JP2003281131A - System and method of processing natural language, and computer program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To handle the conjugation information of a conjugational word without a special process for reference to the conjugated word. <P>SOLUTION: A morpheme including the conjugation information is extracted. Referring to a list of words without requiring insertion of conjugations including specific auxiliary verbs and verbs, a morpheme just after the morpheme with the conjugation information is inserted if the morpheme is included in the list, and the conjugation information is inserted into the result of the morpheme analysis as a character string corresponding to a word. In the stage of the meaning analysis, the difference of the meaning according to the conjugation of the conjugational word is managed as a grammar rule. The conjugation information inserted into the result of the morpheme analysis is referred to according to the grammar rule, so that the meaning of the conjugation can be specified. <P>COPYRIGHT: (C)2004,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a natural language processing system and a natural language processing method for mathematically handling a natural language used by humans for daily communication.
The present invention also relates to a computer program, and more particularly to a natural language processing system and a natural language processing method for performing syntactic and semantic analysis of Japanese syntax, and a computer program.

More specifically, the present invention relates to a natural language processing system and a natural language processing method for performing a semantic analysis according to the conjugation form of a conjugation word in a Japanese sentence, and a computer program. The present invention relates to a natural language processing system, a natural language processing method, and a computer program that handle utilization information of a utilized word without special processing for processing.

[0003]

2. Description of the Related Art Words used by humans for daily communication such as Japanese and English are called "natural language". Many natural languages have a naturally occurring origin and have evolved with the history of humanity, ethnicity, and society. Of course, a person can communicate by gesturing or gesturing, but natural language can realize the most natural and advanced communication.

On the other hand, with the development of information technology, computers have become established in human society and have been deeply permeated in various industries and daily lives. Now, not only computer data but almost all information contents such as images and sounds are handled on the computer, and it is possible to perform advanced processing such as editing / processing, storage, management, transmission and sharing of information. .

Natural language has an abstract and ambiguous nature by nature, but by treating sentences mathematically,
Computer processing can be performed. As a result, various applications related to natural language such as machine translation, dialogue system, search system, etc.
The service is realized.

Natural language processing is generally divided into processing phases of morphological analysis, syntactic analysis, semantic analysis, and context analysis.

In the morphological analysis, a sentence is segmented into morphemes, which are the smallest semantic units, and a part-of-speech recognition process is performed. In the syntactic analysis, a sentence structure such as a phrase structure is analyzed based on grammatical rules. Since the grammar rules have a tree structure,
The syntactic analysis result generally has a tree structure in which individual morphemes are joined based on the dependency relationship. In the semantic analysis, based on the meaning (concept) of the words in the sentence and the semantic relationship between the words, the semantic structure expressing the meaning conveyed by the sentence is obtained, and the semantic structures are synthesized. In context analysis, a sentence (discourse), which is a series of sentences, is regarded as a basic unit of analysis, and a discourse structure is constructed by obtaining a semantic unity between sentences.

In the syntactic and semantic analysis, the relation between the verb and the other constituent elements in the sentence such as the subject (ie, the case frame of the predicate) is described for the structural sentence after the dependency relation is obtained by the syntactic analysis or the like. It is performed that the semantic relation between the predicate and the related words is extracted using the valence index dictionary.

[0009]

Among the many languages, Japanese is said to be highly ambiguous and difficult to process in natural language. One reason for this is that in Japanese sentences, the meaning interpretation differs depending on the conjugation form of the conjugation words such as verbs and adjectives.

For example, in the case of considering two similar sentence examples as shown below, in the sentence example (1), the suffix "so" means "hearing", but in the other sentence example (2), the suffix "Yes" represents guess.

(1) He writes a book. (2) He is likely to write a book.

[0012] These differences are caused by the different conjugations of the verbs associated with the suffix "so". In the above example (1), the verb "writing" is related to "so" in adnominal form, whereas in example (2) the verb "writing" is related to continuous.

In the case of the following two sentence examples, the inflectional form of the inflectional word may have a function corresponding to a specific particle.

(3) He went to the river and she went to the mountain. (4) He went to the river and she went to the mountain.

In the above sentence example (3), the connecting particle "te" connects two phrases "he goes to the river" and "she goes to the mountain". On the other hand, in the sentence example (d), the phrase is connected by the continuous use of “go” instead of the connective particle.

As described above, in the case of performing the semantic analysis process of a Japanese sentence, the conjugation form of the conjugation word is very important information.

