic analysis means 4 will be ?5 described. First, in step 20, flag, pointer, stack, etc., required for the subsequent processing are initialized. Next, in step 21, it is decided from a pointer of the input character train whether or not a ~2358~7 1 character train following the pointer exists in the ke~word dictionary 8, and when the result of the decision indicates that a keyword exists, proceeds to step 24 otherwise proceeds to step 22. When the keyword does not exist, in step 22, one vocabulary is etched from an altexnative solution stack, and it is decided whether or not the fetched vocabulary can be connected with a vocabulary before the fetched one. Here, when the fetched vocabulary is not connected to the preceding vocabulary, and if the stack for alternative solutions is vacant, proceeds to step 23, and undefined words of various lengths are prepared for the input character train follow-ing the pointer under the assumption that any one of the undefined words corresponds to the input character train, and after storing these undefined words in the alternative solution stack returns to the processing of the character train following the pointer in step 21.
When the fetched vocabulary is not connected with the preceding vocabulary, and if alterna~ive solutions are available in the alternative ~olution stack, the process in step 22 is again tried for the altexnative solutions.
In step 21, when the keyword exists, proceeds to step 24, and the longest one of the vocabularies found in step 21 i5 fetc~ed and the rest are stored in the alternative solution stack, and it is checked by using the conjunction relationship dictionary 9 whether or not the fetched vocabulary is connected with a vocabulary appearing just before the fetched vocabulary in the ~X3~
1 input sentence. When ~he connection is possible, proceeds to step 25 and decides whether or not the analysis has progressed to the end of the sentence, whereas when the connection is not possible, the process in step 22 i5 executed. Further, in step 22, when there is an alternative solution and the connection with the preceding vocabulary is possible, proceeds to step 25.
In step 25, when the analysis has not xeached the end of the sentence, the process in step 26 i5 executed, and a post processing for updating the pointer and flag is executed, and returns to step 21. On the other hand, in step 25, when the analysis has reached the end o:E the sentence, proceeds to step 27, and when undefined words exist in the analysis which has reached the end of the sentence, the shortest one among the undeflned words is selected as a irst candidate, and the morphemic analysis is completed. Further, in steps 22 and 24, in the case of checking the possibility of connection, it is supposed that the undefined word can be connected with any vocabulary.
Next, the morphemic analysis which is to be caxried out when the sentence shown in Fig. 2A as an example is inputted will be described. When the sentence is inputted to the morphemic analysis means 4, in the initiali2ing step 20, by looking up in the key-word dictionary of Table 1 a vocabulary in which the pointer is located at the beginning of the sentence, i.e., a position of " ~ " (mi) of ` ~" j 9 (mi gi shi ~3~ L7 l ta ) and the vocabulary begins with " ~ " (mi), "~ ~ "
(mi gi) (address 0007) and " ~ " (mi gi shi ta) (address 0~08) in ~he keyword dictionary of Table l are obtained. Then by applying a rule (the longest coincidence rule1 based on the experimental rule that the longest vocabulary has the highest probability of being correctly punctuated, " ~ " (mi gi shi ta) is obtained, and then in step 21, a forward conjunction condition of al is obtained, and in step 24, from the conjunction relationship dictionary of Table 2, data at the intersection of the beginning of the ssntence and the conjunction condition o al is fetched. Here, the data at the intersection is "l", and since the conjunc-tion condition is met, in step 26 via step 25, the pointer of ~he post processing indicates "~ " (no) in ~ " (no ba sho). Returning to step 21, by looking up in the keyword dictionary (Table l), a case particle ") " (no) at address 0005 is obtained. In the course of further processing in the similar way, although the sentence can be divided into vocabularies until " ~" (ni), the processing is deadlocked at "~ " (ku).
After the deadlock, in steps 22 and 23, undeined words are determined in such a manner as "~ " (ku), ~ 7 `~ ~
(ku u~ \" (ku u ha), "~ ~J~ " (ku u ha ), and these undefined words are stored in the alternative solution stack, and first, "~ " (ku) is selected as an undefined word, and cutting out of a vocabulary is tried for " 7 J\ --" (u ha ), and results in a failure.
~L23~ 7 l Then, by selecting "~ ku u~ as an undefined word and the division into vocabularies is further tried as to the sentence following the " ~`~ " (ku u) and also results in a failure. In this manner, after fetching an undefined word of " 7 ~ " (ku u ha ku i ki), proceeds to step 25, and the cutting out of vocabularies thus reaches the end of the sentence, and vocabularies of undefined words can be cut out as shown in Fig. 2B.
Here, a method of using the inflectional ending dictionary in Table 3 will be described hereinafter.
Specifically, at the time when the vocabulary of ~7t~
(sa ku se i) ~draw up] is cut out in step 21, it is found that this vocabulary is an irregular conjugation verb whose e.nding is in1ected in the "~ " (sa) column in the Japanese syllaba.ry, and as an inflectional ending, two kinds of word's endings including " ~" (shi) o the negative farm and " ~" (.shi) of the continuative form in Table 3 are obtained as candidates. At this time, s.lnce it is impossible to determine which one is to be adopted, these two candidates are stored in the stack, and at the ti~e in which the next vocabulary of ~ 9 ~ ~ (ta i) is cut out, the intersection of a forward conjunction condition a3 o "~ ~ " (ta i~ and a backward conjunction condition ~5 of the "~" (shi) of the negative form is obtained from the conjunction relationship dictionary in Table 2, and since the intersection is "0", it is found that the negative form does not satisfy the conjunction condition.
On the other hand, as to the " ~" (shi) of the continuative ~L~3~ L7 1 form, since the intersection of a backward conjunction condition ~6 of " ~" (shi) o the continuative orm and the forward conjunction condition ~3 is "1" in the conjunction relationship dictionary in Table 2, it is found that "~ ~ " (sa ku se i shi) is the continu-ative form and an auxiliary verb of "9~ " (ta i) expressing a wish is connected.
As a result of ~uch cutting out, a character train of vocabularies is obtained as shown in FigO 2C, and " ~`7J~ " (ku u ha ku i ki) is ob~ained as an undefined word.
1 BACKGROUND OE~ THE INVENTION
The present invention relates to a method and apparatus for processing a natural language such as Japanese, korean, chinese, etc., and in particular, to S a method and apparatus for natural language processing suitable for processing sentences which contain vocabu-laries not registered in a dictonary (hereinafter referred to as undefined words).
In performing sophisticated processing of a natural lanyuage, generally, syntax analysis of a sentence is carried out, and based on tha analysis, composition and editing o~ a new sentence are performed. However, in order to achieve the natural language processing, all vocabularies possible to compose a sentence are not always registered in a dictionary entirely, and thus even when a large dictionary is made available, undefined words which are not registered in the dictionary fre-quently appear in the sentence in many cases.
A technique to solve this problem is disclosed in Japanese patent unexamined publication Nos. 58-175074 (1983), 58-175075 (1983) and 58-175076. In the tech-nique disclosed in these publications, for an undefined word occurring as a result of failure in consulting a dictionary, the user is requested to input a synonym included in the words which have been registered and .,,` ~
~:35~3~7 l existing in the dictionary, and depending on this user's response, information for syntax analysis is added.
However~ in this technique, it is necessary for the user to memorize beforehand synonyms, etc., registered in the dictionary, or it is necessary for the system to output a list of all vocabularies regist~red in the dictionary 50 that the user himself extracts a synonym or the like from the list. Accordingly, in the case of a sophisticated processing, the number of registered words increases, and the burden on the user increases remarkably, and thus it has been difficult to carry out the natural language processing in a simple manner and effectively.
In view of the problems mentioned above, an object of the present invention is to provide a method and apparatus for natural language processing which is capable of performiny a sophisticated and practical na~ural language processing in that even when an un-defined word appears in a sentence to be processed, the user is pr~vided with information of a synonym from the apparatus to enable the processing with a dictionary of a limited storage capacityO
SUMMARY OF THE INVENTION
In the present invention, when an undefined word appears in an input character train of a natural language, an uppex concept of the undefined word is obtained by semantic analysis means which determines a 35~3~7 1 semantic concept of the vocabulary by using a case dictionary, and then by using a vocabulary hierarchy dictionary which registers vocabularies in a hierarchy stxucture formed according to the upper concept, a group of vocabularies corresponding to a lower concept are extracted. Further, it is also a featur~ of khe inven-tion in which the extracted group of vocabularies are displayed on a display device and the user is enabled to select a synonym from the group.
Another feature of the invention resides in that when an undefined word appears in an input character train of a natural language, an upper concept of the unde~ined word is obtained by semantic analysis means which determines a semantlc concept of a vocabulary by using a case dictionary, and then by using a vocabulary hierarchy dictionary which registers vocabularies in a hierarchy structure formed according to the upper concept, a group of vocabularies corresponding to a lower concept are extracted, and a synon~m is selected by inference means which selects from the extracted group of vocabularies of the lower concept a vocabulary by inference, of which vocabulary is in conformity with the intention of the inputted character train, and then the undefined word is replaced by the selected synonym.
BRIEF DESCRIPTION OF T~IE DRAWINGS
Fig. 1 is a block diagram of an overall arrange-a~7 1 ment in accordance with an embodiment of the present invention;
Figs. 2A to 2C show an example of an input sentence;
Fig. 3 is a flowchart for ~xplaining a proces-5 sing procedure of morphemic analysis means;
Fig. 4 is a flowchart for explaining a proces-sing procedure of syntax ana~ysis means, Fig. 5 shows an example of a vocabulary hierarchy dictionary;
Fig. 6 is a flowchart for explaining a proces-sing procedure of semantic analysis means;
Fig. 7 shows an example of a syntax tree;
Fig. 8 shows an example of display of a synonym;
Fig. 9 is a block diagram of an overall arrange-ment in accordance with another embodiment of the present invention;
Figs. lOA and lOB show respectively an example of a picture screen status of an information terminal and an example of representation of a knowledge base (status memory);
Fig~ 11 shows an example o~ a knowledge base (function memory); and Fig. 12 is a flowchart for explaining a processing procedure in accordance with another embodi-ment of the present invention.
-- 4 ~
i8~7 An embodiment of the present invention will be described with reference to the drawings.
Fig. 1 shows an overall arrangement of the invention, that is, shows an intelligent guidance syst~m of the question-answer type in which the user inputs by means of an input device 1 (e.g., a keyboard, voice recognition device, hand-wrikten character input device~
etc.) a natural language (e.g., Japanese) into a proces-sing unit 2, and the processing unit 2 executes processingas regards the natural language, and provides the user with inormation by displaying the contents and results of the processing on a display device 3. Here, the processing unit 2 is composed of a keyword dictionary 8, a con-junction relation5hip dictionary 9, an inflectional ending dictionary 10, and morphemi~ analysis means 4 for dividing an input sentence into vocabularies by use of in~rmation of these dictionaries 8 to 10; a grammatical rule dictionary 11, and syntax analysis means 5 for obtaining a syntax of the input sentence by use of the dictionary 11; a case dictionary 12, a vocabulary hierarchy dictionary 13, and semantic analysis means 5 for extracting a meaning of the input sentence by use of these dictionaries 12 and 13; a sentence pattern dictionary 14 containing several patterns of answer sentences, and sentence producing means 7 for producing a sentence by use of this dictionary 14.
Next, individual processings in the processing l unit 2 will be described in detail by way of an example in which as shown in Fig. 2A~ a sentence in kana charac-ters (Japanese syllabary) in solid writings:
(mi gi shi ta no ba sho ni ku u ha ku i ki wo !wish to draw up a blank area on the sa ku se i shi ta i) lower right place ]
is inputted as an input sentence.
~1) Morphemic Analysis First, in the morphenic analysis, by using the ke~word dictionary 8, aonjunction relationship dictionary 9, and inflectional ending dictionary 10, it is found whether or not a keyword exists in the character train o~ the input sentence, and a conjunction relationship is checked, and then the input sentence is divided into vocabularies. Here, in Tables l, 2, and 3, examples of the keyword dictionary 8, conjunction relationship dictionary 9, and inflectional ending dictionary lO are respectively shown.
