JP7227705B2 - Natural language processing device, search device, natural language processing method, search method and program - Google Patents

Description

The present invention relates to a natural language processing device, a search device, a natural language processing method, a search method and a program.

In natural language processing, which is a technology for processing natural language by computer, for example, different types of analysis processes such as morphological analysis, syntactic analysis, semantic analysis, and contextual analysis are executed in a chain on the sentence to be processed. . Here, the morphological analysis is a process of dividing a sentence into words in a language such as Japanese that does not have clear divisions between words, unlike a language such as English that has spaces between words. Syntax analysis is, for example, a process of identifying hierarchical positional relationships, such as dependencies between words or groups of words that make up a sentence, based on simple grammatical rules. Semantic analysis is, for example, a process of selecting a positional relationship using the meaning of a word when multiple types of positional relationship options exist side by side in syntactic analysis. For example, in semantic analysis, a concept dictionary or the like that defines interrelationships between semantic concepts (classification of semantic attributes) is used (see, for example, Patent Document 1).

In syntactic analysis, in general, a ``constituent'' is used as a unit, and a ``grammatical category (also called a grammatical category)'' of each constituent is specified. Parent-child/sibling relationship) is specified. Here, a constituent means a word (word) that constitutes a sentence and a collection thereof. Also, the grammatical category means, for example, the type of words and phrases, and more specifically, classification based on the grammatical characteristics of constituents. In addition, there is an idea that the grammatical category includes a part-of-speech classification and a classification of grammatical function (hereinafter referred to as "grammatical function classification"). In this case, the part-of-speech classification is classification such as "noun", "verb", "noun phrase", "verb phrase", and "prepositional phrase". Also, the grammatical function classification is a classification of "subject", "predicate", "object" and "complement". In syntactic analysis, a part-of-speech classification is specified as a grammatical category for each constituent. In addition, the hierarchical positional relationship expresses the connections between constituent elements in a hierarchical manner and by showing the anteroposterior relationship (order relationship). In addition, the result of parsing can be represented for each sentence in a syntax tree (hierarchical tree diagram, phrase structure tree), a form using multiple nested parentheses, or the like.

Also, for example, a plurality of analysis tools (programs) for morphological analysis and syntactic analysis are open to the public on the Internet. For example, Japanese morphological analysis tools include "JUMAN", "ChaSen", and "MeCab". Examples of Japanese parsing tools include "KNP" and "CaboCha". Examples of English parsing tools include Berkeley Parser and Stanford Parser.

JP-A-2003-263428

By the way, when retrieving a specified word or sentence from a plurality of documents, it may be easier to obtain a desired retrieval result if, for example, the grammatical function classification of the specified word or sentence is specified. In other words, by specifying the search keyword and the keyword as the subject, or specifying the keyword as the object, the percentage of search results that do not match the intention is reduced, and the search results are improved. It may be possible to increase the accuracy of the results. However, syntactic analysis generally only specifies the part-of-speech classification and not the grammatical function classification. Therefore, for example, it is necessary to specify the grammatical function classification by applying some processing to the syntactic analysis result. For example, when attempting to extract grammatical function classifications by performing semantic analysis, a dictionary such as that described in Patent Document 1 must be prepared. .

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances. An object is to provide a device, a search device, a natural language processing method, a search method, and a program.

According to one aspect of the present invention, a natural language processing device identifies a grammatical function classification indicating a grammatical function classification of a constituent composed of a group of one or more words, which is a constituent unit of a sentence. and syntactically analyzing an input sentence on a sentence-by-sentence basis, and expressing a plurality of said constituents and a constituent type which is a type of each of said constituents by a hierarchical positional relationship between said constituents. and the constituent element belonging to a predetermined grammatical function classification from the plurality of constituent elements based on the constituent type and the hierarchical positional relationship of each constituent element in the syntactic analysis result. An extractor for extracting is provided.

Further, according to one aspect of the present invention, the extracting unit extracts the predetermined constituent type, which is the first constituent type traced from the first constituent type, which is the predetermined constituent type, based on the syntactic analysis result. Two said constituents are specified, and a third said constituent having a predetermined positional relationship with the second said constituent is extracted as said constituent belonging to one of said grammatical function classifications.

Further, according to the aspect of the present invention, the extracting unit extracts from the first constituent whose constituent type is "sentence" to the second constituent whose constituent type is "verb phrase". is traced downward, and the third constituent whose constituent type is "verb" in the lower layer of the second constituent whose constituent type is "verb phrase" is added to the It is extracted as the constituent element belonging to "predicate" which is one of the grammatical function classifications.

Further, according to one aspect of the present invention, the extracting unit traces upward the constituent elements extracted as having the grammatical function classification of "predicate", identifying the constituent having a constituent type, and classifying the constituent having the constituent type of "noun phrase" in the same layer as the constituent of the identified "verb phrase" into one of the grammatical function classifications; It is extracted as the constituent element belonging to the "subject" which is one.

Further, according to the aspect of the present invention, the extracting unit extracts the constituent whose kind of constituent is "verb phrase" and which is connected to the extracted constituent whose grammatical function classification is "predicate". is traced downward, and the constituent element having the constituent type of "noun phrase" existing in the lowest layer is extracted as the constituent element belonging to "object or complement" which is one of the grammatical function classifications. do.

Further, according to the aspect of the present invention, the extraction unit extracts the plurality of from the constituents, the constituents belonging to the predetermined grammatical function classification are extracted.

The constituent element type of the second constituent element that is one layer above the first constituent element is "verb phrase", and it exists before the second constituent element in the same layer, and the second constituent element the constituent element type of the third constituent element connected in an upper layer to the same constituent element as the element is "auxiliary verb", and the constituent element of the fourth constituent element three layers above the first constituent element If the element type is "sentence" and the element does not exist in the four layers above the first element, the extraction unit extracts the first element. , are extracted as the constituents belonging to the "predicate" which is one of the grammatical function classifications.

Further, according to one aspect of the present invention, when the input sentence does not include an auxiliary verb, the constituent type of the second constituent that is one layer above the first constituent is "verb phrase". is, and the constituent type of the third constituent element that is two layers above the first constituent element is "sentence", and there is no constituent element that is three layers above the first constituent element; , the extraction unit extracts the first constituent as the constituent belonging to "predicate", which is one of the grammatical function classifications.

Further, according to one aspect of the present invention, when the input sentence is a compound sentence, the constituent type of the second constituent that is one layer above the first constituent is a "verb phrase." , the constituent element type of the third constituent element that is three layers above the first constituent element is "sentence", and it exists in front of the second constituent element in the same layer, and the second constituent element The constituent type of the constituent connected to the same constituent as the element in the upper layer is "auxiliary verb", and the constituent type of the constituent that is four layers above the first constituent is " is a sentence", the extraction unit extracts the first constituent as a constituent belonging to "predicate", which is one of the grammatical function classifications.