However, in the conventional Japanese parsing system that inputs only the headword of the morphological analysis result,
Such utilization information is lost from the input character string.
In order to handle the difference in meaning as in the above example, it is necessary to return to the stage of morphological analysis and acquire utilization information.

Further, in another conventional technique, "go + continuous use type"
Inflectional forms are connected as features after headwords such as. However, when this method is used, when the usage information plays a role equivalent to a specific part of speech, a rule different from the rule for handling the part of speech is described when performing syntactic analysis or semantic analysis as post-processing. Or you have to do something else.

In short, in the conventional technique, a process different from the grammar rule or a special grammar rule is required to handle the conjugation form of the conjugation word, so that a large cost is generated.

An object of the present invention is to provide an excellent natural language processing system, natural language processing method, and computer program capable of performing a semantic analysis according to the inflectional form of a Japanese sentence. Especially.

A further object of the present invention is to provide an excellent natural language processing system, natural language processing method, and computer program capable of handling utilization information of utilization words without special processing for referring to utilization words. To provide.

[0022]

The present invention has been made in consideration of the above problems, and a first aspect thereof is a natural language processing system or a natural language processing method for handling utilization information of a utilized word. In addition, it is determined whether or not it is necessary to insert utilization information into the extracted utilization word or a utilization word extracting means or step for extracting a morpheme that includes utilization information from the morphemes forming the input sentence. A natural language processing system comprising: utilization information judging means or step; and utilization information inserting means or step for inserting utilization information into an input sentence according to the judgment result by the utilization information judging means or step. It is a natural language processing method.

Here, the utilization information determining means or step refers to the inflectional form insertion unnecessary vocabulary list including a specific auxiliary verb or verb, and the morpheme immediately following the morpheme having the utilization form information is included in the list. If it is, insert it; otherwise, decide not to insert it.

The utilization information inserting means or step may insert the utilization information into the morphological analysis result as a character string corresponding to a word.

Therefore, according to the natural language processing system or the natural language processing method according to the first aspect of the present invention, the information of the utilized word is inserted into the morphological analysis result as a character string corresponding to the word, so that the semantic analysis in the latter stage is performed. In, the difference in meaning can be easily discriminated only by the description of the grammar rule without performing a special process for referring to the utilization information.

That is, the difference in meaning depending on the usage form of the usage word is managed as a grammar rule. For example, by using a context-free grammar, it can be easily handled as a rule. Then, by referring to the usage information inserted in the morpheme analysis result by such a grammar rule, the meaning of the usage form can be specified.

Since the utilization information of the utilization word is inserted in the morphological analysis result as a character string corresponding to the word, it is not necessary to change the mechanism of the semantic analysis. Further, since no special grammar rule is required for the semantic analysis due to the difference in the usage form, the cost does not increase.

A second aspect of the present invention is a computer program written in a computer-readable format so as to execute natural language processing for handling utilization information of a utilization word on a computer system. And a utilization information determining step of determining whether or not it is necessary to insert utilization information into the extracted utilization word, and And a utilization information insertion step of inserting utilization information into an input sentence in accordance with the result of the determination in the information determination step.

A computer according to the second aspect of the present invention
The program defines a computer program written in a computer-readable format so as to realize a predetermined process on a computer system. In other words, by installing the computer program according to the second aspect of the present invention in the computer system, a cooperative action is exerted on the computer system, and the natural language according to the first aspect of the present invention. It is possible to obtain the same effects as those of the processing system or the natural language processing method.

Still other objects, features and advantages of the present invention are as follows.
It will be apparent from the embodiments of the present invention described later and the more detailed description based on the accompanying drawings.

[0031]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below in detail with reference to the drawings.

Natural language syntax analysis methods can be roughly classified into methods based on statistical processing and methods based on grammar rule description. The present invention can exert a remarkable effect particularly when applied to syntactic / semantic analysis based on grammar rule description.

The natural language processing system according to the present invention can be implemented, for example, by incorporating it into a syntactic / semantic analysis process based on the LFG (Lexical-Functional Grammar) grammar theory. In LFG, the linguistic knowledge, or grammar, of native speakers is organized as a component separate from computer processing and other non-grammatical processing parameters that affect the processing behavior of the computer. First, an overview of the natural language processing system will be briefly described. Although the present embodiment is described based on the LFG grammar theory, the present invention can be similarly applied to an analysis system having other grammar rules.