~2~317 ,~, ____ ~ __ .,~ ~ ,~ ,~ ~ ~r er ,~ ,~ ,~ ~r . . .
'C1 ~ ~ ~n ~ c~ c~L ~ a:~ ~L C~L
o ~
; ~o . .
g ~
~ ~ t~ ~1 ~) el' el'r~ ~ r_l ~ ~ ~ ~
n ~ ~ ~ i~3 ~3 ~ ~ ~ ~
O O
_~ _ _ ~ ~ ~ a) Q) ~ ~I ~ 'U 'U 'U
~,~ ~ h ~rl aJ U 1: ~ ~ h ~ ~ ~i h . . .
U ~1 1~ ~ ~1~ O 1~ ~ (~ O O O ~d ,~ O ,~ ~ ~
~ r~ ~
~ ~ ,~ ~, .,., a) a) ~ ~ ~C Ul o~ U~
h 1~ S l ~ h ~ ~ 1~ td P~ ~h U ~-~ n~ U U U
~ - ----- --.----.
~l -~ ~
,_1 ~ ih5 ~< rl E~ ~ _ ~.~
~ ~Q~ _~ ~ ~
C~r~l r~ ~1 _ _ m~So ~ _ *~ _~
d ~ ~ ~ ~0~ ~ ~ ~ ~ 3 _ _ ~_ _ _ ~._ _ _ , _ __ _ __ _ _ ~a G) ~1 ~ ~ ~ u~ ~O 1` 00 a s~ O O O O O O O O O
O O O O O O O O O
~a o o o o o o o o o ~3~ 7 Table 2 Conjunction Relationship Dictionary ~ _ ___ _ _ beginning 1 2 3 4 5 6 sentence ~1 ~
O - ... ... ... -O l _ _ __ _ . _ ~ _ _. _ _ _ ending of sentence _ _ _ _ S8~L7 _ _ _ _ ~ o h O ~ ~ ~ . . .
~1 _, _ ,~ ~
:~ __--~
o u _ ~n _ ~ In ~ .,1 ~0 ca t~.C ...
s~
,~
~ .
h O ~ O
.,~
... .. .
1 In the keyword dictionary 8, there are registered with addresses, keywords, parts of spaech, and conjunction conditions, and in particular, in the conjunction conditions, there are registered with forward and backward conjunction relationships of each keyword.
Further, in the conjunction relationship dictionary 9, there are registered with correspondences between the conjunction relationships of each keyword mentioned above. For example, when a vocabulary having a forward conjunction condition of ~3 and another vocabulary having a backward conjunction condition of ~6 are pos~
sible to he connected with each other, then this xelation-ship is registered as "1". The inflectional ending dictionary 10 is us~d to register beforehand the endings o each keyword corresponding to conjugations of the keyword since the keyword conjugates when it is a verb.
In this respect, in the keyword dlctionary 8, when a keyword ~e.g., t~ 7 ~ ~ ) ; (sa ku se i tsu ru));
~to draw up]) is the word which conjugates, the backward conjunction relationship is registered as ~1 so that the inflectional ending dictionary 10 can be used in looking up immediately.
~ ith reference to Fig. 3, the processing procedure of the morphemic analysis means 4 will be ?5 described. First, in step 20, flag, pointer, stack, etc., required for the subsequent processing are initialized. Next, in step 21, it is decided from a pointer of the input character train whether or not a ~2358~7 1 character train following the pointer exists in the ke~word dictionary 8, and when the result of the decision indicates that a keyword exists, proceeds to step 24 otherwise proceeds to step 22. When the keyword does not exist, in step 22, one vocabulary is etched from an altexnative solution stack, and it is decided whether or not the fetched vocabulary can be connected with a vocabulary before the fetched one. Here, when the fetched vocabulary is not connected to the preceding vocabulary, and if the stack for alternative solutions is vacant, proceeds to step 23, and undefined words of various lengths are prepared for the input character train follow-ing the pointer under the assumption that any one of the undefined words corresponds to the input character train, and after storing these undefined words in the alternative solution stack returns to the processing of the character train following the pointer in step 21.
When the fetched vocabulary is not connected with the preceding vocabulary, and if alterna~ive solutions are available in the alternative ~olution stack, the process in step 22 is again tried for the altexnative solutions.
In step 21, when the keyword exists, proceeds to step 24, and the longest one of the vocabularies found in step 21 i5 fetc~ed and the rest are stored in the alternative solution stack, and it is checked by using the conjunction relationship dictionary 9 whether or not the fetched vocabulary is connected with a vocabulary appearing just before the fetched vocabulary in the ~X3~
1 input sentence. When ~he connection is possible, proceeds to step 25 and decides whether or not the analysis has progressed to the end of the sentence, whereas when the connection is not possible, the process in step 22 i5 executed. Further, in step 22, when there is an alternative solution and the connection with the preceding vocabulary is possible, proceeds to step 25.
In step 25, when the analysis has not xeached the end of the sentence, the process in step 26 i5 executed, and a post processing for updating the pointer and flag is executed, and returns to step 21. On the other hand, in step 25, when the analysis has reached the end o:E the sentence, proceeds to step 27, and when undefined words exist in the analysis which has reached the end of the sentence, the shortest one among the undeflned words is selected as a irst candidate, and the morphemic analysis is completed. Further, in steps 22 and 24, in the case of checking the possibility of connection, it is supposed that the undefined word can be connected with any vocabulary.
Next, the morphemic analysis which is to be caxried out when the sentence shown in Fig. 2A as an example is inputted will be described. When the sentence is inputted to the morphemic analysis means 4, in the initiali2ing step 20, by looking up in the key-word dictionary of Table 1 a vocabulary in which the pointer is located at the beginning of the sentence, i.e., a position of " ~ " (mi) of ` ~" j 9 (mi gi shi ~3~ L7 l ta ) and the vocabulary begins with " ~ " (mi), "~ ~ "
(mi gi) (address 0007) and " ~ " (mi gi shi ta) (address 0~08) in ~he keyword dictionary of Table l are obtained. Then by applying a rule (the longest coincidence rule1 based on the experimental rule that the longest vocabulary has the highest probability of being correctly punctuated, " ~ " (mi gi shi ta) is obtained, and then in step 21, a forward conjunction condition of al is obtained, and in step 24, from the conjunction relationship dictionary of Table 2, data at the intersection of the beginning of the ssntence and the conjunction condition o al is fetched. Here, the data at the intersection is "l", and since the conjunc-tion condition is met, in step 26 via step 25, the pointer of ~he post processing indicates "~ " (no) in ~ " (no ba sho). Returning to step 21, by looking up in the keyword dictionary (Table l), a case particle ") " (no) at address 0005 is obtained. In the course of further processing in the similar way, although the sentence can be divided into vocabularies until " ~" (ni), the processing is deadlocked at "~ " (ku).
After the deadlock, in steps 22 and 23, undeined words are determined in such a manner as "~ " (ku), ~ 7 `~ ~
(ku u~ \" (ku u ha), "~ ~J~ " (ku u ha ), and these undefined words are stored in the alternative solution stack, and first, "~ " (ku) is selected as an undefined word, and cutting out of a vocabulary is tried for " 7 J\ --" (u ha ), and results in a failure.
~L23~ 7 l Then, by selecting "~ ku u~ as an undefined word and the division into vocabularies is further tried as to the sentence following the " ~`~ " (ku u) and also results in a failure. In this manner, after fetching an undefined word of " 7 ~ " (ku u ha ku i ki), proceeds to step 25, and the cutting out of vocabularies thus reaches the end of the sentence, and vocabularies of undefined words can be cut out as shown in Fig. 2B.
Here, a method of using the inflectional ending dictionary in Table 3 will be described hereinafter.
Specifically, at the time when the vocabulary of ~7t~
(sa ku se i) ~draw up] is cut out in step 21, it is found that this vocabulary is an irregular conjugation verb whose e.nding is in1ected in the "~ " (sa) column in the Japanese syllaba.ry, and as an inflectional ending, two kinds of word's endings including " ~" (shi) o the negative farm and " ~" (.shi) of the continuative form in Table 3 are obtained as candidates. At this time, s.lnce it is impossible to determine which one is to be adopted, these two candidates are stored in the stack, and at the ti~e in which the next vocabulary of ~ 9 ~ ~ (ta i) is cut out, the intersection of a forward conjunction condition a3 o "~ ~ " (ta i~ and a backward conjunction condition ~5 of the "~" (shi) of the negative form is obtained from the conjunction relationship dictionary in Table 2, and since the intersection is "0", it is found that the negative form does not satisfy the conjunction condition.
On the other hand, as to the " ~" (shi) of the continuative ~L~3~ L7 1 form, since the intersection of a backward conjunction condition ~6 of " ~" (shi) o the continuative orm and the forward conjunction condition ~3 is "1" in the conjunction relationship dictionary in Table 2, it is found that "~ ~ " (sa ku se i shi) is the continu-ative form and an auxiliary verb of "9~ " (ta i) expressing a wish is connected.
As a result of ~uch cutting out, a character train of vocabularies is obtained as shown in FigO 2C, and " ~`7J~ " (ku u ha ku i ki) is ob~ained as an undefined word.
(2) Syntax Analysis The character train obtained in the morphemic analysls means 4 is the input to the next processing in the syntax analy~is means S. The syntax analysis means 5 executes a processing which determines a grammatical rule applicable by using the grammatical rule dictionary ll, and determine with which role in the grammatical rule the inputted vocabulary is burdened.
Here, an example of the grammatical rule dictionary ll is shown in Table 4.
~3~
Table 4 Grammatical Rule Dictionary _.
No. of Rule Left Side Right Side _.
1 sentence clause + "end mark"
2 clause predicate _
Here, an example of the grammatical rule dictionary ll is shown in Table 4.
~3~
Table 4 Grammatical Rule Dictionary _.
No. of Rule Left Side Right Side _.
1 sentence clause + "end mark"
2 clause predicate _
3 clause adverb phrase _
4 clause noun phase + clause noun phrase noun ~ particle _ _ _ _ _ _ 6 noun phrase clause ~ noun phrase _ . . _ _ _ _ _ _ 7 clause noun +
., .
1 The gra~natical rule dictionary 11 registers systematized grammatical rules, and or example, as a result o ~he morphemic analysis, when " ~ ~ j Y " (mi gi shi ta) and "~ " (no) are cut out, these resultant S vocabularies corre~pond to a rule of No. 5, the right side~ i.e., noun ~ particle, by referring to the left side, it is found that it is a noun phrase.
With reference to Fig. 4, the processing procedure of the syntax analysis means 5 will be described.
First, in step 30, a variable, stack, etc., re~uired for the syntax analysis are initialized.
Next, in step 31, a grammatical relationship ~23~i8~
1 which is identical with the one between a phrase at the beginning of the stack and grammatical information (exists in the keyword dictionary 8) of a vocabulary at the beginning of the input sentence is searched from the right side of the grammatical rule dictionary 11 shown in Table 4, and it is decided whether or not a rule corresponding to that grammatical relationship exists, and when the rule exists in the dictionary 11, executes step 32, and when the rul~ does not exist in the dictionary 11l step 33 is executed~ In the case in which the rule exists in the dictionary 11, proceeds to step 32, and when a plurality of rules are found, these rules are stored in data for alternative solutions, and by using one rule from these rules, the phrase at the beginning of the stack is combined with the beginning of the input sentence, and by assigning the grammatical condition to the left side of the grammatical rule, the combination is made as a vocabulary (phrase in this case) at the beginning o the input sentence. However, when there is an undefined word, since it is impossible to determine the grammatical condition of the undefined word, it is supposed that the undefined word can be connected with any vocabulary. When the rule does not exist in the dictionary, in step 33, a processing to stack the phrase at the beginning of the input sentence is executed, and in step 34, it is decided whether or not the input sentence after stac~ing on the stack has reached the end of the sentence, and when the end of the 3~i8~7 1 sentence has not yet been reached, the processing in step 31 is executed, whereas when the sentence-end has been reached, the processing in step 35 is executed. In step 35, it is decided whether or not an alternative solution has existed in the processing in step 31, and when the alternative solution has existed, proceeds to step 36, and when the alternative solution has not existed~ the syntax analysis is completed. In the step 36, one of the candidates of alternative solution stored in step 32, and data obtained as a result of the analysis conducted up to step 31 are set, and this enables to continue the processing subsequent to the processing in which the candidates for alternative solution~ have been ound.