In addition, according to one aspect of the present invention, the search device performs a search using a grammatical function classification that indicates the grammatical function classification of a constituent consisting of a group of one or more words, which is a constituent unit of a sentence. and syntactically analyzing an input sentence on a sentence-by-sentence basis, and expressing a plurality of said constituents and a constituent type which is a type of each of said constituents by a hierarchical positional relationship between said constituents. and the constituent element belonging to a predetermined grammatical function classification from the plurality of constituent elements based on the constituent type and the hierarchical positional relationship of each constituent element in the syntactic analysis result. an extraction unit for extracting; a search processing unit for retrieving the result of grammatical function classification extracted by the extracting unit as a search target for the sentence containing the constituent element corresponding to the specified grammatical function classification and keyword; Prepare.

Further, according to one aspect of the present invention, the natural language processing method specifies a grammatical function classification indicating a grammatical function classification of a constituent consisting of a group of one or more words, which is a constituent unit of a sentence. A processing method, wherein an input sentence is syntactically analyzed on a sentence-by-sentence basis, and a plurality of said constituents and a constituent type, which is the type of each constituent, are represented by a hierarchical positional relationship between said constituents. generating an analysis result; and selecting the constituents belonging to a predetermined grammatical function classification from the plurality of constituents based on the constituent type and the hierarchical positional relationship of each constituent in the syntactic analysis result. and extracting.

Further, according to one aspect of the present invention, the search method searches using a grammatical function classification that indicates the classification of the grammatical function of a constituent consisting of a group of one or more words, which is a constituent unit of a sentence. and syntactically analyzing an input sentence on a sentence-by-sentence basis, and expressing a plurality of said constituents and a constituent type which is a type of each of said constituents by a hierarchical positional relationship between said constituents. and extracting the constituents belonging to a predetermined grammatical function classification from the plurality of constituents based on the constituent type and the hierarchical positional relationship of each constituent in the syntactic analysis result. and searching the sentence containing the construct corresponding to the specified grammatical function classification and keyword, using the grammatical function classification extracted by the extracting unit as a search target.

Further, according to one aspect of the present invention, the program uses a natural language processing method for specifying a grammatical function classification indicating a grammatical function classification of a constituent consisting of a group of one or more words, which is a constituent unit of a sentence. A program to be executed by a computer, parsing an input sentence on a sentence-by-sentence basis, determining a plurality of said constituents and a constituent type which is a type of each of said constituents, and determining a hierarchical positional relationship between said constituents. a step of generating a syntactic analysis result represented by and based on the constituent type and the hierarchical positional relationship of each constituent in the syntactic analysis result, from the plurality of constituents belonging to a predetermined grammatical function classification and extracting constituents.

According to each of the embodiments described above, it is possible to specify a predetermined grammatical function classification for a constituent by performing a simple process on the syntactic analysis result.

1 is a system diagram showing a configuration example of a natural language processing device according to an embodiment; FIG. FIG. 5 is a schematic diagram for explaining an operation example of the syntax analysis unit according to the embodiment; It is a figure which shows an example of the constituent contained in the parsing result which concerns on embodiment. FIG. 4 is a diagram showing a list of grammatical function classifications extracted by a grammatical function classification extraction unit according to the embodiment; 4 is a flow chart showing a first operation example of the grammatical function class extraction unit according to the embodiment; FIG. 10 is an explanatory diagram for explaining a first operation example of the grammatical function class extraction unit according to the embodiment; FIG. 5 is a schematic diagram for explaining a first operation example of the grammatical function class extraction unit according to the embodiment; FIG. 10 is a diagram showing a configuration example of a grammatical function classification extraction result according to the embodiment; 9 is a flow chart showing a second operation example of the grammatical function class extraction unit according to the embodiment; FIG. 11 is a schematic diagram for explaining a second operation example of the grammatical function class extraction unit according to the embodiment; FIG. 11 is a schematic diagram for explaining a second operation example of the grammatical function class extraction unit according to the embodiment; FIG. 11 is a schematic diagram for explaining a second operation example of the grammatical function class extraction unit according to the embodiment; 1 is a schematic block diagram showing the configuration of a computer according to an embodiment; FIG.

Hereinafter, each embodiment will be described with reference to the drawings.

<Embodiment>
FIG. 1 is a system diagram showing a configuration example of a natural language processing device 1 according to one embodiment.
The natural language processing device 1 shown in FIG. 1 is configured using one or more computers. Each of the one or more computers has a CPU (central processing unit), a main memory, an auxiliary memory, an input/output device, a communication device, etc., and the CPU executes a predetermined program stored in the auxiliary memory. Executes a predetermined process.
A natural language processing device 1 shown in FIG. The syntactic analysis unit 2, the grammatical function classification extraction unit 3, the storage unit 4, and the search unit 5 are configured by a combination of hardware and software provided in each computer that constitutes the natural language processing device 1. FIG. Further, the natural language processing device 1 may have a configuration including a grammatical function classification extraction unit 3 and a storage unit 4 configured using one or a plurality of computers (configuration not including the search unit 5 and the like), for example. However, a configuration including the syntax analysis unit 2 and the grammatical function classification extraction unit 3 (configuration not including the storage unit 4 and the search unit 5) may be used, or the syntax analysis unit 2, the grammatical function classification extraction unit 3, and the search unit 5 (not including the storage unit 4), or a device comprising the grammatical function classification extraction unit 3 and the search unit 5 alone. In addition, below, the case where the natural language processing apparatus 1 which concerns on this embodiment performs predetermined processing for English is demonstrated. Note that the grammatical function class extraction unit 3 is an example of an extraction unit.

The storage unit 4 stores a plurality of sets of documents 41, syntax analysis results 42, and grammatical function classification extraction results 43 that are associated with each other. The document 41 is information representing a document to be searched by the search unit 5 . For example, the document 41 is input from outside the natural language processing apparatus 1 via a communication line or a storage medium and stored in the storage unit 4 . The parsing result 42 is information representing the parsing result of the parsing unit 2 for each sentence included in the document 41 . The grammatical function classification extraction result 43 is information representing the grammatical function classification extraction result by the grammatical function classification extraction unit 3 for the syntactic analysis result included in the syntactic analysis result 42 .

The syntactic analysis unit 2 uses an existing syntactic analysis method to parse the natural language sentences included in the document 41 stored in the storage unit 4 based on the formal grammar for each sentence. 42 is stored in the storage unit 4 . That is, the syntactic analysis unit 2 performs syntactic analysis of an input sentence on a sentence-by-sentence basis, and expresses a plurality of constituents and a constituent type, which is the type of each constituent, by a hierarchical positional relationship between the constituents. produce results. An example of the syntax analysis result 42 is shown in FIG.

FIG. 2 is a schematic diagram for explaining an operation example of the parser 2 shown in FIG.
In FIG. 2, the parser 2 inputs one sentence 410 included in the document 41 stored in the memory 4 . Then, the parsing unit 2 outputs a parse tree 420 as a result of parsing one sentence 410 . A parse tree 420 forms part of the parse result 42 . Note that in this case, the sentence 410 is "This tree is illustrating the constitution relation."

The syntax tree 420 includes nodes 4202 to 4208 corresponding to the constituents "This", "tree", "is", "illustrating", "the", "constituency", and "relation" that make up the sentence 410, and a node It is constructed by connecting nodes 4209 to 4213 corresponding to a plurality of constituent elements hierarchically grouping 4202 to 4208 with a plurality of branches 4201 . Also, node 4202 is tagged with "ROOT" 4200 indicating that it is the node of the highest hierarchy. Each of the nodes 4201 to 4213 is indicated by a symbol indicating the type of constituent (constituent type) to which it corresponds. FIG. 3 shows examples of constituent symbols.