FIG. 10 schematically shows the configuration of the LFG-based natural language processing system 1.

The morpheme analysis unit 2 has a morpheme rule 2A and a morpheme dictionary 2B related to a specific language such as Japanese, and divides an input sentence into morphemes, which are the smallest semantic units, and performs part-of-speech recognition processing. For example, if the sentence "My daughter speaks English." Is entered, the result of morphological analysis is "I {Noun} {up} daughter {Noun} is {up} English {Noun}.
{Up} speak {Verb1} {tr} and {jp}. {pt} "is output.

The result of such morphological analysis is then input to the syntactic / semantic analysis unit 3. The syntactic and semantic analysis department
It has dictionaries such as grammar rules 3A and valence valence dictionary 3B, and analyzes the phrase structure based on grammar rules and expresses the meaning conveyed by the sentence based on the meaning of words in the sentence and the meaning relation between words. Analyze the semantic structure (The valence dictionary describes the relationship between verbs and other constituent elements in the sentence, such as the subject, and can extract the semantic relationship between predicates and related words).

Then, as a result of the syntactic analysis, "c-structure (constituent structur) which represents a phrase structure of a sentence composed of words and morphemes as a tree structure.
e) ", and as a result of semantically and functionally analyzing the input sentence based on the case structure of the subject, object, etc., such as interrogative sentence, past tense, and polite sentence," f-structure (functional stru
cture) "is output.

11 and 12, c-structure and f-obtained as a result of processing the input sentence "My daughter speaks English" by the syntactic / semantic analysis unit 1.
Each structure is shown.

The c-structure is a tree structure representing the structure of words and phrases in a sentence, and is defined by the syntax category. For example, a phonological interpretation for generating a phoneme string can be performed based on c-structure. On the other hand, f-structure is
It is a clear expression of grammatical functions and is composed of grammatical function names, semantic forms, and feature symbols. By referring to f-structure, the subject, the object, and the complement
t) and modifiers (adjunct) can be understood. f-structure is c-structure
It is a set of features attached to each node of ure, and is expressed in the form of an attribute-attribute value matrix as shown in FIG. In other words, the left side in [] is the feature (attribute)
And the right side is the feature value (attribute value).

Regarding the details of LFG, see R.
Paper "Lexical-Functio" co-authored by M. Kaplan and J. Bresnan
nal Grammar: A Formal System for Grammatical Repre
sentation "(The MIT Press, Cambridge (1982). Repr
inted in Formal Issues in Lexical-Functional Gramma
r, pp. 29-130.CSLI publications, Stanford Univers
ity (1995).).

As described above, in Japanese sentences, the meaning interpretation is different depending on the inflectional form of the inflectional word, so that the inflectional form of the inflectional word is very important information when performing the semantic analysis of the Japanese sentence. .

FIG. 1 shows, in the form of a flow chart, a processing procedure for handling a conjugation form of a conjugation word according to an embodiment of the present invention. In the following, it is assumed that "He writes a book." Is input.

First, morphological analysis is performed on the input sentence. As a result, the output result as shown in Table 1 below can be obtained (step S1).

[0044]

[Table 1]

Next, from the output result of the morphological analysis, headwords delimited by delimiters such as blanks are stored in the buffer as a character string (step S2). Here, the character string "He writes a book" is stored.

Then, it is judged whether or not the output result includes the inflectional information (step S3). In this example, the condition is true because the conjugation form information of the verb "writing" includes "conjunctive form", and the conjugation form information of the auxiliary verb "da" includes "termination form."

Then, if the information on the inflectional form is included, it is further judged whether or not the inflectional form should be inserted in the output result (step S4). In this judgment step, the morphological insertion unnecessary vocabulary list is referred to, and if the morpheme immediately following the morpheme having the conjugation information is included in the list, it is judged that the morpheme is inserted.

The inflectional insertion unnecessary vocabulary list includes specific auxiliary verbs and verbs. In this case, "yes" is not included in the list, but "." Is included, and "write"
It is judged that the utilization information of should be inserted.

Next, the inquired form-character string correspondence table as shown in Table 2 below is inquired for a character string corresponding to the "continuously used form" (step S5). Here, since the character string corresponding to the “continuously used form” is RY, the character string RY is acquired.

[0050]

[Table 2]

Then, the acquired character string is inserted after the conjugation word included in the character string stored in the buffer (step S6). Here, since the utilization word is "write", RY as utilization information is inserted after "writing". Therefore,
The string stored in the buffer becomes "RY he reads the book, yes."