Next, an example of operation for this proces-sing will be described. As an input sentence, the sentence as shown in Fig. 2B and additional grammatical information are inputted. Since the stack is vacant at first, steps 31 and 33 are executed, and "~ ~ j 9 "
(mi gi shi ta) is stored in the stack. Mext, in step 31, a conjunction relationship between "~ " ~no) and ~ " (mi gi shi ta) in the input sentence is consulted with the grammatical rule dictionary in Table 4, and it is found that noun + particle in the right side in rule No. 5 is applicable, and thus these portions of the input sentence are stored in the stack as one noun phrase. Next, since "l~ " (ba sho) is not connected with the aforementioned noun phrase, the ~23~
1 noun ")\ ~ ~ " (ba sho) is stacked on the stack.
Supposing that ")~ ~ ~ " (ba sho) and " _" (ni) constitute a noun phrase, and the undefined word of " ~ ~ J~ " (ku u ha ku i ki) is a noun, it is ound S that " ~ " (ku u ha ku i ki) and "~ " (wo) constitute a noun phrase, and " ~ ~t~ ~ 9 ~ " (sa ku sei shi ta i) is a predicate, and that by applying the left side of rule No. 2, these portions of the input sentence constitute a clause. Further, from rule No. 4, since noun phrase ~ clause (right side) is clause (left side), it i5 decided that " ~ ~91 (ku u ha ku i ki) (wo) (sa ku se i shi ta i) is a clause. In this manner, by analizing until the end of ~he inpu~ sentence is reache.d, it i9 found in accord-ance with clau~e + "end mark" in rule No. 1 that theinput sentence is a grammatical or meaning sentence, and the analysi~ is succe~sful.
(3) Seman~ic Analysis ~ased on the result of the syntax analysis, the semantic analysis means 6 obtains an upper concept of the undefined word from a grammatical condition by using the case dictionary 12, and urther performs a semantic analysis by using the vocabulary hierarchy dictionary 13 in which the upper concept is further formed in a hierarchy structure.
Here, an example of the case dictionary 12 is shown in Table 5, and an example of the vocabulary ~3~8~L7 1 hierarchy dictionary 13 is shown in Fig. 5.
TablP 5 Ca~e Dictionary ~1~ h~ ~ 3 (sa ku se i su ru) [draw up]
_ _ _ _ _ _ semantic condition grammatical condition concept default value _ _ _ _ .
subject noun + ~`(ga~ animal user place noun ~ l~tni) position case noun ~ ~(wo) indication _.
priority -The case dictionary 12 registers a semantic concept in a corresponding relationship with the grammatical condition of a vocabulary (in this case,
., .
1 The gra~natical rule dictionary 11 registers systematized grammatical rules, and or example, as a result o ~he morphemic analysis, when " ~ ~ j Y " (mi gi shi ta) and "~ " (no) are cut out, these resultant S vocabularies corre~pond to a rule of No. 5, the right side~ i.e., noun ~ particle, by referring to the left side, it is found that it is a noun phrase.
With reference to Fig. 4, the processing procedure of the syntax analysis means 5 will be described.
First, in step 30, a variable, stack, etc., re~uired for the syntax analysis are initialized.
Next, in step 31, a grammatical relationship ~23~i8~
1 which is identical with the one between a phrase at the beginning of the stack and grammatical information (exists in the keyword dictionary 8) of a vocabulary at the beginning of the input sentence is searched from the right side of the grammatical rule dictionary 11 shown in Table 4, and it is decided whether or not a rule corresponding to that grammatical relationship exists, and when the rule exists in the dictionary 11, executes step 32, and when the rul~ does not exist in the dictionary 11l step 33 is executed~ In the case in which the rule exists in the dictionary 11, proceeds to step 32, and when a plurality of rules are found, these rules are stored in data for alternative solutions, and by using one rule from these rules, the phrase at the beginning of the stack is combined with the beginning of the input sentence, and by assigning the grammatical condition to the left side of the grammatical rule, the combination is made as a vocabulary (phrase in this case) at the beginning o the input sentence. However, when there is an undefined word, since it is impossible to determine the grammatical condition of the undefined word, it is supposed that the undefined word can be connected with any vocabulary. When the rule does not exist in the dictionary, in step 33, a processing to stack the phrase at the beginning of the input sentence is executed, and in step 34, it is decided whether or not the input sentence after stac~ing on the stack has reached the end of the sentence, and when the end of the 3~i8~7 1 sentence has not yet been reached, the processing in step 31 is executed, whereas when the sentence-end has been reached, the processing in step 35 is executed. In step 35, it is decided whether or not an alternative solution has existed in the processing in step 31, and when the alternative solution has existed, proceeds to step 36, and when the alternative solution has not existed~ the syntax analysis is completed. In the step 36, one of the candidates of alternative solution stored in step 32, and data obtained as a result of the analysis conducted up to step 31 are set, and this enables to continue the processing subsequent to the processing in which the candidates for alternative solution~ have been ound.
Next, an example of operation for this proces-sing will be described. As an input sentence, the sentence as shown in Fig. 2B and additional grammatical information are inputted. Since the stack is vacant at first, steps 31 and 33 are executed, and "~ ~ j 9 "
(mi gi shi ta) is stored in the stack. Mext, in step 31, a conjunction relationship between "~ " ~no) and ~ " (mi gi shi ta) in the input sentence is consulted with the grammatical rule dictionary in Table 4, and it is found that noun + particle in the right side in rule No. 5 is applicable, and thus these portions of the input sentence are stored in the stack as one noun phrase. Next, since "l~ " (ba sho) is not connected with the aforementioned noun phrase, the ~23~
1 noun ")\ ~ ~ " (ba sho) is stacked on the stack.
Supposing that ")~ ~ ~ " (ba sho) and " _" (ni) constitute a noun phrase, and the undefined word of " ~ ~ J~ " (ku u ha ku i ki) is a noun, it is ound S that " ~ " (ku u ha ku i ki) and "~ " (wo) constitute a noun phrase, and " ~ ~t~ ~ 9 ~ " (sa ku sei shi ta i) is a predicate, and that by applying the left side of rule No. 2, these portions of the input sentence constitute a clause. Further, from rule No. 4, since noun phrase ~ clause (right side) is clause (left side), it i5 decided that " ~ ~91 (ku u ha ku i ki) (wo) (sa ku se i shi ta i) is a clause. In this manner, by analizing until the end of ~he inpu~ sentence is reache.d, it i9 found in accord-ance with clau~e + "end mark" in rule No. 1 that theinput sentence is a grammatical or meaning sentence, and the analysi~ is succe~sful.
(3) Seman~ic Analysis ~ased on the result of the syntax analysis, the semantic analysis means 6 obtains an upper concept of the undefined word from a grammatical condition by using the case dictionary 12, and urther performs a semantic analysis by using the vocabulary hierarchy dictionary 13 in which the upper concept is further formed in a hierarchy structure.
Here, an example of the case dictionary 12 is shown in Table 5, and an example of the vocabulary ~3~8~L7 1 hierarchy dictionary 13 is shown in Fig. 5.
TablP 5 Ca~e Dictionary ~1~ h~ ~ 3 (sa ku se i su ru) [draw up]
_ _ _ _ _ _ semantic condition grammatical condition concept default value _ _ _ _ .
subject noun + ~`(ga~ animal user place noun ~ l~tni) position case noun ~ ~(wo) indication _.
priority -The case dictionary 12 registers a semantic concept in a corresponding relationship with the grammatical condition of a vocabulary (in this case,
5 "~ ~3" (.sakuseisuro) [draw up]). For example, when "noun ~ ~`(ga)" is being connected with " ~ 3 "
(sakuseisuru), according to this dictionary, i.t can be understood that the noun represents an animal as its concept. Here, the default value is intended to mean a value which is applied to an omittable item (which will be described later) in the course of the semantic analysis. Further, in the vocabulary heirarchy dictionary ~235~3~7 1 13 shown in Fig. 5, there are registered with vocabu-laries of a lower concept in a hierarchy structure whose upper concept is the concept in the case dictionary 12.
With reference to Fig. 6, a processing 5 procedure of the semantic analysis means 6 will be described. First, in step 40, from the output of the syntax analysis means 5, a vocabulary (e~g., a verb constituting a predicate) of a highest priority is taken out wi~h the exception of an undefined word, and infor-mation of the vocabulary is obtained by using the casedictionary 12 as shown in Table 5.
Next, in step 41, from phrases in the input sentence, a phrase which satisies the gra~matical condition oE the vocabulary obtained in step 40 is selected, and the selected phrase is substituted in a slot (a concept enclos~d in a block in Fig. 7) of a syntax tree. And in step 42, i~ is checked whether or not the completed syntax tree satisfies the semantic condition, and when an undeined word exists, proceeds to step 43, and when the semantic condition is not satisfied proceeds to step 44, and in other cases returns. In step 43, a concept of ~he slot in which the undefined word is to be entered is obtained, and its lower concept is obtained from the vocabulary hierarchy dictionary 13 in order to use as data for producing a sentence. In step 44, the processing is programmed to obtain an alternative solution of the syntax analysis, when the semantic analysis is not successful.
~235~3~7 l This operation will be described as regards the example used in the previous description. In step 40, it is made clear that the verb constituting the predicate is ~ 3 ~ (sa~useisuru~ [draw up~, and the case dickionary in Table 5 is fetched. ~ phrase corresponding to an item of the grammatical rule of the case dictionary is fetched from the output of the syntax analysis means 5, and this phrase is applied to the item~ For example, since a phrase "noun ~ ~r"(ga)" coxresponding to the subject of action is not found in the sentence, it is assumed that this phrase is omitted, and applies ~ `" (u za) of the default value. Next, it is found that the place case is an indispensable case which can not be omitted, and that this place case corresponds lS to word~ and phrases " ~ " (migishita no basho) [place a~ the lower right] ~ " ~ " (ni), and that its objective case is " ~ ~1 J~ " (ku u ha ku i ki) +
" ~" (wo). Next, in step 42~ it is checked whether ~ " (migishita no basho ni) represents a place.
In this case, the semantic condition for representing a position is satisfied~ Since " ~ ~ J~ " (ku u ha ku i ki) is an undefined word, the semantic condition is not satisfied. However, it is also found that for a semantic condition to be satisfied, it is required that the semantic condition is a concept of "indication object".
Fig. 5 is an example of the hierarchy structure of a vocabulary formed by putting such concept in order, and it is seen that " ~ " (ku u ha ku i ki) is a ~2 -.
~23533~7 l lowex concept of the "indication object", and is a kirld of "window" ~ "character". A syntax tree formed in this manner is shown in Fig. 7.
(4) Sentence Production The sentence producing means 7~ based on the result of the semantic analysis means 6, calls a question pattern of the undefined word from the sentence pattern dictionary 14, and fills the ~uestion pattern with vocabularies of the hierarchy structure obtained by the semantic analysis means 6, and displays it as a question sentence as shown in Fig. 8, and enables the user to select a synonym.
When the user selects the synonym, the system proceeds with the analysis further by using various information of the synonym in the dictionary.
By vir~ue of such an arrangement, even when the sentence as shown in Fig. 2A is inputted, the undefined word of " ~ " (ku u ha ku i ki) i5 fetched as a vocabulary, and the user is enabled to select a synonym from the vocabularies displayed on the display device 3 as shown in Fig. 8. When the user inputs a vocabulary o~ " [blank], the system proceeds wi~h subsequent analysis by using various kinds of informa~ion of the vocabulary of "~ " [blank].