FIG. 3 is a diagram showing an example of the correspondence between symbols, contents, and constituent levels of each kind of constituents, for a plurality of kinds of constituents.
For example, the constituent content of the symbol "S" is "sentence" and the constituent level is "clause". Further, for example, the constituent content of the symbol "SBAR" is "subordinate clause" and the constituent level is "clause". Further, for example, the constituent content of the symbol "NP" is "noun phrase" and the constituent level is "phrase". Also, for example, the constituent content of the symbol "VP" is "verb phrase" and the constituent level is "phrase". Also, for example, the constituent content of the symbol "PP" is "prepositional phrase" and the constituent level is "phrase". Further, for example, the content of the constituent of the symbol "NN" is "noun" and the level of the constituent is "word". Further, for example, the constituent content of the symbol "VB" is "verb" and the constituent level is "word".

In the parse tree 420 shown in FIG. 2, for example, the node 4206 corresponds to the constituent of the "determinant (DT)" corresponding to the constituent "the", and the node 4207 corresponds to the constituent "noun (DT)" corresponding to the constituent "consistency". NN)”, and node 4208 corresponds to the constituent of “noun (NN)” corresponding to the constituent “relation”. Also, node 4209 corresponds to a constituent of "noun phrase (NP)". This node 4209 is located one layer higher than the nodes 4206 to 4208, and the nodes 4206 to 4208 are located in the same layer. In this case, a node 4209 (parent) and nodes 4206 to 4208 (children) are in a parent-child relationship, and the nodes 4206 to 4208 are siblings to each other. In addition, constituents positioned in a higher hierarchy include one or more constituents connected via one or more branches 4201 positioned in a lower hierarchy. In addition, the left and right positional relationship of each node 4202 to 4213 in the figure indicates the front and rear positional relationship of each corresponding constituent element. For example, each of the nodes 4202 to 4208 corresponding to the "word" level constituents shows the positional relationship (order relationship) between the left and right sides of the sentence regardless of the layer. In addition, for each node 4209 to 4213 corresponding to the constituent elements of the level of "phrase" or "clause", the positional relationship between the left and right indicates the positional relationship (order relationship) before and after the sentence for each layer.

Note that the syntax analysis unit 2 may output the result of syntax analysis in a format using nested multiple parentheses, for example. In the example shown in FIG. 2, the parsing result is, for example, "(ROOT (S (NP (DT This) (NN tree)) (VP (VBZ is) (VP (VBG illustration) (NP (DT the) (NN consistency ) (NN relation))))))”.

On the other hand, the grammatical function classification extraction unit 3 shown in FIG. Constituents belonging to a predetermined grammatical function classification are extracted from a plurality of constituents based on the hierarchical positional relationship of . Then, the grammatical function class extraction unit 3 stores the analysis result in the storage unit 4 as the grammatical function class extraction result 43 . In this embodiment, the grammatical function classification includes the classification of "subject", "predicate", "object" or "complement", and a predetermined Classifications to which constituents having a positional relationship belong ("initial subordinate clause", "noun phrase following predicate", etc.) are included (see FIG. 4 for details).

FIG. 4 is a diagram showing a grammatical function classification list extracted by the grammatical function classification extraction unit 3 shown in FIG.
FIG. 4 shows the correspondence between the symbols and contents of each grammatical function classification and the conditions of the constituents of the classification. The classification content of the symbol "F_SBAR1" is "initial subordinate clause", and the constituent condition is to be located "before a clause or phrase containing a subject, predicate, object or complement". The content of the classification of the symbol "F_PP1" is "prepositional phrase at the beginning", and the constituent condition is to be located "before a clause or phrase containing a subject, predicate, object or complement". The content of the classification of the symbol "F_S" is "subject", and the condition of the constituent is to be located "in a clause or phrase not contained in a subordinate clause". The content of the classification of the symbol "F_V" is "predicate (predicate verb)", and the condition of the constituent is to be located "in a clause or phrase not included in a subordinate clause". The content of the classification of the symbol "F_OC" is "object or complement" and the condition of the constituent is to be located "within a clause or phrase not contained in a subordinate clause". The content of the classification of the symbol "F_NP" is "a noun phrase that follows the predicate", and the condition of the constituent is that it is located "in a clause or phrase that is not included in a subordinate clause". The content of the classification of the symbol "F_ADJP" is "an adjective phrase following the predicate", and the condition of the constituent is to be located "in a clause or phrase not included in a subordinate clause". The content of the classification of the symbol "F_to" is "a To-infinitive that follows the predicate", and the condition of the constituent is that it is located "in a clause or phrase not included in a subordinate clause". The content of the classification of the symbol "F_PP2" is "a prepositional phrase after the predicate", and the constituent condition is "after a clause or phrase containing a subject, predicate, object or complement". . The content of the classification of the symbol "F_SBAR2" is "the subordinate clause following the predicate", and the constituent condition is positioned "after the clause or phrase containing the subject, predicate, object or complement". is.

When extracting constituents belonging to the grammatical function classification shown in FIG. Extract the corresponding constituents as follows. That is, when the sentence to be extracted is a simple sentence, the sentence includes only one clause and only one pair of subject and predicate. Identify predicates, or objects or complements, as "F_S", "F_V" or "F_OC" categories. If the sentence to be extracted is a complex sentence, the sentence includes a main clause and a subordinate clause, and a plurality of sets of subjects and predicates. , or to identify objects or complements in the "F_S", "F_V" or "F_OC" categories. If the sentence to be extracted is a compound sentence, the sentence includes a plurality of coordinately connected sentences. Therefore, if each sentence is a simple sentence, the grammatical function classification extraction unit 3 Identify the predicate or object or complement as being classified as "F_S", "F_V" or "F_OC", and if each sentence is a compound sentence, the subject, predicate or object contained in the main clause of that compound sentence Or identify the complement as being classified as "F_S", "F_V" or "F_OC".

In addition, the predicate verb means, for example, a verb that is the core of a predicate in English, for example. A predicate verb is a verb that expresses the action or state of the subject, and is generally placed immediately after the subject.

Next, a first operation example of the grammatical function class extraction unit 3 shown in FIG. 1 will be described with reference to FIGS. 5 to 8. FIG. FIG. 5 is a flow chart showing a first operation example of the grammatical function class extraction unit 3 shown in FIG. FIG. 6 is an explanatory diagram for explaining a first operation example of the grammatical function class extraction unit 3 shown in FIG. FIG. 7 is a schematic diagram for explaining a first operation example of the grammatical function classification extraction unit 3 shown in FIG. was having breakfast." as a syntax tree 422. In FIG. 7, the layers of constituent elements are shown as layer numbers 0-5. Layer number 0 is the highest layer (also referred to as the highest layer), and layer number 5 is the lowest layer (also referred to as the lowest layer). FIG. 8 is a diagram showing a configuration example of the grammatical function classification extraction result 43 shown in FIG.