Finally, the character string in the buffer is output as a result (step S7). In this way, by inserting the information of the inflection word into the morpheme analysis result as a character string corresponding to the word, the description of the grammar rule 3A can be performed in the subsequent semantic analysis without performing a special process for referring to the inflection information. The difference in meaning can be easily discriminated by only

Next, the processing for handling the case in which the difference in meaning occurs due to the difference in usage type will be described.

FIG. 2 shows a simple rule description by a free context grammar for handling the difference in meaning of the suffix "so". When this rule is applied, it is possible to determine that the input sentence “RY he writes a book, yes,” that is obtained as an output result is a “guess”.

Next, the processing when the inflectional form of the inflectional word corresponds to a specific particle will be described.

FIG. 3 shows a rule description for handling a compound sentence in which two sentences are connected by a connective particle. If you apply this rule, as shown in FIG. 4, for the statement that "he went to the river, she went to the mountain.", "He goes to the river." Sentence and "her mountain It is possible to obtain the analysis result that the sentence "haeta" is connected by the connecting particle "te".

By the processing procedure shown in FIG. 1, "he went to the mountain and she went to the river" is converted into the character string "RY he went to the mountain RY she went to the river." . When the rule shown in FIG. 3 is applied to this character string, the result shown in FIG.
The result as shown in can be obtained.

As shown in Table 1, according to the present invention, the verb usage information in the input sentence is inserted in the morpheme analysis result as a character string corresponding to a word. Therefore, no special processing is required to refer to the utilization information in the verb. Further, the difference in meaning due to the difference in inflectional form in the verb can be described as a simple rule by the free context grammar as shown in FIGS. Further, as can be seen by comparing FIGS. 4 and 5, it is possible to analyze the difference in meaning according to the difference in the inflectional form of the verb according to the normal procedure of referring to the grammar rule.

In the above description, the handling of the utilization information regarding the verb as the utilization word has been described by taking "he seems to write a book" as an example. Next, the handling of utilization information in the case of adjectives will be described below.

FIG. 6 shows a rule in which an adjective becomes an adverb when the adjective appears in the plural form.

When the input sentence "She read a book in detail" is processed in the same manner according to the processing procedure shown in FIG. 1, it is converted into the character string "She read a detailed RY book." . When the rule of FIG. 6 is applied to this character string, the result as shown in FIG. 7 can be obtained.

Further, FIG. 8 shows a rule for handling a case in which an adjective appears in the adnominal form and is associated with a noun.

When the input sentence "She read a book in detail." Is processed in the same manner according to the processing procedure shown in FIG. 1, it is converted into the character string "She read a detailed RT book." . When the rule shown in FIG. 8 is applied to this character string, the result shown in FIG. 9 can be obtained.

As described above, according to the present invention, the utilization information of the adjective in the input sentence is inserted as a character string corresponding to a word in the morphological analysis result. Therefore, no special processing is required to refer to the utilization information in the adjective. Further, the difference in meaning due to the difference in inflectional form of adjectives can be described as a simple rule by the free context grammar as shown in FIGS. 6 and 8. And FIG.
Also, as can be seen by comparing FIG. 9 and FIG. 9, it is possible to analyze the difference in meaning according to the difference in the inflectional form of the verb according to the normal procedure of referring to the grammar rule.

[Supplement] The present invention has been described in detail with reference to the specific embodiments. However, it is obvious that those skilled in the art can modify or substitute the embodiments without departing from the scope of the present invention. That is, the present invention has been disclosed in the form of exemplification, and the contents of this specification should not be construed in a limited manner. In order to determine the gist of the present invention, the section of the claims described at the beginning should be taken into consideration.

[0066]

As described above in detail, according to the present invention,
It is possible to provide an excellent natural language processing system, natural language processing method, and computer program capable of performing a semantic analysis according to a usage form of a Japanese sentence.

Further, according to the present invention, an excellent natural language processing system, natural language processing method, and computer program capable of handling utilization information of utilization words without special processing for referencing utilization words. To provide.

In the present invention, the information on the inflection word is inserted into the morphological analysis result as a character string corresponding to the word. by this,
No special process is required to refer to the utilization information, and the difference in meaning can be easily identified only by describing the grammar rules.

[Brief description of drawings]

FIG. 1 is a flowchart showing a processing procedure for handling an inflectional form of an inflection word according to an embodiment of the present invention.