In the prior art, when it is not clear whether or not a synonym exists, a vocabulary is entered, and when there is no synonym, another vocabulary had to be l entered again anew. However, in the present invention, vocabularies which belong to the concept of the undefined word are displayed, and these vocabularies are displayed without fail if synonyms thereof are registered in the dictionary. Accordingly, the system is very easy for the user to use~
Next, another embodiment of the present inven~
tion will be described with reference to Figs. 9 to 12. Differences in this embodiment from that shown in Fig. 1 reside in that as shown in Fig. 9, an object 100, knowledge base processing means 101, knowledge base 102, and inference means 103 are added. Here, the object 100 serves to restrict a theme, and information is stored in the knowledge base 102 following on a variation in status, through the knowledge base processing means 101 which produces and update~ the knowledge base 102. On the other hand, in the inference means 103, the syntax tree (Fig. 7) which is the output of the semantic analysis means 6 mentioned in the previous embodiment is inputted, and by looking up in the knowledge base 102, a synonym of the undefined word i~ automatically substi~uted and displayed ~or the user.
Here, aq an example of the object 100, a processing of a natural language as regards an "informa-tion terminal system" will be described. Here, the"information terminal system" means an apparatus which includes an interface which deals with information between human and a computer. Thus, in the example of , :
~3~ 7 l the language processing which will be described herein-after, in conducting the inference by using the knowledge base, the object of a theme is restricted to a concxete operation method or processing method of the "inf~rmation terminal system". Naturally, in the present invention, the object 100 is not limited to the l'infor-mation terminal system", but the natural language processing is possible for other objects of themes.
In the knowledge base processing means 101 connected with the information terminal on line, as a method of representing its knowledge ba~e 102, various representations axe available such as predicate logic, frame, and the like. For example, here, in representing a status of the picture screen of the information terminal as shown in Fig. lOA, if the picture screen status i9 represented as shown in Fig. lOB, the knowledge base processing means 101 actually performs a data conversion processing or converting an internal representation of the information terminal to a knowledge base description type. Here, in Fig. lOB, it is represented to mean that information given a name "place l" belongs to a klnd of place, and its position is at the lower right place, and a sentence given a name "sentence l" exists at the place. Similarly, "place 2" represents infor~
Z5 mation of a kind of place, and its position is at upper left place, and a figure given a name "Fig. 1" exists.
Such information is represented so that the content of the picture screen display is varied each 1 time the user manipulates the information terminal, and for example, it is arranged that when the figure at the upper left in Fig. 10A is erased, the items of (place 2 ( ~ in Fig. 10B disappear.
On ~he other hand, in the knowledge base 102, in addition to the information ~a set of such information is called a status memory) which varies following on a status change of the information terminal as described above and shown in Fig. 10B, a function memory which expresses the functions of the in~ormation terminalas shown in Fig. 11 i5 included. This function memory is composed of, for each address, ~ verb, ~ command name, ~ precondition, and ~ goalO In the ~ command name, in view of the fact that the status of the information terminal i~ varied each time a key is mani-pulated, a command name corresponding to the manipulation i~ stored. In the ~ precondition, a status of the inormation terminal which is khe premise required to execute the key manipulation is stored. In the ~ goal, a new status produced after the execution of the key manipulation is stored. And in the ~ verb, verbs (each verb and its synonyms, etc., are collected in a group, and included in an upper concept or the like) in the input sentence are stored, and each verb is made to correspond to the ~ command name which is the action representing the key manipulation in the input sentenceO
Further, the representation method of these knowledge base i5 hereinaf~er referred to as a rame.
_ 26 ~23~ L7 1 Since the function memory of the knowledge base provides a general representation, names having a mark # attached thereto are variable~, and when correspondences of these ~ariables with the input sentence and the status memory of the knowledge base are established, the variables are bound with respective constants with which the corxespondences have been established.
When executing the processing describad in the foregoing embodiment, the output of the semantic lQ analysis is the syntax tree (.Fig. 7). Upon receipt of this syntax tree, the inference means 103 executes a processing as shown in Fig. 12.
~ ere, since the syntax tree which is the output of the semantic analysis means 6 differs from the frame representation describing the knowledge base, a syntax tree conversion rule dictionary (not shown in ~igure) shown in Table 6, and a sentence type determination dictionary (not shown in figure) shown in Table 7 are used~
~7 -~;~3S~7 Table 6 Syntax Tree Conversion Rule Dictionary !
draw up _ _ _ _ _ syntax tree frame structure , tdraw up ~.draw up (kind (meaning))) ~ (branch 1 (subject of action)) (place ~exist : ( " 2 (place)~ objective case))) ( " 3 (objective case))) _ _ _ Table 7 Sentence Type Determination Dictionary sentence-end o sentence type input qentence (tài) (wish) HOW type _ _ ~ ~ ~ 3 ~` RESULT type (dou naru ka) . _ (suru houhou ha) ~OW type _ l In the processing pxocedure shown in Fig. 12, in step 110, by using the syntax tree CQnverSiOn rule dictionary shown in Table 6, the syntax tree is converted to the frame representation, and in step 120, undefined 3~
1 words in this frame are made variahles.
Next, from the syntax tree in Fig. 7, since it is understood that this sentence is a wish, that is, a question as to an operation method of the in~ormation S terminal, in step 130, by using this information and the sentence type determination dictionary sho~n in Table 7, the sentence type is decided and proceeds to step 140. In step 140, from the verbs in the input sentence, all commands which coin~ide with these verbs are selected from the knowledge base. In order to execute these commands, in step 150, a matching between the precondition which has to be satisfied by the infor-mation terminal and the status o~ the information terminal is executed, and also a binding of the variables is executed, and commands which have passed the matching are selected.
In step 160, the goal status of the selected command is matched with the content of the syntax tree, and the binding of the variables is executed, and items which coincide with the goal status are fetched. In step 170, the results o the binding between the variable and the constant (e.g., place 1, Fig. 1, sentenca 1, etc.,) executed in steps 150 and 160 are fetched. Before entering the main processing, since it can not be determined what the undefined words indicate, the un-defined words have been regarded as variables, however, as a result of the execution in step 170, the constant which has been bound to the variable of the undefined 5~
1 word becomes a synonym of the undefined word.
Here, the method of matching and binding of variable is descri~ed in detail in LISP by P~trick H.
Winston & Berthold K. P. Horn (Addison-Wesley Publishing Company, Inc., Reading, Mass., U.S.A ).
This processing will be described by way of an example. The syntax tree in Fig. 7 is a logical sum o~ 3 [draw upJ (kind (action))) and (~ ~ 5~
[place at lower right] (exi~t (# ~ (ku u ha ku i ki) [blank area~))), and it is converted in accordance with the syntax tree conversion rule dictionary in Table
(sakuseisuru), according to this dictionary, i.t can be understood that the noun represents an animal as its concept. Here, the default value is intended to mean a value which is applied to an omittable item (which will be described later) in the course of the semantic analysis. Further, in the vocabulary heirarchy dictionary ~235~3~7 1 13 shown in Fig. 5, there are registered with vocabu-laries of a lower concept in a hierarchy structure whose upper concept is the concept in the case dictionary 12.
With reference to Fig. 6, a processing 5 procedure of the semantic analysis means 6 will be described. First, in step 40, from the output of the syntax analysis means 5, a vocabulary (e~g., a verb constituting a predicate) of a highest priority is taken out wi~h the exception of an undefined word, and infor-mation of the vocabulary is obtained by using the casedictionary 12 as shown in Table 5.
Next, in step 41, from phrases in the input sentence, a phrase which satisies the gra~matical condition oE the vocabulary obtained in step 40 is selected, and the selected phrase is substituted in a slot (a concept enclos~d in a block in Fig. 7) of a syntax tree. And in step 42, i~ is checked whether or not the completed syntax tree satisfies the semantic condition, and when an undeined word exists, proceeds to step 43, and when the semantic condition is not satisfied proceeds to step 44, and in other cases returns. In step 43, a concept of ~he slot in which the undefined word is to be entered is obtained, and its lower concept is obtained from the vocabulary hierarchy dictionary 13 in order to use as data for producing a sentence. In step 44, the processing is programmed to obtain an alternative solution of the syntax analysis, when the semantic analysis is not successful.
~235~3~7 l This operation will be described as regards the example used in the previous description. In step 40, it is made clear that the verb constituting the predicate is ~ 3 ~ (sa~useisuru~ [draw up~, and the case dickionary in Table 5 is fetched. ~ phrase corresponding to an item of the grammatical rule of the case dictionary is fetched from the output of the syntax analysis means 5, and this phrase is applied to the item~ For example, since a phrase "noun ~ ~r"(ga)" coxresponding to the subject of action is not found in the sentence, it is assumed that this phrase is omitted, and applies ~ `" (u za) of the default value. Next, it is found that the place case is an indispensable case which can not be omitted, and that this place case corresponds lS to word~ and phrases " ~ " (migishita no basho) [place a~ the lower right] ~ " ~ " (ni), and that its objective case is " ~ ~1 J~ " (ku u ha ku i ki) +
" ~" (wo). Next, in step 42~ it is checked whether ~ " (migishita no basho ni) represents a place.
In this case, the semantic condition for representing a position is satisfied~ Since " ~ ~ J~ " (ku u ha ku i ki) is an undefined word, the semantic condition is not satisfied. However, it is also found that for a semantic condition to be satisfied, it is required that the semantic condition is a concept of "indication object".
Fig. 5 is an example of the hierarchy structure of a vocabulary formed by putting such concept in order, and it is seen that " ~ " (ku u ha ku i ki) is a ~2 -.
~23533~7 l lowex concept of the "indication object", and is a kirld of "window" ~ "character". A syntax tree formed in this manner is shown in Fig. 7.
(4) Sentence Production The sentence producing means 7~ based on the result of the semantic analysis means 6, calls a question pattern of the undefined word from the sentence pattern dictionary 14, and fills the ~uestion pattern with vocabularies of the hierarchy structure obtained by the semantic analysis means 6, and displays it as a question sentence as shown in Fig. 8, and enables the user to select a synonym.
When the user selects the synonym, the system proceeds with the analysis further by using various information of the synonym in the dictionary.
By vir~ue of such an arrangement, even when the sentence as shown in Fig. 2A is inputted, the undefined word of " ~ " (ku u ha ku i ki) i5 fetched as a vocabulary, and the user is enabled to select a synonym from the vocabularies displayed on the display device 3 as shown in Fig. 8. When the user inputs a vocabulary o~ " [blank], the system proceeds wi~h subsequent analysis by using various kinds of informa~ion of the vocabulary of "~ " [blank].
In the prior art, when it is not clear whether or not a synonym exists, a vocabulary is entered, and when there is no synonym, another vocabulary had to be l entered again anew. However, in the present invention, vocabularies which belong to the concept of the undefined word are displayed, and these vocabularies are displayed without fail if synonyms thereof are registered in the dictionary. Accordingly, the system is very easy for the user to use~
Next, another embodiment of the present inven~
tion will be described with reference to Figs. 9 to 12. Differences in this embodiment from that shown in Fig. 1 reside in that as shown in Fig. 9, an object 100, knowledge base processing means 101, knowledge base 102, and inference means 103 are added. Here, the object 100 serves to restrict a theme, and information is stored in the knowledge base 102 following on a variation in status, through the knowledge base processing means 101 which produces and update~ the knowledge base 102. On the other hand, in the inference means 103, the syntax tree (Fig. 7) which is the output of the semantic analysis means 6 mentioned in the previous embodiment is inputted, and by looking up in the knowledge base 102, a synonym of the undefined word i~ automatically substi~uted and displayed ~or the user.
Here, aq an example of the object 100, a processing of a natural language as regards an "informa-tion terminal system" will be described. Here, the"information terminal system" means an apparatus which includes an interface which deals with information between human and a computer. Thus, in the example of , :
~3~ 7 l the language processing which will be described herein-after, in conducting the inference by using the knowledge base, the object of a theme is restricted to a concxete operation method or processing method of the "inf~rmation terminal system". Naturally, in the present invention, the object 100 is not limited to the l'infor-mation terminal system", but the natural language processing is possible for other objects of themes.
In the knowledge base processing means 101 connected with the information terminal on line, as a method of representing its knowledge ba~e 102, various representations axe available such as predicate logic, frame, and the like. For example, here, in representing a status of the picture screen of the information terminal as shown in Fig. lOA, if the picture screen status i9 represented as shown in Fig. lOB, the knowledge base processing means 101 actually performs a data conversion processing or converting an internal representation of the information terminal to a knowledge base description type. Here, in Fig. lOB, it is represented to mean that information given a name "place l" belongs to a klnd of place, and its position is at the lower right place, and a sentence given a name "sentence l" exists at the place. Similarly, "place 2" represents infor~
Z5 mation of a kind of place, and its position is at upper left place, and a figure given a name "Fig. 1" exists.