In the first operation example, when activated, the grammatical function classification extraction unit 3 first determines whether or not there is a document to be extracted (step S10). If there is a document to be extracted, that is, if there is a document 41 that has not been extracted by the grammatical function classification extraction unit 3 among the plurality of documents 41 stored in the storage unit 4 ("Yes" in step S10). case), the grammatical function classification extraction unit 3 selects the document 41 to be extracted (step S11). On the other hand, if there is no document to be extracted ("NO" in step S10), the grammatical function class extraction unit 3 terminates the process.

After selecting the document 41 to be extracted in step S11, the grammatical function classification extraction unit 3 next selects an unprocessed sentence from the selected document 41 (step S12). Next, the grammatical function classification extraction unit 3 refers to the syntactic analysis result of the selected sentence included in the syntactic analysis result 42 corresponding to the selected document 41 (step S13).

Next, the grammatical function classification extraction unit 3 identifies predicates included in the selected sentence (step S14). In step S14, the grammatical function classification extracting unit 3 selects the uppermost layer (the smallest layer number) "S" (the lowest layer number) of the syntax tree, as shown in FIG. "VP" (verb phrase) is traced from the sentence), and the verb to the lower left of "VP" (verb phrase), that is, the verb preceding the verb phrase is specified as the predicate verb. In the example of the syntax tree 422 shown in FIG. 7, the grammatical function classification extraction unit 3 first traces from the "S" node 4221 to the "VP" node 4222 and then to the "VP" node 4223 . Next, the grammatical function classification extraction unit 3 extracts the word "was" corresponding to the "VBD" (verb) node 4224 at the lower left of the node 4222 of "VP" and "VBG" (the lower left of the node 4223 of "VP"). verb) is identified as the predicate verb.

Next, the grammatical function class extraction unit 3 identifies the subject included in the selected sentence (step S15). In step S15, the grammatical function classification extraction unit 3, as shown in FIG. Identify the 'NP' (noun phrase) in the same layer as the contained VP' as the subject'. In the example of the syntax tree 422 shown in FIG. 7, the grammatical function classification extraction unit 3 extracts the ``PRP'' (personal pronoun) connected below the ``NP'' node 4226 in the same layer as the ``VP'' node 4222. Identify the word "I" corresponding to node 4227 as the subject.

Next, the grammatical function classification extraction unit 3 identifies objects or complements included in the selected sentence (step S16). In step S16, the grammatical function classification extraction unit 3 selects the "lowest layer (largest layer number)" 'NP' in the same layer as the lower left verb of 'VP' containing the predicate verb is identified as the object/complement'. In the example of the syntax tree 422 shown in FIG. 7, the grammatical function classification extraction unit 3 is connected below the node 4226 of "NP" in the same layer as the node 4225 of "VBG" on the lower left of the node 4223 of "VP". Identify the word "breakfast" corresponding to node 4228 of "NN" (noun) as an object or complement.

Next, the grammatical function classification extraction unit 3 identifies grammatical function classifications before and after the clause or phrase containing the subject, predicate, object or complement (step S17). In step S17, the grammatical function classification extraction unit 3 classifies the symbol "F_SBAR1" (the subordinate clause at the beginning), the symbol "F_PP1" (the prepositional phrase at the beginning), the symbol "F_PP2" (after the predicate). Identify the constituent corresponding to the classification of a certain prepositional phrase) or the classification of the symbol "F_SBAR2" (the subordinate clause following the predicate).

In this embodiment, the constituent corresponding to the classification of the symbol "F_SBAR1" (the subordinate clause at the beginning) is the constituent kind "SBAR" ( subordinate clause). Constituents corresponding to the classification of the symbol "F_PP1" (prepositional phrases at the beginning) are constituents of the constituent type "PP" (prepositional phrases) that precede clauses or phrases containing subjects, predicates, objects or complements. It is raw. Constituents corresponding to the classification of the symbol "F_PP2" (prepositional phrases after predicate) are of the constituent type "PP" (prepositional phrases) after clauses or phrases containing subjects, predicates, objects or complements. is a constituent of Constructs corresponding to the classification of symbol "F_SBAR2" (subordinate clauses after predicate) are of the constituent kind "SBAR" (subordinate clauses) located after clauses or phrases containing subjects, predicates, objects or complements. is a constituent of

In the example of the syntax tree 422 shown in FIG. 7, in step S17, the grammatical function classification extraction unit 3 extracts the word corresponding to the constituent corresponding to the node 4229 of "SBAR" that precedes the node 4226 of "NP" that includes the subject. Identify "When", "she", "woke", and "up" as constituents corresponding to the classification of the symbol "F_SBAR1" (the subordinate clause at the beginning).

Next, the grammatical function class extraction unit 3 saves the identification results in steps S14 to S17 in the storage unit 4 as the grammatical function class extraction result 43 (step S18). FIG. 8 shows a portion 430 as a configuration example of the grammatical function classification extraction result 43 . A portion 430 shown in FIG. 8 is the grammatical function classification extraction result 43 for one document 41 including the sentence "When she woke up, I was having breakfast." used as an example in FIG. In this case, the portion 430 is identified by associating the identification information of the document 41 (“document identification information A”) with the identification information of the sentence “When she woke up, I was having breakfast” (“sentence identification information B”). It contains information indicating the words corresponding to the symbols and constituents of the grammatical functional taxonomy. The "document identification information A" is, for example, the file name of the document 41, URI (Uniform Resource Identifier), and the like. "Sentence identification information B" is information for identifying the sentence "When she woke up, I was having breakfast." can be For example, the retrieval unit 5 shown in FIG. 1 can refer to the document 41 and a specific sentence in the document 41 by designating the document identification information and the sentence identification information.

Next, the grammatical function classification extraction unit 3 determines whether or not there is an unprocessed sentence in the selected document 31 (step S19). If there is an unprocessed sentence ("YES" in step S19), the grammatical function class extraction unit 3 executes the processes after step S12 again. On the other hand, if there is no unprocessed sentence ("NO" in step S19), the grammatical function classification extraction unit 3 executes the processes after step S10 again.

As described above, in the first operation example, the grammatical function classification extracting unit 3 extracts a predetermined structure traced from a constituent of a predetermined constituent type (referred to as a first constituent) based on the syntactic analysis result. A constituent that is a class of primes (referred to as a second constituent) is specified, and a constituent that has a predetermined positional relationship with the second constituent (referred to as a third constituent) is classified as one of the grammatical function classifications. are extracted as constituents belonging to . Therefore, a predetermined grammatical function classification can be specified for a constituent by performing a simple process on the syntactic analysis result without using a concept dictionary or the like.

In the first operation example, the grammatical function classification extraction unit 3 extracts constituents (corresponding to the first constituent) whose constituent type is "sentence" from constituents whose constituent type is "verb phrase". (corresponding to the second constituent) is traced downward, and the Constituents that are "verbs" (corresponding to the third constituent) are extracted as constituents that belong to "predicate", which is one of the grammatical function classifications. That is, according to the first operation example, the grammatical function classification of the "predicate" can be specified for the constituent by performing a simple process on the syntactic analysis result.

Further, in the first operation example, the grammatical function class extraction unit 3 traces upward the constituents extracted as having the grammatical function class of "predicate", and finds the constituent type of "verb phrase" existing in the highest layer. , and the constituents whose constituent type is "noun phrase" in the same layer as the constituent of the identified "verb phrase" belong to "subject", which is one of the grammatical functional classifications Extract as constituents. That is, according to the first operation example, the grammatical function classification of the "subject" can be specified for the constituent by performing a simple process on the syntactic analysis result.