FIG. 2 is a diagram showing a simple rule description by a free context grammar for handling a difference in meaning of the suffix “so” in an input sentence.

FIG. 3 is a diagram showing a rule description for handling a compound sentence in which two sentences are connected by a connection particle.

FIG. 4 is a diagram showing a result of applying the rule shown in FIG. 3 to an input sentence.

FIG. 5 is a diagram showing a result of applying the rule shown in FIG. 3 to an input sentence.

FIG. 6 is a diagram showing a rule that becomes an adverb when an adjective appears in a plural form.

FIG. 7 is a diagram showing a result of applying the rule shown in FIG. 6 to a character string “She read a detailed RY book”.

FIG. 8 is a diagram showing a rule for handling a case in which an adjective appears in an adnominal form and is associated with a noun.

FIG. 9 is a diagram showing a result of applying the rule shown in FIG. 8 to the character string “She read a detailed RT book”.

FIG. 10 is a diagram schematically showing a configuration of a natural language processing system 1 based on LFG.

[Fig. 11] c- obtained as a result of processing the input sentence "My daughter speaks English" by the syntactic and semantic analysis unit 1.
It is the figure which showed structure.

[Fig. 12] f-obtained as a result of processing the input sentence "My daughter speaks English" by the syntactic and semantic analysis unit 1
It is the figure which showed structure.

[Explanation of symbols]

1. Natural language processing system 2 ... Morphological analysis unit 2A ... Morphological rules, 2B ... Morphological dictionary 3 ... Syntactic and Semantic Analysis Department 3A ... Grammar rule, 3B ... Bond valence dictionary

Claims (9)

[Claims] 1. A natural language processing system for handling utilization information of a utilization word, comprising a utilization word extraction means for extracting a morpheme that comprises utilization information from among morphemes forming an input sentence, and the extracted utilization word. Utilization information judging means for determining whether or not utilization information needs to be inserted, and utilization information inserting means for inserting utilization information into an input sentence according to a result of the determination by the utilization information judging means. Characteristic natural language processing system. 2. The conjugation information judging means refers to a conjugation unnecessary vocabulary list containing a specific auxiliary verb or verb, and the morpheme immediately following the morpheme having the conjugation information is included in the list. The natural language processing system according to claim 1, wherein the natural language processing system is characterized in that if it is inserted, otherwise it is determined not to be inserted. 3. The natural language processing system according to claim 1, wherein the utilization information inserting means inserts utilization information into a morphological analysis result as a character string corresponding to a word. 4. A grammar rule managing means for managing a difference in meaning according to a usage form of a usage word as a grammar rule, and the usage information inserted by the usage information insertion means is referred to by the grammar rule to input a sentence. The natural language processing system according to claim 1, further comprising: a meaning specifying unit that specifies a meaning. 5. A natural language processing method for handling utilization information of a utilization word, comprising a utilization word extracting step of extracting a morpheme that includes utilization information from morphemes forming an input sentence, and extracting the utilization word into the extracted utilization word. A utilization information determining step of determining whether or not utilization information needs to be inserted; and a utilization information inserting step of inserting utilization information into an input sentence according to a result of the determination in the utilization information determining step. Characteristic natural language processing method. 6. In the utilization information determining step, a morpheme immediately after a morpheme having inflectional information is included in the list by referring to a vocabulary list that does not require inflectional insertion including a specific auxiliary verb or verb. The natural language processing method according to claim 5, characterized in that if it is inserted, otherwise it is determined not to be inserted. 7. The utilization information inserting step inserts utilization information into a morpheme analysis result as a character string corresponding to a word.
The natural language processing method according to claim 5, wherein. 8. A grammar rule managing step for managing a difference in meaning according to a usage form of a usage word as a grammar rule, and a usage sentence inserted in the usage information inserting step by referring to the usage information inserted in the usage information inserting step. The natural language processing method according to claim 5, further comprising a meaning specifying step of specifying a meaning. 9. A computer program written in a computer-readable format so as to execute natural language processing for handling utilization information of a utilization word on a computer system, wherein utilization information is included from among morphemes forming an input sentence. Depending on the result of the utilization word extraction step for extracting the word containing the utilization information, the utilization information determination step for determining whether or not the utilization information needs to be inserted into the extracted utilization word, and the determination result in the utilization information determination step. A computer program comprising: a utilization information inserting step of inserting utilization information into an input sentence.