Such information is represented so that the content of the picture screen display is varied each 1 time the user manipulates the information terminal, and for example, it is arranged that when the figure at the upper left in Fig. 10A is erased, the items of (place 2 ( ~ in Fig. 10B disappear.
On ~he other hand, in the knowledge base 102, in addition to the information ~a set of such information is called a status memory) which varies following on a status change of the information terminal as described above and shown in Fig. 10B, a function memory which expresses the functions of the in~ormation terminalas shown in Fig. 11 i5 included. This function memory is composed of, for each address, ~ verb, ~ command name, ~ precondition, and ~ goalO In the ~ command name, in view of the fact that the status of the information terminal i~ varied each time a key is mani-pulated, a command name corresponding to the manipulation i~ stored. In the ~ precondition, a status of the inormation terminal which is khe premise required to execute the key manipulation is stored. In the ~ goal, a new status produced after the execution of the key manipulation is stored. And in the ~ verb, verbs (each verb and its synonyms, etc., are collected in a group, and included in an upper concept or the like) in the input sentence are stored, and each verb is made to correspond to the ~ command name which is the action representing the key manipulation in the input sentenceO
Further, the representation method of these knowledge base i5 hereinaf~er referred to as a rame.
_ 26 ~23~ L7 1 Since the function memory of the knowledge base provides a general representation, names having a mark # attached thereto are variable~, and when correspondences of these ~ariables with the input sentence and the status memory of the knowledge base are established, the variables are bound with respective constants with which the corxespondences have been established.
When executing the processing describad in the foregoing embodiment, the output of the semantic lQ analysis is the syntax tree (.Fig. 7). Upon receipt of this syntax tree, the inference means 103 executes a processing as shown in Fig. 12.
~ ere, since the syntax tree which is the output of the semantic analysis means 6 differs from the frame representation describing the knowledge base, a syntax tree conversion rule dictionary (not shown in ~igure) shown in Table 6, and a sentence type determination dictionary (not shown in figure) shown in Table 7 are used~
~7 -~;~3S~7 Table 6 Syntax Tree Conversion Rule Dictionary !
draw up _ _ _ _ _ syntax tree frame structure , tdraw up ~.draw up (kind (meaning))) ~ (branch 1 (subject of action)) (place ~exist : ( " 2 (place)~ objective case))) ( " 3 (objective case))) _ _ _ Table 7 Sentence Type Determination Dictionary sentence-end o sentence type input qentence (tài) (wish) HOW type _ _ ~ ~ ~ 3 ~` RESULT type (dou naru ka) . _ (suru houhou ha) ~OW type _ l In the processing pxocedure shown in Fig. 12, in step 110, by using the syntax tree CQnverSiOn rule dictionary shown in Table 6, the syntax tree is converted to the frame representation, and in step 120, undefined 3~
1 words in this frame are made variahles.
Next, from the syntax tree in Fig. 7, since it is understood that this sentence is a wish, that is, a question as to an operation method of the in~ormation S terminal, in step 130, by using this information and the sentence type determination dictionary sho~n in Table 7, the sentence type is decided and proceeds to step 140. In step 140, from the verbs in the input sentence, all commands which coin~ide with these verbs are selected from the knowledge base. In order to execute these commands, in step 150, a matching between the precondition which has to be satisfied by the infor-mation terminal and the status o~ the information terminal is executed, and also a binding of the variables is executed, and commands which have passed the matching are selected.
In step 160, the goal status of the selected command is matched with the content of the syntax tree, and the binding of the variables is executed, and items which coincide with the goal status are fetched. In step 170, the results o the binding between the variable and the constant (e.g., place 1, Fig. 1, sentenca 1, etc.,) executed in steps 150 and 160 are fetched. Before entering the main processing, since it can not be determined what the undefined words indicate, the un-defined words have been regarded as variables, however, as a result of the execution in step 170, the constant which has been bound to the variable of the undefined 5~
1 word becomes a synonym of the undefined word.
Here, the method of matching and binding of variable is descri~ed in detail in LISP by P~trick H.
Winston & Berthold K. P. Horn (Addison-Wesley Publishing Company, Inc., Reading, Mass., U.S.A ).
This processing will be described by way of an example. The syntax tree in Fig. 7 is a logical sum o~ 3 [draw upJ (kind (action))) and (~ ~ 5~
[place at lower right] (exi~t (# ~ (ku u ha ku i ki) [blank area~))), and it is converted in accordance with the syntax tree conversion rule dictionary in Table
6 in steps 110 and 120. Here, the mark #indicates that the attached word is a variable. From the sentence type determination dictionary in Table 7 t it is found that the type of the sentence is a HOW type ~a sentence which states a goal status and inquiries a means ~or realizing the goall as determined from an auxiliary verb o~ the wish. When searching from the knowledge base (function memory~ of Fig. 11, a verb of the first item of the frame in step 120 r of which verb means (~
[draw up] ~kind (action))), the matching is established with the "space" command and "draw flgure" command.
Next, matching is tried between the st~tus memory (this iS al50 included in the knowledge base) of Fig. 10B
which being the present status and the precondition.
When the "place at lower right" is substituted for the '# place name' in the precondition, it becomes clear in st~p 150 that the matching with the "space" command is ~3~8~ 7 1 successful whereas the matching with the "draw figure"
coxNmand is not successful. Step 160 is executed tc obtain matching between the status directly represented by the input sentence, i.e., ( ~ place at lower right]
(~ ~ ~exist] (# 7 ~ )~ 7 ~ ~ (ku u ha ku i ki) [blank area]))) and (~ [place at lower right] ( ~ ~
[exist] (~ ~ [blank]))) obtained in step 150 by substi-tuting the value of the variable for the goal in Fig. 11, and # ~ `~ f~ (ku u ha ku i ki) [blank area] and ~ ~ [blank] are bound. As a result, it i~ found that ~ ~ 7 1~ (ku u ha ku i ki) [blank area] is a synonym o~ " (yohaku) [blank]. In the sentence producing mean~ 7, it i5 displayed for the user that " ~ ~ J~ (ku u ha ku i ki) should be interpreted by replacing it by "~ b ~ ~blank].
In this embodiment of the present invention, it is shown that a synonym i~ made automatically select-able.
In this embodiment, although the description is made in connection with the 'linformation terminal system", it will be apparent from the embodiment that the present invention is applicable not only to the "information terminal system" but to other systems.
According to the present invention, even when the undefined word appears, information o~ the synonym is extracted and provided for the user. Thus, the user is not required to store all synonyms, and it is ; possible to achieve the natural language processing with ~2~ 7 a dictionary having a limited storage capacity, of which processing is practical also to the user and efficient.
, , . ..... . .
[draw up] ~kind (action))), the matching is established with the "space" command and "draw flgure" command.
Next, matching is tried between the st~tus memory (this iS al50 included in the knowledge base) of Fig. 10B
which being the present status and the precondition.
When the "place at lower right" is substituted for the '# place name' in the precondition, it becomes clear in st~p 150 that the matching with the "space" command is ~3~8~ 7 1 successful whereas the matching with the "draw figure"
coxNmand is not successful. Step 160 is executed tc obtain matching between the status directly represented by the input sentence, i.e., ( ~ place at lower right]
(~ ~ ~exist] (# 7 ~ )~ 7 ~ ~ (ku u ha ku i ki) [blank area]))) and (~ [place at lower right] ( ~ ~
[exist] (~ ~ [blank]))) obtained in step 150 by substi-tuting the value of the variable for the goal in Fig. 11, and # ~ `~ f~ (ku u ha ku i ki) [blank area] and ~ ~ [blank] are bound. As a result, it i~ found that ~ ~ 7 1~ (ku u ha ku i ki) [blank area] is a synonym o~ " (yohaku) [blank]. In the sentence producing mean~ 7, it i5 displayed for the user that " ~ ~ J~ (ku u ha ku i ki) should be interpreted by replacing it by "~ b ~ ~blank].
In this embodiment of the present invention, it is shown that a synonym i~ made automatically select-able.
In this embodiment, although the description is made in connection with the 'linformation terminal system", it will be apparent from the embodiment that the present invention is applicable not only to the "information terminal system" but to other systems.
According to the present invention, even when the undefined word appears, information o~ the synonym is extracted and provided for the user. Thus, the user is not required to store all synonyms, and it is ; possible to achieve the natural language processing with ~2~ 7 a dictionary having a limited storage capacity, of which processing is practical also to the user and efficient.
, , . ..... . .
Claims (7)
1. A natural language processing apparatus for processing a character train comprising:
morphemic analysis means for dividing an inputted character train of a natural language into vocabularies;
syntax analysis means for deciding whether or not a conjunction of the vocabularies is in conformity with a grammatical rule;
a case dictionary for registering a semantic role between the vocabularies which are in conformity with the grammatical rule;
semantic analysis means for determining a semantic concept of the vocabulary in an output of said syntax analysis means by using said case dictionary; and a vocabulary hierarchy dictionary for register-ing vocabularies by classifying the same from an upper order to a lower order in accordance with the semantic concept of each of the vocabularies, wherein when said inputted character train includes an undefined word, the semantic concept of said undefined word is obtained by said semantic analysis means, and a group of vocabularies corresponding to a lower concept of the semantic concept of said undefined word are extracted from said vocabulary hierarchy dictionary.
morphemic analysis means for dividing an inputted character train of a natural language into vocabularies;
syntax analysis means for deciding whether or not a conjunction of the vocabularies is in conformity with a grammatical rule;
a case dictionary for registering a semantic role between the vocabularies which are in conformity with the grammatical rule;
semantic analysis means for determining a semantic concept of the vocabulary in an output of said syntax analysis means by using said case dictionary; and a vocabulary hierarchy dictionary for register-ing vocabularies by classifying the same from an upper order to a lower order in accordance with the semantic concept of each of the vocabularies, wherein when said inputted character train includes an undefined word, the semantic concept of said undefined word is obtained by said semantic analysis means, and a group of vocabularies corresponding to a lower concept of the semantic concept of said undefined word are extracted from said vocabulary hierarchy dictionary.
2. An apparatus according to Claim 1, wherein said semantic analysis means fetches a vocabulary having a high priority except said undefined word from the output of said syntax analysis means, and the semantic concept of said undefined word is obtained by using said case dictionary.
3. A system according to Claim 2, wherein said vocabulary having a high priority is a verb constituting a predicate.
4. A natural language processing apparatus for processing a character train comprising:
morphemic analysis means for dividing an inputted character train of a natural language into vocabularies;
syntax analysis means for deciding whether or not a conjunction of the vocabularies is in conformity with a grammatical rule;
a display device for displaying a content and result of the processing of said character train;
a case dictionary for registering a semantic role between the vocabularies which are in conformity with the grammatical rule;
semantic analysis means for determining a semantic concept of the vocabulary in the output of said syntax analysis means by using said case dictionary;
and a vocabulary hierarchy dictionary for register-ing vocabularies by classifying the same from an upper order to a lower order in accordance with a semantic concept of each of the vocabularies, wherein when said inputted character train includes an undefined word, a semantic concept of said undefined word is obtained by said semantic analysis means, and a group of vocabularies corresponding to a lower concept of the semantic concept of said undefined word are extracted from said vocabulary hierarchy dictionary, and said extracted group of vocabularies of the lower concept are displayed on said display device so that a synonym of said undefined word is selected from said displayed group of vocabularies.
morphemic analysis means for dividing an inputted character train of a natural language into vocabularies;
syntax analysis means for deciding whether or not a conjunction of the vocabularies is in conformity with a grammatical rule;
a display device for displaying a content and result of the processing of said character train;
a case dictionary for registering a semantic role between the vocabularies which are in conformity with the grammatical rule;
semantic analysis means for determining a semantic concept of the vocabulary in the output of said syntax analysis means by using said case dictionary;
and a vocabulary hierarchy dictionary for register-ing vocabularies by classifying the same from an upper order to a lower order in accordance with a semantic concept of each of the vocabularies, wherein when said inputted character train includes an undefined word, a semantic concept of said undefined word is obtained by said semantic analysis means, and a group of vocabularies corresponding to a lower concept of the semantic concept of said undefined word are extracted from said vocabulary hierarchy dictionary, and said extracted group of vocabularies of the lower concept are displayed on said display device so that a synonym of said undefined word is selected from said displayed group of vocabularies.