Further, in the first operation example, the grammatical function class extraction unit 3 directs constituents of which the class of constituents is "verb phrase" connected to the constituents extracted as having the grammatical function class of "predicate" to the lower layer. Then, the constituent element having the constituent type of "noun phrase" existing in the lowest layer is extracted as a constituent element belonging to "object or complement" which is one of the grammatical function classifications. That is, according to the first operation example, the grammatical function classification of "object or complement" can be specified for constituents by performing a simple process on the syntactic analysis result.

Next, a second operation example of the grammatical function class extraction unit 3 shown in FIG. 1 will be described with reference to FIGS. 9 to 12. FIG. FIG. 9 is a flow chart showing a second operation example of the grammatical function class extraction unit 3 shown in FIG. FIG. 10 is a schematic diagram for explaining a second operation example of the grammatical function class extraction unit 3 shown in FIG.
FIG. 10(a) shows an example of a pattern used by the grammatical function classification extraction unit 3, and FIG. 10(b) shows an example of the result of syntactic analysis to be processed when the input sentence is "The Student should read an interesting book .” is shown as a syntax tree 423 .
Here, this input sentence contains the auxiliary verb "should".
FIG. 11(a) shows an example of another pattern used by the grammatical function classification extraction unit 3, and FIG. 11(b) shows an example of the parsing result to be processed when the input sentence is "Members are loved by a "big dog that eats sugar toasts and have a good amount of muscle." Here, this input sentence does not contain an auxiliary verb.
FIG. 12(a) shows an example of another pattern used by the grammatical function classification extraction unit 3, and FIG. 12(b) shows an example of the syntactic analysis result to be processed when the input sentence is "In the library the student should read good.""Workbooks and the Librarian should ensure that all the books are vandal resistant." Here, the input sentences are compound sentences, each sentence containing the auxiliary verb "should".
In the second operation example, the grammatical function classification extraction unit 3 extracts the grammatical function classifications shown in FIG. 4 as in the first operation example.

In the second operation example, when activated, the grammatical function classification extraction unit 3 first determines whether or not there is a document to be extracted (step S20). If there is a document to be extracted, that is, if there is a document 41 that has not been extracted by the grammatical function class extraction unit 3 among the plurality of documents 41 stored in the storage unit 4 ("Yes" in step S20). case), the grammatical function classification extraction unit 3 selects the document 41 to be extracted (step S21). On the other hand, if there is no document to be extracted ("NO" in step S20), the grammatical function class extraction unit 3 terminates the process.

After selecting the document 41 to be extracted in step S21, the grammatical function classification extraction unit 3 next selects an unprocessed sentence from the selected document 41 (step S22). Next, the grammatical function classification extraction unit 3 refers to the syntactic analysis result of the selected sentence included in the syntactic analysis result 42 corresponding to the selected document 41 (step S23).

Next, the grammatical function class extraction unit 3 selects one or more patterns corresponding to the selected sentence type (step S24). Here, the pattern corresponding to the type of sentence is the form of the syntactic analysis result (syntax tree) set in advance for each predetermined sentence type and for each grammatical function classification. The sentence type is, for example, a sentence classification such as whether the sentence is a compound sentence, a non-compound sentence, or a sentence containing an auxiliary verb. A pattern corresponds to a form of connection between a plurality of nodes (constituents), and is set for each grammatical function classification. For each pattern, for each type of sentence, for example, if a certain node belongs to a certain grammatical function class, what kind of constituent type and what kind of connection relationship other nodes have with respect to that node? It can be defined by showing that Each pattern can be created by trial and error, for example, for a plurality of example sentences.

FIG. 10(a) shows a pattern for identifying the grammatical function class "F_V" (predicate) when the sentence is of a type containing the constituent "MD" (auxiliary verb). In this case, the pattern includes rules 1 through 4. When all of rules 1 to 4 are satisfied, the grammatical function class extraction unit 3 specifies that the grammatical function class of the node is "F_V" (predicate). Rule 1 is "first parent is VP (verb phrase)", rule 2 is "first parent's left brother is MD", rule 3 is "third parent is S (sentence)", rule 4 is "missing a fourth parent".

In the parse tree 423, which is the result of syntactic analysis of a sentence containing auxiliary verbs shown in FIG. The constituent kind of 4232 is "VP" and rule 1 is satisfied. Also, the constituent type of the left sibling node 4233 of the first parent node 4232 is "MD", and rule 2 is satisfied. The constituent type of node 4235, which is the parent (third parent) of node 4234, which is the second parent of node 4231, is "S (sentence)", and rule 3 is satisfied. And since the third parent node 4235 is "ROOT", rule 4 is satisfied. Therefore, based on the syntax tree 423 shown in FIG. 10B, the grammatical function classification extraction unit 3 can identify that the word "provide" corresponding to the node 4231 is the grammatical function classification "F_V" (predicate). .

Also, FIG. 11(a) shows a pattern for identifying the grammatical function class "F_V" (predicate) when the sentence is of a type that does not contain the constituent "MD" (auxiliary verb). In this case, the pattern includes rules 1 through 3. When all of rules 1 to 3 are satisfied, the grammatical function class extraction unit 3 specifies that the grammatical function class of the node is "F_V" (predicate). Rule 1 is "first parent is VP (verb phrase)", rule 2 is "second parent is S (sentence)", and rule 3 is "no third parent".

In the parse tree 424, which is the result of syntactic analysis of a sentence that does not include auxiliary verbs, shown in FIG. The constituent kind of node 4242 is "VP" and rule 1 is satisfied. Also, the constituent type of node 4243, which is the second parent of node 4241, is "S (sentence)", and rule 2 is satisfied. And since the second parent node 4243 is "ROOT", rule 3 is satisfied. Therefore, based on the syntax tree 424 shown in FIG. 11B, the grammatical function class extraction unit 3 can identify that the word "are" corresponding to the node 4241 is the grammatical function class "F_V" (predicate). .

Also, FIG. 12(a) shows a pattern for identifying the grammatical function class "F_V" (predicate) when the sentence type is a compound sentence and each sentence includes an auxiliary verb. In this case, the pattern includes rules 1 through 4. When all of rules 1 to 4 are satisfied, the grammatical function class extraction unit 3 specifies that the grammatical function class of the node is "F_V" (predicate). Rule 1 is "The first parent is VP (verb phrase)", Rule 2 is "The third parent is S (sentence)", Rule 3 is "The left brother of the first parent is MD (auxiliary verb)" ', and rule 4 is 'the fourth parent is S (sentence)'.

In the parse tree 425, which is the result of parsing that each sentence shown in FIG. The constituent kind of eye parent node 4252 is "VP" and rule 1 is satisfied. Also, the constituent type of node 4253, which is the third parent of node 4251, is "S (sentence)", and rule 2 is satisfied. Also, the constituent kind of the left sibling node 4254 of the first parent node 4252 is "MD", and rule 3 is satisfied. The constituent type of node 4255, which is the fourth parent of node 4251, is "S (sentence)", and rule 4 is satisfied. Therefore, based on the syntax tree 425 shown in FIG. 12(b), the grammatical function class extraction unit 3 can identify that the word "enable" corresponding to the node 4251 is the grammatical function class "F_V" (predicate). .

In addition, in the syntax tree 425, when attention is focused on the node 4256 whose constituent type is "verb (VB)", the constituent type of the first parent node 4257 of the node 4256 is "VP", and rule 1 is It is filled. Also, the constituent type of node 4258, which is the third parent of node 4256, is "S (sentence)", and rule 2 is satisfied. Also, the constituent type of the left sibling node 4259 of the first parent node 4257 is "MD", and rule 3 is satisfied. The constituent type of node 4255, which is the fourth parent of node 4256, is "S (sentence)", and rule 4 is satisfied. Therefore, based on the syntax tree 425 shown in FIG. 12(b), the grammatical function classification extraction unit 3 can identify that the word "enable" corresponding to the node 4256 is the grammatical function classification "F_V" (predicate). .

Now, after selecting a pattern corresponding to the type of sentence selected in step S24, the grammatical function classification extraction unit 3 selects the Each constituent belonging to a predetermined grammatical function classification is specified as described with reference to FIGS. 10 to 12 (step S25).

Next, the grammatical function class extraction unit 3 stores the result of identification in step S25 in the storage unit 4 as the grammatical function class extraction result 43 in the same manner as in the first operation example (step S26).

Next, the grammatical function classification extraction unit 3 determines whether or not there is an unprocessed sentence in the selected document 41 (step S27). If there is an unprocessed sentence ("YES" in step S27), the grammatical function class extraction unit 3 executes the processes after step S22 again. On the other hand, if there is no unprocessed sentence ("NO" in step S27), the grammatical function class extraction unit 3 executes the processes after step S20 again.

As described above, in the second operation example, the grammatical function classification extraction unit 3 selects a plurality of constituents when the constituent type of each constituent and the hierarchical positional relationship between the constituents conform to a predetermined pattern. , to extract constituents belonging to a predetermined grammatical function classification. Therefore, a predetermined grammatical function classification can be specified for a constituent by performing a simple process on the syntactic analysis result without using a concept dictionary or the like.

Further, in the second operation example, the grammatical function classification extraction unit 3 determines that the constituent type of a constituent (assumed to be a second constituent) that is one layer above a certain constituent (assumed to be a first constituent) is "verb The constituent type of a constituent (referred to as the third constituent) that is "phrase" and exists in front of the second constituent in the same layer and is connected in the upper layer to the same constituent as the second constituent is an "auxiliary verb", the constituent type of a constituent element three layers above the first constituent element (assumed to be the fourth constituent element) is "sentence", and four constituent elements above the first constituent element With a predetermined pattern that a constituent does not exist in the grammatical function classification is extracted as a constituent belonging to "predicate" which is one of That is, according to the second operation example, by performing a simple process on the syntactic analysis result, it is possible to specify the grammatical functional classification of the "predicate" for the constituent when the input sentence includes an auxiliary verb. can.

Further, in the second operation example, the grammatical function classification extraction unit 3 extracts a constituent (a second constituent) that is one layer higher than a certain constituent (assumed to be the first constituent) when the input sentence does not contain an auxiliary verb. The constituent type of the first constituent is "verb phrase", and the constituent type of the constituent two levels above the first constituent (which is assumed to be the third constituent) is "sentence", and With a predetermined pattern that there is no constituent element three layers above the constituent element of 1, when the constituent type of each constituent element and the hierarchical positional relationship between the constituent elements conform to the pattern, the first is extracted as a constituent belonging to "predicate" which is one of the grammatical function classifications. That is, according to the second operation example, by performing a simple process on the syntactic analysis result, it is possible to specify the grammatical functional classification of the "predicate" for the constituent when the input sentence includes an auxiliary verb. can.

Further, in the second operation example, when the input sentence is a compound sentence, the grammatical function classification extraction unit 3 extracts a constituent element that is one layer higher than a constituent element (referred to as a first constituent element) The constituent type of the first constituent is "verb phrase", the constituent type of the constituent three levels above the first constituent (which is called the third constituent) is "sentence", and the second constituent is A constituent element of a constituent element (fifth constituent element) that exists in front of the same layer of the element and is connected in the upper layer to another constituent element (fourth constituent element) like the second constituent element The type is "auxiliary verb", and the constituent type of the constituent four layers above the first constituent is "sentence" is used as a predetermined pattern, the constituent type of each constituent and each constituent When the hierarchical positional relationship between the elements matches the pattern, the first constituent is extracted as a constituent belonging to "predicate" which is one of the grammatical function classifications. That is, according to the second operation example, by performing a simple process on the syntactic analysis result, it is possible to specify the grammatical functional classification of the "predicate" for the constituent when the input sentence includes an auxiliary verb. can.

On the other hand, the search unit 5 (search device) shown in FIG. 1 includes a search processing unit 51 . The search processing unit 51 searches the grammatical function classification extraction result 43 as a search target for sentences containing constituents corresponding to the designated grammatical function classification 52 and keyword 53 , and outputs a search result 54 . The search unit 5 extracts, for example, a sentence including one or more words extracted as the specified grammatical function classification 52 that matches the specified keyword 53 and outputs it as a search result 54 . That is, the search processing unit 51 extracts an input sentence based on a plurality of constituents that are constituent units of the sentence, constituent types that are the types of each constituent, and hierarchical positional relationships between the constituents. with reference to the syntactic analysis result 42 representing, based on the constituent type of each constituent in the syntactic analysis result 42 and the hierarchical positional relationship between each constituent, a constituent belonging to a predetermined grammatical function classification is selected from a plurality of constituents The grammatical function class extraction result 43, which is the result of extracting

According to the search unit 5, by considering the grammatical function classification, it is possible to extract the intended sentence with higher precision than a simple keyword search that does not consider the grammatical function classification.

The grammatical function classification 52 and the keyword 53 input by the search unit 5 may be one each, or may be plural. In the case of a plurality of numbers, they can be combined using, for example, a logical sum, logical product, negation, or the like. For example, the search unit 5 includes one or more words that match the first keyword 53 in the first designated grammatical function classification 52 and the second keyword 53 in the second designated grammatical function classification 52 . can be searched for sentences that do not contain one or more words matching .

As described above, according to the present embodiment, it is possible to specify a predetermined grammatical function classification for a constituent by performing a simple process on the syntactic analysis result. Further, according to the present embodiment, by considering the grammatical function classification, it is possible to extract the intended sentence with higher accuracy than a simple keyword search that does not consider the grammatical function classification.

Note that the natural language processing device 1 of the present embodiment can process natural languages other than English, such as Japanese. In that case, the natural language processing device 1 may be configured to perform morphological analysis. The content of the document 41 is not limited. can be the document 41 to be processed. In this case, when equipment names, parts names, material names, and their attributes are used as keywords, more detailed searches can be performed by using grammatical function classifications such as subjects and objects, and search accuracy can be improved. There is expected. For example, measurement conditions are often written at the beginning, and it may be effective to perform a search to exclude words that match the keyword written at the beginning, for example. Further, the grammatical function class extraction unit 3 may extract (specify) a grammatical function class by combining the above-described first operation example and second operation example.