5. A natural language processing apparatus for processing a character train comprising:
morphemic analysis means for dividing an inputted character train of a natural language into vocabularies;
syntax analysis means for deciding whether or not a conjunction of the vocabularies is in conformity with a grammatical rule;
a case dictionary for registering a semantic role between the vocabularies which are in conformity with the grammatical rule;
semantic analysis means for determining a semantic concept of the vocabulary in the output of said syntax analysis means by using said case dictionary;
a vocabulary hierarchy dictionary for register-ing vocabularies by classifying the same from an upper order to a lower order in accordance with a semantic concept of each of the vocabularies; and inference means for selecting by inference a vocabulary which is in conformity with an intention of said inputted character train from said group of vocabularies registered as the lower concept in said vocabulary hierarchy dictionary, wherein when said inputted character train includes an undefined word, a semantic concept of said undefined word is obtained by said semantic analysis means, and a group of vocabularies corresponding to a lower concept of the semantic concept of said undefined word are extracted from said vocabulary hierarchy dictionary, and a synonym which is in conformity with the intention of said inputted character train is selected by said inference means from said extracted group of vocabularies of the lower concept so that said undefined word is replaced by said selected synonym.
morphemic analysis means for dividing an inputted character train of a natural language into vocabularies;
syntax analysis means for deciding whether or not a conjunction of the vocabularies is in conformity with a grammatical rule;
a case dictionary for registering a semantic role between the vocabularies which are in conformity with the grammatical rule;
semantic analysis means for determining a semantic concept of the vocabulary in the output of said syntax analysis means by using said case dictionary;
a vocabulary hierarchy dictionary for register-ing vocabularies by classifying the same from an upper order to a lower order in accordance with a semantic concept of each of the vocabularies; and inference means for selecting by inference a vocabulary which is in conformity with an intention of said inputted character train from said group of vocabularies registered as the lower concept in said vocabulary hierarchy dictionary, wherein when said inputted character train includes an undefined word, a semantic concept of said undefined word is obtained by said semantic analysis means, and a group of vocabularies corresponding to a lower concept of the semantic concept of said undefined word are extracted from said vocabulary hierarchy dictionary, and a synonym which is in conformity with the intention of said inputted character train is selected by said inference means from said extracted group of vocabularies of the lower concept so that said undefined word is replaced by said selected synonym.
6. An apparatus according to Claim 5, wherein an information terminal system is operated by using said inputted natural language.
7. A method for processing a natural language by extracting an undefined word from an input character train, comprising the steps of:
analyzing morphemes by dividing said input character train of the natural language into vocabularies, said input character train containing the undefined word not registered in a dictionary;
analyzing syntax by deciding whether or not a conjunction of the vocabularies is in conformity with a grammatical rule;
obtaining a semantic concept of said undefined word in the output obtained in said step of analyzing syntax, by using a case dictionary which registers a semantic role between the vocabularies; and extracting a group of vocabularies corresponding to a lower concept of the semantic concept of said undefined word by using a vocabulary hierarchy dictionary which registers vocabularies from an upper concept to a lower concept.
analyzing morphemes by dividing said input character train of the natural language into vocabularies, said input character train containing the undefined word not registered in a dictionary;
analyzing syntax by deciding whether or not a conjunction of the vocabularies is in conformity with a grammatical rule;
obtaining a semantic concept of said undefined word in the output obtained in said step of analyzing syntax, by using a case dictionary which registers a semantic role between the vocabularies; and extracting a group of vocabularies corresponding to a lower concept of the semantic concept of said undefined word by using a vocabulary hierarchy dictionary which registers vocabularies from an upper concept to a lower concept.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP59227251A JPS61105671A (en) | 1984-10-29 | 1984-10-29 | Natural language processing device |
| JP227251/84 | 1984-10-29 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CA1235817A true CA1235817A (en) | 1988-04-26 |
Family
ID=16857884
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CA000494041A Expired CA1235817A (en) | 1984-10-29 | 1985-10-28 | Method and apparatus for natural language processing |
Country Status (6)
| Country | Link |
|---|---|
| US (1) | US5109509A (en) |
| EP (1) | EP0180888B1 (en) |
| JP (1) | JPS61105671A (en) |
| KR (1) | KR920003498B1 (en) |
| CA (1) | CA1235817A (en) |
| DE (1) | DE3587993T2 (en) |
Families Citing this family (184)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5696916A (en) * | 1985-03-27 | 1997-12-09 | Hitachi, Ltd. | Information storage and retrieval system and display method therefor |
| DE3616011A1 (en) * | 1985-05-14 | 1986-11-20 | Sharp K.K., Osaka | TRANSLATION DEVICE |
| US4800522A (en) * | 1985-05-14 | 1989-01-24 | Sharp Kabushiki Kaisha | Bilingual translation system capable of memorizing learned words |
| US6443840B2 (en) * | 1986-03-10 | 2002-09-03 | Response Reward Systems, L.C. | Evaluation of responses of participatory broadcast audience with prediction of winning contestants; monitoring, checking and controlling of wagering, and automatic crediting and couponing |
| JP2654001B2 (en) * | 1986-05-08 | 1997-09-17 | 株式会社東芝 | Machine translation method |
| JPS62272359A (en) * | 1986-05-20 | 1987-11-26 | Sharp Corp | Mechanical translating device |
| JPS6310265A (en) * | 1986-06-30 | 1988-01-16 | Sharp Corp | Translating device |
| NL8702359A (en) * | 1986-10-03 | 1988-05-02 | Ricoh Kk | LANGUAGE ANALYSIS DEVICE. |
| US5225981A (en) * | 1986-10-03 | 1993-07-06 | Ricoh Company, Ltd. | Language analyzer for morphemically and syntactically analyzing natural languages by using block analysis and composite morphemes |
| GB2197510A (en) * | 1986-10-27 | 1988-05-18 | Sharp Kk | Translation apparatus |
| JPS63121977A (en) * | 1986-11-11 | 1988-05-26 | Fujitsu Ltd | Mechanical translating device |
| GB2199170A (en) * | 1986-11-28 | 1988-06-29 | Sharp Kk | Translation apparatus |
| JPH0827797B2 (en) * | 1986-12-01 | 1996-03-21 | 松下電器産業株式会社 | Parser |
| US4890230A (en) * | 1986-12-19 | 1989-12-26 | Electric Industry Co., Ltd. | Electronic dictionary |
| JPH0731677B2 (en) * | 1987-09-29 | 1995-04-10 | シャープ株式会社 | Document creation / proofreading support device |
| JPH01314373A (en) * | 1988-06-15 | 1989-12-19 | Hitachi Ltd | Translated word selecting system in machine translating system |
| US5111398A (en) * | 1988-11-21 | 1992-05-05 | Xerox Corporation | Processing natural language text using autonomous punctuational structure |
| US5276616A (en) * | 1989-10-16 | 1994-01-04 | Sharp Kabushiki Kaisha | Apparatus for automatically generating index |
| JPH03222065A (en) * | 1990-01-26 | 1991-10-01 | Sharp Corp | Machine translation device |
| US5255386A (en) * | 1990-02-08 | 1993-10-19 | International Business Machines Corporation | Method and apparatus for intelligent help that matches the semantic similarity of the inferred intent of query or command to a best-fit predefined command intent |
| FR2660085A1 (en) * | 1990-03-20 | 1991-09-27 | Philips Electronique Lab | DATA PROCESSING DEVICE AND METHOD FOR SELECTING DATA WORDS CONTAINED IN A DICTIONARY. |
| JP3266246B2 (en) * | 1990-06-15 | 2002-03-18 | インターナシヨナル・ビジネス・マシーンズ・コーポレーシヨン | Natural language analysis apparatus and method, and knowledge base construction method for natural language analysis |
| US5477451A (en) * | 1991-07-25 | 1995-12-19 | International Business Machines Corp. | Method and system for natural language translation |
| US5265065A (en) * | 1991-10-08 | 1993-11-23 | West Publishing Company | Method and apparatus for information retrieval from a database by replacing domain specific stemmed phases in a natural language to create a search query |
| US5488725A (en) * | 1991-10-08 | 1996-01-30 | West Publishing Company | System of document representation retrieval by successive iterated probability sampling |
| US5541836A (en) * | 1991-12-30 | 1996-07-30 | At&T Corp. | Word disambiguation apparatus and methods |
| US5523946A (en) * | 1992-02-11 | 1996-06-04 | Xerox Corporation | Compact encoding of multi-lingual translation dictionaries |
| US5268840A (en) * | 1992-04-30 | 1993-12-07 | Industrial Technology Research Institute | Method and system for morphologizing text |
| US5477447A (en) * | 1992-05-27 | 1995-12-19 | Apple Computer, Incorporated | Method and apparatus for providing computer-implemented assistance |
| US5390281A (en) * | 1992-05-27 | 1995-02-14 | Apple Computer, Inc. | Method and apparatus for deducing user intent and providing computer implemented services |
| US5434777A (en) * | 1992-05-27 | 1995-07-18 | Apple Computer, Inc. | Method and apparatus for processing natural language |
| JPH0728815A (en) * | 1993-02-18 | 1995-01-31 | Matsushita Electric Ind Co Ltd | Syntax analyzing device |
| US5448474A (en) * | 1993-03-03 | 1995-09-05 | International Business Machines Corporation | Method for isolation of Chinese words from connected Chinese text |
| US5359514A (en) * | 1993-08-03 | 1994-10-25 | International Business Machines Corporation | Method and apparatus for facilitating comprehension of on-line documents |
| US5523945A (en) * | 1993-09-17 | 1996-06-04 | Nec Corporation | Related information presentation method in document processing system |
| US5546573A (en) * | 1993-10-27 | 1996-08-13 | International Business Machines Corporation | Specification of cultural bias in database manager |
| CA2175187A1 (en) * | 1993-10-28 | 1995-05-04 | William K. Thomson | Database search summary with user determined characteristics |
| US5510981A (en) * | 1993-10-28 | 1996-04-23 | International Business Machines Corporation | Language translation apparatus and method using context-based translation models |
| US5692176A (en) * | 1993-11-22 | 1997-11-25 | Reed Elsevier Inc. | Associative text search and retrieval system |
| KR100304654B1 (en) * | 1993-11-30 | 2001-11-22 | 윤종용 | Method and apparatus for analyzing korean document |
| ZA948426B (en) * | 1993-12-22 | 1995-06-30 | Qualcomm Inc | Distributed voice recognition system |
| JPH07182465A (en) * | 1993-12-22 | 1995-07-21 | Hitachi Ltd | Character recognition method |