FIG. 13 is a schematic block diagram showing the configuration of the computer according to the above embodiment.
The computer 9 includes a CPU 91 , a main storage device 92 , an auxiliary storage device 93 and an interface 94 .
The natural language processing device 1 described above comprises a computer 9 . The operation of each processing unit described above is stored in the auxiliary storage device 93 in the form of a program. The CPU 91 reads out the program from the auxiliary storage device 93, develops it in the main storage device 92, and executes the above process according to the program. For example, the CPU 91 may be the syntactic analysis unit 2, the grammatical function classification extraction unit 3, and the search unit 5 (search processing unit 51) described above.
Further, the CPU 91 may secure a storage area corresponding to the above-described storage section 4 in the main storage device 92 or the auxiliary storage device 93 according to the program.
Some or all of the programs and data constituting the natural language processing device 1, the parser 2, the grammatical function class extraction unit 3, the search unit 5, the grammatical function class extraction result 43, etc. are computer readable. It can be distributed via recording media or communication lines. Moreover, when the natural language processing device 1 is composed of a plurality of computers, the plurality of computers may be arranged in a distributed manner via a network.

Although the embodiments of the present invention have been described in detail above with reference to the drawings, the specific configuration is not limited to these embodiments, and designs and the like are included within the scope of the gist of the present invention.

1 Natural language processing device 2 Syntax analysis unit 3 Grammar function classification extraction unit (extraction unit)
4 storage unit 5 search unit 51 search processing unit

Claims (12)