| US6594628B1 (en) | 1995-09-21 | 2003-07-15 | Qualcomm, Incorporated | Distributed voice recognition system |
| US5682538A (en) * | 1994-08-12 | 1997-10-28 | Wall Data Incorporated | Automatic adaptive computer screen generation |
| US6993471B1 (en) * | 1995-11-13 | 2006-01-31 | America Online, Inc. | Integrated multilingual browser |
| US5987404A (en) * | 1996-01-29 | 1999-11-16 | International Business Machines Corporation | Statistical natural language understanding using hidden clumpings |
| JP2987099B2 (en) * | 1996-03-27 | 1999-12-06 | 株式会社日立国際ビジネス | Document creation support system and term dictionary |
| US5878385A (en) * | 1996-09-16 | 1999-03-02 | Ergo Linguistic Technologies | Method and apparatus for universal parsing of language |
| US6144968A (en) * | 1997-03-04 | 2000-11-07 | Zellweger; Paul | Method and apparatus for menu access to information objects indexed by hierarchically-coded keywords |
| DE19740793C2 (en) * | 1997-09-17 | 2003-03-20 | Bosch Gmbh Robert | Process for coating surfaces by means of a system with sputter electrodes and use of the process |
| US6411926B1 (en) | 1999-02-08 | 2002-06-25 | Qualcomm Incorporated | Distributed voice recognition system |
| US6574596B2 (en) | 1999-02-08 | 2003-06-03 | Qualcomm Incorporated | Voice recognition rejection scheme |
| US6324509B1 (en) | 1999-02-08 | 2001-11-27 | Qualcomm Incorporated | Method and apparatus for accurate endpointing of speech in the presence of noise |
| WO2000062193A1 (en) | 1999-04-08 | 2000-10-19 | Kent Ridge Digital Labs | System for chinese tokenization and named entity recognition |
| US6401061B1 (en) * | 1999-05-13 | 2002-06-04 | Yuri L. Zieman | Combinatorial computational technique for transformation phrase text-phrase meaning |
| JP3266586B2 (en) * | 1999-07-07 | 2002-03-18 | インターナショナル・ビジネス・マシーンズ・コーポレーション | Data analysis system |
| JP2001125897A (en) * | 1999-10-28 | 2001-05-11 | Sony Corp | Device and method for learning language |
| US6834280B2 (en) | 2000-02-07 | 2004-12-21 | Josiah Lee Auspitz | Systems and methods for determining semiotic similarity between queries and database entries |
| US8645137B2 (en) | 2000-03-16 | 2014-02-04 | Apple Inc. | Fast, language-independent method for user authentication by voice |
| US7792676B2 (en) * | 2000-10-25 | 2010-09-07 | Robert Glenn Klinefelter | System, method, and apparatus for providing interpretive communication on a network |
| US6665642B2 (en) | 2000-11-29 | 2003-12-16 | Ibm Corporation | Transcoding system and method for improved access by users with special needs |
| US7013263B1 (en) | 2001-10-25 | 2006-03-14 | Mindfabric, Inc. | Online interaction processing |
| US6685188B2 (en) * | 2001-10-25 | 2004-02-03 | Rehbein Juerg | Method and apparatus for performing a transaction without the use of spoken communication between the transaction parties |
| US20030157470A1 (en) * | 2002-02-11 | 2003-08-21 | Michael Altenhofen | E-learning station and interface |
| US20040012643A1 (en) * | 2002-07-18 | 2004-01-22 | August Katherine G. | Systems and methods for visually communicating the meaning of information to the hearing impaired |
| US7349839B2 (en) * | 2002-08-27 | 2008-03-25 | Microsoft Corporation | Method and apparatus for aligning bilingual corpora |
| WO2005124599A2 (en) * | 2004-06-12 | 2005-12-29 | Getty Images, Inc. | Content search in complex language, such as japanese |
| KR100682897B1 (en) * | 2004-11-09 | 2007-02-15 | 삼성전자주식회사 | Method and apparatus for updating dictionary |
| US9002725B1 (en) | 2005-04-20 | 2015-04-07 | Google Inc. | System and method for targeting information based on message content |
| US8677377B2 (en) | 2005-09-08 | 2014-03-18 | Apple Inc. | Method and apparatus for building an intelligent automated assistant |
| US7949714B1 (en) * | 2005-12-05 | 2011-05-24 | Google Inc. | System and method for targeting advertisements or other information using user geographical information |
| US8601004B1 (en) * | 2005-12-06 | 2013-12-03 | Google Inc. | System and method for targeting information items based on popularities of the information items |
| US9318108B2 (en) | 2010-01-18 | 2016-04-19 | Apple Inc. | Intelligent automated assistant |
| US8977255B2 (en) | 2007-04-03 | 2015-03-10 | Apple Inc. | Method and system for operating a multi-function portable electronic device using voice-activation |
| US9251172B2 (en) * | 2007-06-01 | 2016-02-02 | Getty Images (Us), Inc. | Method and system for searching for digital assets |
| US8996433B2 (en) * | 2007-10-11 | 2015-03-31 | Steven Ginzberg | Automated natural language formula translator and data evaluator |
| US9330720B2 (en) | 2008-01-03 | 2016-05-03 | Apple Inc. | Methods and apparatus for altering audio output signals |
| US8996376B2 (en) | 2008-04-05 | 2015-03-31 | Apple Inc. | Intelligent text-to-speech conversion |
| US10496753B2 (en) | 2010-01-18 | 2019-12-03 | Apple Inc. | Automatically adapting user interfaces for hands-free interaction |
| US20100030549A1 (en) | 2008-07-31 | 2010-02-04 | Lee Michael M | Mobile device having human language translation capability with positional feedback |
| WO2010067118A1 (en) | 2008-12-11 | 2010-06-17 | Novauris Technologies Limited | Speech recognition involving a mobile device |
| US20120311585A1 (en) | 2011-06-03 | 2012-12-06 | Apple Inc. | Organizing task items that represent tasks to perform |
| US9858925B2 (en) | 2009-06-05 | 2018-01-02 | Apple Inc. | Using context information to facilitate processing of commands in a virtual assistant |
| US10241644B2 (en) | 2011-06-03 | 2019-03-26 | Apple Inc. | Actionable reminder entries |
| US10241752B2 (en) | 2011-09-30 | 2019-03-26 | Apple Inc. | Interface for a virtual digital assistant |
| US9431006B2 (en) | 2009-07-02 | 2016-08-30 | Apple Inc. | Methods and apparatuses for automatic speech recognition |
| US10276170B2 (en) | 2010-01-18 | 2019-04-30 | Apple Inc. | Intelligent automated assistant |
| US10705794B2 (en) | 2010-01-18 | 2020-07-07 | Apple Inc. | Automatically adapting user interfaces for hands-free interaction |
| US10679605B2 (en) | 2010-01-18 | 2020-06-09 | Apple Inc. | Hands-free list-reading by intelligent automated assistant |
| US10553209B2 (en) | 2010-01-18 | 2020-02-04 | Apple Inc. | Systems and methods for hands-free notification summaries |
| US8682667B2 (en) | 2010-02-25 | 2014-03-25 | Apple Inc. | User profiling for selecting user specific voice input processing information |
| US10762293B2 (en) | 2010-12-22 | 2020-09-01 | Apple Inc. | Using parts-of-speech tagging and named entity recognition for spelling correction |
| US9262612B2 (en) | 2011-03-21 | 2016-02-16 | Apple Inc. | Device access using voice authentication |
| US10057736B2 (en) | 2011-06-03 | 2018-08-21 | Apple Inc. | Active transport based notifications |
| US8994660B2 (en) | 2011-08-29 | 2015-03-31 | Apple Inc. | Text correction processing |
| US10134385B2 (en) | 2012-03-02 | 2018-11-20 | Apple Inc. | Systems and methods for name pronunciation |
| US9483461B2 (en) | 2012-03-06 | 2016-11-01 | Apple Inc. | Handling speech synthesis of content for multiple languages |
| US9280610B2 (en) | 2012-05-14 | 2016-03-08 | Apple Inc. | Crowd sourcing information to fulfill user requests |
| US9721563B2 (en) | 2012-06-08 | 2017-08-01 | Apple Inc. | Name recognition system |
| US9495129B2 (en) | 2012-06-29 | 2016-11-15 | Apple Inc. | Device, method, and user interface for voice-activated navigation and browsing of a document |
| US9576574B2 (en) | 2012-09-10 | 2017-02-21 | Apple Inc. | Context-sensitive handling of interruptions by intelligent digital assistant |
| US9547647B2 (en) | 2012-09-19 | 2017-01-17 | Apple Inc. | Voice-based media searching |
| US9530405B2 (en) * | 2012-11-30 | 2016-12-27 | Mitsubishi Electric Corporation | Intention estimating device and intention estimating method |
| KR102516577B1 (en) | 2013-02-07 | 2023-04-03 | 애플 인크. | Voice trigger for a digital assistant |
| US9368114B2 (en) | 2013-03-14 | 2016-06-14 | Apple Inc. | Context-sensitive handling of interruptions |
| WO2014144949A2 (en) | 2013-03-15 | 2014-09-18 | Apple Inc. | Training an at least partial voice command system |
| WO2014144579A1 (en) | 2013-03-15 | 2014-09-18 | Apple Inc. | System and method for updating an adaptive speech recognition model |
| US9582608B2 (en) | 2013-06-07 | 2017-02-28 | Apple Inc. | Unified ranking with entropy-weighted information for phrase-based semantic auto-completion |
| WO2014197336A1 (en) | 2013-06-07 | 2014-12-11 | Apple Inc. | System and method for detecting errors in interactions with a voice-based digital assistant |
| WO2014197334A2 (en) | 2013-06-07 | 2014-12-11 | Apple Inc. | System and method for user-specified pronunciation of words for speech synthesis and recognition |
| WO2014197335A1 (en) | 2013-06-08 | 2014-12-11 | Apple Inc. | Interpreting and acting upon commands that involve sharing information with remote devices |
| EP3008641A1 (en) | 2013-06-09 | 2016-04-20 | Apple Inc. | Device, method, and graphical user interface for enabling conversation persistence across two or more instances of a digital assistant |
| US10176167B2 (en) | 2013-06-09 | 2019-01-08 | Apple Inc. | System and method for inferring user intent from speech inputs |
| CN105265005B (en) | 2013-06-13 | 2019-09-17 | 苹果公司 | System and method for the urgent call initiated by voice command |
| WO2015020942A1 (en) | 2013-08-06 | 2015-02-12 | Apple Inc. | Auto-activating smart responses based on activities from remote devices |
| US9620105B2 (en) | 2014-05-15 | 2017-04-11 | Apple Inc. | Analyzing audio input for efficient speech and music recognition |
| US10592095B2 (en) | 2014-05-23 | 2020-03-17 | Apple Inc. | Instantaneous speaking of content on touch devices |
| US9502031B2 (en) | 2014-05-27 | 2016-11-22 | Apple Inc. | Method for supporting dynamic grammars in WFST-based ASR |
| US10078631B2 (en) | 2014-05-30 | 2018-09-18 | Apple Inc. | Entropy-guided text prediction using combined word and character n-gram language models |
| US9633004B2 (en) | 2014-05-30 | 2017-04-25 | Apple Inc. | Better resolution when referencing to concepts |
| US10289433B2 (en) | 2014-05-30 | 2019-05-14 | Apple Inc. | Domain specific language for encoding assistant dialog |
| US9430463B2 (en) | 2014-05-30 | 2016-08-30 | Apple Inc. | Exemplar-based natural language processing |
| US9760559B2 (en) | 2014-05-30 | 2017-09-12 | Apple Inc. | Predictive text input |
| US9842101B2 (en) | 2014-05-30 | 2017-12-12 | Apple Inc. | Predictive conversion of language input |
| US9715875B2 (en) | 2014-05-30 | 2017-07-25 | Apple Inc. | Reducing the need for manual start/end-pointing and trigger phrases |
| US10170123B2 (en) | 2014-05-30 | 2019-01-01 | Apple Inc. | Intelligent assistant for home automation |
| US9785630B2 (en) | 2014-05-30 | 2017-10-10 | Apple Inc. | Text prediction using combined word N-gram and unigram language models |
| EP3149728B1 (en) | 2014-05-30 | 2019-01-16 | Apple Inc. | Multi-command single utterance input method |
| US9734193B2 (en) | 2014-05-30 | 2017-08-15 | Apple Inc. | Determining domain salience ranking from ambiguous words in natural speech |
| US10659851B2 (en) | 2014-06-30 | 2020-05-19 | Apple Inc. | Real-time digital assistant knowledge updates |
| US9338493B2 (en) | 2014-06-30 | 2016-05-10 | Apple Inc. | Intelligent automated assistant for TV user interactions |
| US10446141B2 (en) | 2014-08-28 | 2019-10-15 | Apple Inc. | Automatic speech recognition based on user feedback |
| US9818400B2 (en) | 2014-09-11 | 2017-11-14 | Apple Inc. | Method and apparatus for discovering trending terms in speech requests |
| US10789041B2 (en) | 2014-09-12 | 2020-09-29 | Apple Inc. | Dynamic thresholds for always listening speech trigger |
| US9606986B2 (en) | 2014-09-29 | 2017-03-28 | Apple Inc. | Integrated word N-gram and class M-gram language models |
| US9886432B2 (en) | 2014-09-30 | 2018-02-06 | Apple Inc. | Parsimonious handling of word inflection via categorical stem + suffix N-gram language models |
| US10074360B2 (en) | 2014-09-30 | 2018-09-11 | Apple Inc. | Providing an indication of the suitability of speech recognition |
| US9646609B2 (en) | 2014-09-30 | 2017-05-09 | Apple Inc. | Caching apparatus for serving phonetic pronunciations |
| US10127911B2 (en) | 2014-09-30 | 2018-11-13 | Apple Inc. | Speaker identification and unsupervised speaker adaptation techniques |
| US9668121B2 (en) | 2014-09-30 | 2017-05-30 | Apple Inc. | Social reminders |
| US10552013B2 (en) | 2014-12-02 | 2020-02-04 | Apple Inc. | Data detection |
| US9711141B2 (en) | 2014-12-09 | 2017-07-18 | Apple Inc. | Disambiguating heteronyms in speech synthesis |
| CN104573028B (en) * | 2015-01-14 | 2019-01-25 | 百度在线网络技术（北京）有限公司 | Realize the method and system of intelligent answer |
| US9865280B2 (en) | 2015-03-06 | 2018-01-09 | Apple Inc. | Structured dictation using intelligent automated assistants |
| US10567477B2 (en) | 2015-03-08 | 2020-02-18 | Apple Inc. | Virtual assistant continuity |
| US9721566B2 (en) | 2015-03-08 | 2017-08-01 | Apple Inc. | Competing devices responding to voice triggers |
| US9886953B2 (en) | 2015-03-08 | 2018-02-06 | Apple Inc. | Virtual assistant activation |
| US9899019B2 (en) | 2015-03-18 | 2018-02-20 | Apple Inc. | Systems and methods for structured stem and suffix language models |
| US9842105B2 (en) | 2015-04-16 | 2017-12-12 | Apple Inc. | Parsimonious continuous-space phrase representations for natural language processing |
| US10083688B2 (en) | 2015-05-27 | 2018-09-25 | Apple Inc. | Device voice control for selecting a displayed affordance |
| US10127220B2 (en) | 2015-06-04 | 2018-11-13 | Apple Inc. | Language identification from short strings |
| US10101822B2 (en) | 2015-06-05 | 2018-10-16 | Apple Inc. | Language input correction |
| US11025565B2 (en) | 2015-06-07 | 2021-06-01 | Apple Inc. | Personalized prediction of responses for instant messaging |
| US10186254B2 (en) | 2015-06-07 | 2019-01-22 | Apple Inc. | Context-based endpoint detection |
| US10255907B2 (en) | 2015-06-07 | 2019-04-09 | Apple Inc. | Automatic accent detection using acoustic models |
| US10671428B2 (en) | 2015-09-08 | 2020-06-02 | Apple Inc. | Distributed personal assistant |
| US10747498B2 (en) | 2015-09-08 | 2020-08-18 | Apple Inc. | Zero latency digital assistant |
| US9697820B2 (en) | 2015-09-24 | 2017-07-04 | Apple Inc. | Unit-selection text-to-speech synthesis using concatenation-sensitive neural networks |
| US11010550B2 (en) | 2015-09-29 | 2021-05-18 | Apple Inc. | Unified language modeling framework for word prediction, auto-completion and auto-correction |
| US10366158B2 (en) | 2015-09-29 | 2019-07-30 | Apple Inc. | Efficient word encoding for recurrent neural network language models |
| US11587559B2 (en) | 2015-09-30 | 2023-02-21 | Apple Inc. | Intelligent device identification |
| US10691473B2 (en) | 2015-11-06 | 2020-06-23 | Apple Inc. | Intelligent automated assistant in a messaging environment |
| US10049668B2 (en) | 2015-12-02 | 2018-08-14 | Apple Inc. | Applying neural network language models to weighted finite state transducers for automatic speech recognition |
| US10223066B2 (en) | 2015-12-23 | 2019-03-05 | Apple Inc. | Proactive assistance based on dialog communication between devices |
| US10446143B2 (en) | 2016-03-14 | 2019-10-15 | Apple Inc. | Identification of voice inputs providing credentials |
| US9934775B2 (en) | 2016-05-26 | 2018-04-03 | Apple Inc. | Unit-selection text-to-speech synthesis based on predicted concatenation parameters |
| US9972304B2 (en) | 2016-06-03 | 2018-05-15 | Apple Inc. | Privacy preserving distributed evaluation framework for embedded personalized systems |
| US10249300B2 (en) | 2016-06-06 | 2019-04-02 | Apple Inc. | Intelligent list reading |
| US10191899B2 (en) | 2016-06-06 | 2019-01-29 | Comigo Ltd. | System and method for understanding text using a translation of the text |
| US10049663B2 (en) | 2016-06-08 | 2018-08-14 | Apple, Inc. | Intelligent automated assistant for media exploration |
| DK179309B1 (en) | 2016-06-09 | 2018-04-23 | Apple Inc | Intelligent automated assistant in a home environment |
| US10067938B2 (en) | 2016-06-10 | 2018-09-04 | Apple Inc. | Multilingual word prediction |
| US10490187B2 (en) | 2016-06-10 | 2019-11-26 | Apple Inc. | Digital assistant providing automated status report |
| US10586535B2 (en) | 2016-06-10 | 2020-03-10 | Apple Inc. | Intelligent digital assistant in a multi-tasking environment |
| US10509862B2 (en) | 2016-06-10 | 2019-12-17 | Apple Inc. | Dynamic phrase expansion of language input |
| US10192552B2 (en) | 2016-06-10 | 2019-01-29 | Apple Inc. | Digital assistant providing whispered speech |
| DK201670540A1 (en) | 2016-06-11 | 2018-01-08 | Apple Inc | Application integration with a digital assistant |
| DK179049B1 (en) | 2016-06-11 | 2017-09-18 | Apple Inc | Data driven natural language event detection and classification |
| DK179343B1 (en) | 2016-06-11 | 2018-05-14 | Apple Inc | Intelligent task discovery |
| DK179415B1 (en) | 2016-06-11 | 2018-06-14 | Apple Inc | Intelligent device arbitration and control |
| US10593346B2 (en) | 2016-12-22 | 2020-03-17 | Apple Inc. | Rank-reduced token representation for automatic speech recognition |
| KR102375800B1 (en) * | 2017-04-28 | 2022-03-17 | 삼성전자주식회사 | electronic device providing speech recognition service and method thereof |
| DK179745B1 (en) | 2017-05-12 | 2019-05-01 | Apple Inc. | SYNCHRONIZATION AND TASK DELEGATION OF A DIGITAL ASSISTANT |
| DK201770431A1 (en) | 2017-05-15 | 2018-12-20 | Apple Inc. | Optimizing dialogue policy decisions for digital assistants using implicit feedback |
Family Cites Families (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4193119A (en) * | 1977-03-25 | 1980-03-11 | Xerox Corporation | Apparatus for assisting in the transposition of foreign language text |
| GB1596411A (en) * | 1977-09-19 | 1981-08-26 | Shadin Hojin Kyodo Tsushinsha | Translation system |
| US4164025A (en) * | 1977-12-13 | 1979-08-07 | Bell Telephone Laboratories, Incorporated | Spelled word input directory information retrieval system with input word error corrective searching |
| CA1129194A (en) * | 1978-02-17 | 1982-08-10 | Jung-Eui Son | Portable word translator device |
| JPS5936795B2 (en) * | 1979-08-17 | 1984-09-05 | シャープ株式会社 | electronic dictionary |
| US4623985A (en) * | 1980-04-15 | 1986-11-18 | Sharp Kabushiki Kaisha | Language translator with circuitry for detecting and holding words not stored in dictionary ROM |
| JPS5748159A (en) * | 1980-09-03 | 1982-03-19 | Sharp Corp | Electronic interpreter |
| JPS5924452B2 (en) * | 1980-09-22 | 1984-06-09 | 株式会社日立製作所 | Japanese text input device |
| JPS57201958A (en) * | 1981-06-05 | 1982-12-10 | Hitachi Ltd | Device and method for interpretation between natural languages |
| JPS57201926A (en) * | 1981-06-05 | 1982-12-10 | Hitachi Ltd | "kanji" selecting method for "kanji" input device |
| JPS5840684A (en) * | 1981-09-04 | 1983-03-09 | Hitachi Ltd | Automatic translating system between natural languages |
| JPS58175074A (en) * | 1982-04-07 | 1983-10-14 | Toshiba Corp | Analyzing system of sentence structure |
| JPS58201175A (en) * | 1982-05-20 | 1983-11-22 | Kokusai Denshin Denwa Co Ltd <Kdd> | Machine translation system |
| US4468756A (en) * | 1982-05-20 | 1984-08-28 | Chan Computer Corporation | Method and apparatus for processing languages |
| US4544276A (en) * | 1983-03-21 | 1985-10-01 | Cornell Research Foundation, Inc. | Method and apparatus for typing Japanese text using multiple systems |
| JPS6140672A (en) * | 1984-07-31 | 1986-02-26 | Hitachi Ltd | Processing system for dissolution of many parts of speech |
-
1984
- 1984-10-29 JP JP59227251A patent/JPS61105671A/en active Granted
-
1985
- 1985-10-28 DE DE3587993T patent/DE3587993T2/en not_active Expired - Fee Related
- 1985-10-28 KR KR1019850007969A patent/KR920003498B1/en not_active IP Right Cessation
- 1985-10-28 EP EP85113695A patent/EP0180888B1/en not_active Expired - Lifetime
- 1985-10-28 CA CA000494041A patent/CA1235817A/en not_active Expired
-
1988
- 1988-11-28 US US07/277,526 patent/US5109509A/en not_active Expired - Fee Related
Also Published As
| Publication number | Publication date |
|---|---|
| US5109509A (en) | 1992-04-28 |
| EP0180888B1 (en) | 1995-02-22 |
| EP0180888A3 (en) | 1986-08-27 |
| JPS61105671A (en) | 1986-05-23 |
| EP0180888A2 (en) | 1986-05-14 |
| JPH0510703B2 (en) | 1993-02-10 |
| KR920003498B1 (en) | 1992-05-01 |
| DE3587993D1 (en) | 1995-03-30 |
| DE3587993T2 (en) | 1995-06-14 |
| KR860003560A (en) | 1986-05-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CA1235817A (en) | Method and apparatus for natural language processing | |
| US5225981A (en) | Language analyzer for morphemically and syntactically analyzing natural languages by using block analysis and composite morphemes | |
| CN107209759B (en) | Annotation support device and recording medium | |
| US5579224A (en) | Dictionary creation supporting system | |
| JP3717730B2 (en) | Electronic dictionary | |
| JP2002215617A (en) | Method for attaching part of speech tag | |
| JPH083815B2 (en) | Natural language co-occurrence relation dictionary maintenance method | |
| JPH0351020B2 (en) | ||
| KR20040111715A (en) | A Method for Synthesizing a Self Learning System for Extraction of Knowledge from Textual Documents for Use in Search Systems | |
| JPH06195373A (en) | Machine translation system | |
| JPH05314166A (en) | Electronic dictionary and dictionary retrieval device | |
| JPH08292955A (en) | Language processing method and data processor applying the same | |
| JPH08235182A (en) | Method and device for document processing | |
| JPH05120324A (en) | Language processing system | |
| JPS63228326A (en) | Automatic key word extracting system | |
| JPH0561902A (en) | Mechanical translation system | |
| JPS62267872A (en) | Language analyzing device | |
| JPH09245051A (en) | Device and method for retrieving natural language instance | |
| JP3949874B2 (en) | Translation translation learning method, translation translation learning device, storage medium, and translation system | |
| JPH0816910B2 (en) | Language analyzer | |
| JP2002183134A (en) | Translating device | |
| JP2880600B2 (en) | Word string translator | |
| Branco et al. | EtiFac: A facilitating tool for manual tagging | |
| JPH0724057B2 (en) | Electronics | |
| JPH1063665A (en) | Method and device for translating text |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| MKEX | Expiry |