A natural language processing device that identifies a grammatical function classification indicating a grammatical function classification of a constituent consisting of a group of one or more words that is a constituent unit of a sentence,
A syntax for syntactically analyzing an input sentence on a sentence-by-sentence basis and generating a parsing result representing a plurality of said constituents and a constituent type which is a type of each said constituent by a hierarchical positional relationship between said constituents. an analysis unit;
an extraction unit that extracts the constituents belonging to a predetermined grammatical function classification from the plurality of constituents based on the constituent type and the hierarchical positional relationship of each constituent in the syntactic analysis result,
The predetermined grammatical function classification determines whether at least a constituent whose constituent type is "subordinate clause" or "prepositional phrase" is at the beginning of the sentence or after the predicate of the sentence. , has a distinguishable division ,
The extraction unit extracts the predetermined grammatical function classification from the plurality of constituents when the constituent type of each constituent and the hierarchical positional relationship between the constituents conforms to a predetermined pattern. extracting said constituents belonging to
the constituent type of the second constituent that is one layer above the first constituent is a "verb phrase";
the constituent type of the third constituent that exists in front of the second constituent in the same layer and is connected to the same constituent as the second constituent in the upper layer is an "auxiliary verb";
The constituent type of the fourth constituent element that is three layers above the first constituent element is "sentence", and
there is no said constituent in four layers above the first said constituent,
If it fits the pattern with
The extracting unit extracts the first constituent element as the constituent element belonging to "predicate", which is one of the grammatical function classifications.
Natural language processor. The extractor is
Based on the syntactic analysis result, identifying the second constituent that is the predetermined constituent type traced from the first constituent that is the predetermined constituent type,
2. The natural language processing device according to claim 1, wherein a third constituent having a predetermined positional relationship with the second constituent is extracted as the constituent belonging to one of the grammatical function classifications. The extractor is
Tracing downward from the first constituent whose constituent type is "sentence" to the second constituent whose constituent type is "verb phrase";
The third constituent element whose constituent type is "verb", which is under the second constituent element whose constituent type is "verb phrase", is selected according to one of the grammatical function classifications. 3. The natural language processing device according to claim 2, wherein the constituent element belonging to a certain "predicate" is extracted. The extractor is
tracing upward the constituent elements extracted as having the grammatical function classification of "predicate", and identifying the constituent element having the constituent type of "verb phrase" existing in the highest layer;
The constituent element whose constituent type is "noun phrase" and which is in the same layer as the constituent element of the specified "verb phrase" is regarded as the constituent element belonging to "subject" which is one of the grammatical function classifications. The natural language processing device according to claim 3, which extracts. the extracting unit traces downward the constituents of which the constituent type is "verb phrase" connected to the constituents extracted as having the grammatical function classification of "predicate";
5. The constituent element of claim 3 or 4, wherein the constituent element having the constituent type of "noun phrase" existing in the lowest layer is extracted as the constituent element belonging to "object or complement" which is one of the grammatical function classifications. The natural language processing device described. A search device for searching using a grammatical function classification that indicates the grammatical function classification of a constituent consisting of a group of one or more words, which is a constituent unit of a sentence,
A syntax for syntactically analyzing an input sentence on a sentence-by-sentence basis and generating a parsing result representing a plurality of said constituents and a constituent type which is a type of each said constituent by a hierarchical positional relationship between said constituents. an analysis unit;
an extraction unit for extracting the constituents belonging to a predetermined grammatical function classification from the plurality of constituents based on the constituent type and the hierarchical positional relationship of each constituent in the syntactic analysis result;
a search processing unit for searching, as a search target, the sentence containing the constituent element corresponding to the specified grammatical function classification and keyword, from the results of the grammatical function classification extracted by the extracting unit;
with
The predetermined grammatical function classification determines whether at least a constituent whose constituent type is "subordinate clause" or "prepositional phrase" is at the beginning of the sentence or after the predicate of the sentence. , has a distinguishable division ,
The extraction unit extracts the predetermined grammatical function classification from the plurality of constituents when the constituent type of each constituent and the hierarchical positional relationship between the constituents conforms to a predetermined pattern. extracting said constituents belonging to
the constituent type of the second constituent that is one layer above the first constituent is a "verb phrase";
the constituent type of the third constituent that exists in front of the second constituent in the same layer and is connected to the same constituent as the second constituent in the upper layer is an "auxiliary verb";
The constituent type of the fourth constituent element that is three layers above the first constituent element is "sentence", and
there is no said constituent in four layers above the first said constituent,
If it fits the pattern with
The extracting unit extracts the first constituent element as the constituent element belonging to "predicate", which is one of the grammatical function classifications.
search device. A natural language processing method for identifying a grammatical function classification indicating a grammatical function classification of a constituent consisting of a group of one or more words that is a constituent unit of a sentence,
parsing an input sentence sentence by sentence, and generating a syntactic analysis result representing a plurality of said constituents and a constituent kind which is a kind of each said constituent by a hierarchical positional relationship between said constituents; and,
a step of extracting the constituents belonging to a predetermined grammatical function classification from the plurality of constituents based on the constituent type and the hierarchical positional relationship of each constituent in the syntactic analysis result;
has
The predetermined grammatical function classification determines whether at least a constituent whose constituent type is "subordinate clause" or "prepositional phrase" is at the beginning of the sentence or after the predicate of the sentence. , has a distinguishable division ,
In the extracting step, the predetermined grammatical function is extracted from the plurality of constituents when the constituent type of each constituent and the hierarchical positional relationship between the constituents conform to a predetermined pattern. extracting the constituents belonging to the classification;
the constituent type of the second constituent that is one layer above the first constituent is a "verb phrase";
the constituent type of the third constituent that exists in front of the second constituent in the same layer and is connected to the same constituent as the second constituent in the upper layer is an "auxiliary verb";
The constituent type of the fourth constituent element that is three layers above the first constituent element is "sentence", and
there is no said constituent in four layers above the first said constituent,
If it fits the pattern with
In the extracting step, the first constituent element is extracted as the constituent element belonging to "predicate", which is one of the grammatical function classifications.
Natural language processing method. A retrieval method for retrieving using a grammatical function classification indicating the grammatical function classification of a constituent consisting of a group of one or more words that is a constituent unit of a sentence,
parsing an input sentence sentence by sentence, and generating a syntactic analysis result representing a plurality of said constituents and a constituent kind which is a kind of each said constituent by a hierarchical positional relationship between said constituents; and,
a step of extracting the constituents belonging to a predetermined grammatical function classification from the plurality of constituents based on the constituent type and the hierarchical positional relationship of each constituent in the syntactic analysis result;
a step of searching, as a search target, the sentence containing the constituent element corresponding to the specified grammatical function classification and keyword;
has
The predetermined grammatical function classification determines whether at least a constituent whose constituent type is "subordinate clause" or "prepositional phrase" is at the beginning of the sentence or after the predicate of the sentence. , has a distinguishable division ,
In the extracting step, the predetermined grammatical function is extracted from the plurality of constituents when the constituent type of each constituent and the hierarchical positional relationship between the constituents conform to a predetermined pattern. extracting the constituents belonging to the classification;
the constituent type of the second constituent that is one layer above the first constituent is a "verb phrase";
the constituent type of the third constituent that exists in front of the second constituent in the same layer and is connected to the same constituent as the second constituent in the upper layer is an "auxiliary verb";
The constituent type of the fourth constituent element that is three layers above the first constituent element is "sentence", and
there is no said constituent in four layers above the first said constituent,
If it fits the pattern with
In the extracting step, the first constituent element is extracted as the constituent element belonging to "predicate", which is one of the grammatical function classifications.
retrieval method. A program for causing a computer to execute a natural language processing method for identifying a grammatical function classification indicating a grammatical function classification of a constituent consisting of a group of one or more words that is a constituent unit of a sentence,
parsing an input sentence sentence by sentence, and generating a syntactic analysis result representing a plurality of said constituents and a constituent kind which is a kind of each said constituent by a hierarchical positional relationship between said constituents; and,
a step of extracting the constituents belonging to a predetermined grammatical function classification from the plurality of constituents based on the constituent type and the hierarchical positional relationship of each constituent in the syntactic analysis result;
A program that executes
The predetermined grammatical function classification determines whether at least a constituent whose constituent type is "subordinate clause" or "prepositional phrase" is at the beginning of the sentence or after the predicate of the sentence. , has a distinguishable division ,
In the extracting step, the predetermined grammatical function is extracted from the plurality of constituents when the constituent type of each constituent and the hierarchical positional relationship between the constituents conform to a predetermined pattern. extracting the constituents belonging to the classification;
the constituent type of the second constituent that is one layer above the first constituent is a "verb phrase";
the constituent type of the third constituent that exists in front of the second constituent in the same layer and is connected to the same constituent as the second constituent in the upper layer is an "auxiliary verb";
The constituent type of the fourth constituent element that is three layers above the first constituent element is "sentence", and
there is no said constituent in four layers above the first said constituent,
If it fits the pattern with
In the extracting step, the first constituent element is extracted as the constituent element belonging to "predicate", which is one of the grammatical function classifications.
program. A natural language processing device that identifies a grammatical function classification indicating a grammatical function classification of a constituent consisting of a group of one or more words that is a constituent unit of a sentence,
A syntax for syntactically analyzing an input sentence on a sentence-by-sentence basis and generating a parsing result representing a plurality of said constituents and a constituent type which is a type of each said constituent by a hierarchical positional relationship between said constituents. an analysis unit;
an extraction unit that extracts the constituents belonging to a predetermined grammatical function classification from the plurality of constituents based on the constituent type and the hierarchical positional relationship of each constituent in the syntactic analysis result,
The extraction unit extracts the predetermined grammatical function classification from the plurality of constituents when the constituent type of each constituent and the hierarchical positional relationship between the constituents conforms to a predetermined pattern. extracting said constituents belonging to
the constituent type of the second constituent that is one layer above the first constituent is a "verb phrase";
the constituent type of the third constituent that exists in front of the second constituent in the same layer and is connected to the same constituent as the second constituent in the upper layer is an "auxiliary verb";
The constituent type of the fourth constituent element that is three layers above the first constituent element is "sentence", and
there is no said constituent in four layers above the first said constituent,
If it fits the pattern with
The extracting unit extracts the first constituent element as the constituent element belonging to "predicate", which is one of the grammatical function classifications.
Natural language processor. A natural language processing device that identifies a grammatical function classification indicating a grammatical function classification of a constituent consisting of a group of one or more words that is a constituent unit of a sentence,
A syntax for syntactically analyzing an input sentence on a sentence-by-sentence basis and generating a parsing result representing a plurality of said constituents and a constituent type which is a type of each said constituent by a hierarchical positional relationship between said constituents. an analysis unit;
an extraction unit that extracts the constituents belonging to a predetermined grammatical function classification from the plurality of constituents based on the constituent type and the hierarchical positional relationship of each constituent in the syntactic analysis result,
The extraction unit extracts the predetermined grammatical function classification from the plurality of constituents when the constituent type of each constituent and the hierarchical positional relationship between the constituents conforms to a predetermined pattern. extracting said constituents belonging to
When the input sentence does not contain an auxiliary verb,
the constituent type of the second constituent that is one layer above the first constituent is a "verb phrase";
The constituent type of the third constituent that is two layers above the first constituent is "sentence", and
There are no constituents three layers above the first said constituent,
If it fits the pattern with
The extracting unit extracts the first constituent element as the constituent element belonging to "predicate", which is one of the grammatical function classifications.
Natural language processor. A natural language processing device that identifies a grammatical function classification indicating a grammatical function classification of a constituent consisting of a group of one or more words that is a constituent unit of a sentence,
A syntax for syntactically analyzing an input sentence on a sentence-by-sentence basis and generating a parsing result representing a plurality of said constituents and a constituent type which is a type of each said constituent by a hierarchical positional relationship between said constituents. an analysis unit;
an extraction unit that extracts the constituents belonging to a predetermined grammatical function classification from the plurality of constituents based on the constituent type and the hierarchical positional relationship of each constituent in the syntactic analysis result,
The extraction unit extracts the predetermined grammatical function classification from the plurality of constituents when the constituent type of each constituent and the hierarchical positional relationship between the constituents conforms to a predetermined pattern. extracting said constituents belonging to
When the input sentence is a compound sentence,
the constituent type of the second constituent that is one layer above the first constituent is a "verb phrase";
the constituent type of the third constituent that is three layers above the first constituent is "sentence";
The constituent type of the constituent that exists in front of the second constituent in the same layer and is connected to the constituent that is the same as the second constituent in the upper layer is an "auxiliary verb", and
The constituent element type of the constituent element four layers above the first constituent element is "sentence",
If it fits the pattern with
The extracting unit extracts the first constituent element as the constituent element belonging to "predicate", which is one of the grammatical function classifications.
Natural language processor.

Images