JP4653375B2 - Structured document generation apparatus and structured document generation program - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a structured document generation device and a structured document generation program, and more particularly to a structured document generation device and a structured document generation program suitable for generating data that can be directly used for retrieval.
[0002]
[Prior art and problems to be solved by the invention]
In Japanese Patent Laid-Open No. 2001-290801, a structured document system, a structured document program, and a computer-readable storage medium (hereinafter, “conventional document”) that inputs a text document and outputs a structured document using the document structure. Technology 1 ”) has been proposed.
[0003]
The prior art 1 is intended to reconstruct a text-format document for display on a Web page, and in order to decompose the document structure, a document decomposition definition file is defined in advance, and the document definition The document is structured using patterns described in the file. The document structure generated in this way has a surface format such as title, creator, and date.
[0004]
However, the portion written in the natural text as the text is output as it is without any conversion, so it is not in a format that can be used in a computer system. For this reason, the structured document output from the prior art 1 is used in the database only for surface information, but there is a problem that specific contents such as text written in a natural sentence are not used in the database. It was.
[0005]
The present invention has been proposed in order to solve the above-described problems, and a structured document generation apparatus and structure for generating a structured document having a semantic structure that can be easily recognized by a computer from a natural sentence. An object is to provide a computerized document generation program.
[0018]
[Means for Solving the Problems]
According to the first aspect of the present invention, a document input unit for inputting a document including a natural sentence, and a phrase analysis for the document including the natural sentence input by the document input unit are configured to configure the document. A phrase analysis means for outputting a phrase-analyzed document in which a phrase is associated with a part-of-speech type; Provided for each concept representing the meaning of the word string, Composition of part-of-speech types of phrase strings containing attribute values to be extracted And patterns corresponding to each concept The part-of-speech type composition pattern that represents the attribute value and the corresponding tag is structured. For outputting structured elements for each concept A concept dictionary storage means for storing a concept dictionary having output format information; a part-of-speech type configuration pattern of a phrase string of the phrase-analyzed document; and a part-of-speech type configuration of the concept dictionary stored in the concept dictionary storage means Matching a pattern, selecting a concept dictionary corresponding to the semantic content of the phrase string from the concept dictionary storage means, and using the selected concept dictionary, the phrase analysis completed corresponding to the part of speech type composition pattern Based on attribute value extraction means for extracting an attribute value from a word sequence of a document, attribute values extracted by the attribute value extraction means, and output format information of the selected concept dictionary, Represents a structured element for the concept corresponding to the selected concept dictionary Document generating means for generating a structured document.
[0019]
Claim 1 The invention described in claim 1 is claimed in a computer. 3 It is constituted by installing the invention described in 1.
[0020]
According to a third aspect of the present invention, the computer executes a phrase analysis on a document input unit that inputs a document including a natural sentence, and a document including the natural sentence input by the document input unit, and the document A phrase analysis means for outputting a phrase-analyzed document in which the phrases constituting the phrase and part-of-speech types are associated with each other; Provided for each concept representing the meaning of the word string, Composition of part-of-speech types of phrase strings containing attribute values to be extracted And the pattern corresponding to the concept The part-of-speech type composition pattern that represents the attribute value and the corresponding tag is structured. A structured element of the concept Output for A concept dictionary having output format information The A concept dictionary storage means for storing, a composition pattern of a part of speech type of a phrase string of the phrase analyzed document, Of the concept dictionary stored in the concept dictionary storage means. Compare the part-of-speech type composition pattern, select a concept dictionary corresponding to the semantic content of the word string from the concept dictionary storage means, and use the selected concept dictionary to correspond to the part-of-speech type composition pattern Based on attribute value extraction means for extracting an attribute value from the phrase string of the phrase analyzed document, the attribute value extracted by the attribute value extraction means, and the output format information of the selected concept dictionary, Represents a structured element for the concept corresponding to the selected concept dictionary It is a structured document generation program that functions as a document generation unit that generates a structured document.
[0021]
The phrase analysis unit performs phrase analysis on a document including a natural sentence. Here, the document including the natural sentence may be only a document including only the natural sentence, or may be a document in a predetermined format having the natural sentence and a predetermined tag. Then, the phrase analysis means can phrase Each part-of-speech type to phrase And the phrase-analyzed document in which each part-of-speech type is associated with each other.
[0022]
The concept dictionary storage means stores a concept dictionary having a part-of-speech type configuration pattern and output format information. Here, the part-of-speech type composition pattern is the composition of the part-of-speech type of the phrase string including the attribute value to be extracted when trying to extract the attribute value from the document. A pattern corresponding to a concept It represents. The output format information indicates the output format of the document, and the attribute value and corresponding tag are structured. A structured element of the concept Output for Information.
[0023]
The attribute value extraction means phrase Match the part-of-speech type composition pattern in the column with the part-of-speech type composition pattern in the concept dictionary to correspond to this part-of-speech type composition pattern phrase Find a column from a parsed document. And found out phrase Extract attribute values from columns.
The document generation means generates a structured document of a predetermined format based on the extracted attribute value and the output format information of the concept dictionary. This structured document is a document of a format composed of an attribute value and a tag even if it is a natural sentence at the beginning.
[0024]
Therefore, according to the first and third aspects of the invention, the part-of-speech type representing the composition pattern of the part-of-speech type of the phrase string of the phrase-analyzed document and the part-of-speech type of the phrase string including the attribute value to be extracted Attribute values are extracted by comparing with configuration patterns, and the extracted attribute values are structured based on the output format information Represents the element By generating a structured document, a structured document whose meaning is extracted from a natural sentence can be obtained.
[0025]
The invention according to claim 2 is the invention according to claim 1, in front The attribute value extracting means selects a concept dictionary having a part-of-speech type configuration pattern that matches a part-of-speech type configuration pattern of the word-analyzed document from a plurality of concept dictionaries stored in the concept dictionary storage unit. The attribute value is extracted using the selected concept dictionary, and the document creating means converts the attribute value extracted by the attribute value extraction means and the output format information of the concept dictionary selected by the attribute value extraction means. Based on this, a structured document is generated.
[0026]
Claim 2 The invention described in claim 1 is claimed in a computer. 4 It is constituted by installing the invention described in 1.
[0027]
The invention according to claim 4 is the invention according to claim 3, in front The attribute value extracting means selects a concept dictionary having a part-of-speech type composition pattern that matches a part-of-speech type composition pattern of the word-analyzed document from a plurality of concept dictionaries stored in the concept dictionary storage means. The attribute value is extracted using the selected concept dictionary, and the document creating means converts the attribute value extracted by the attribute value extraction means and the output format information of the concept dictionary selected by the attribute value extraction means. Based on this, a structured document is generated.
[0028]
There are various concept dictionary storage means Represents the semantic content of a word string Each concept dictionary is stored for each concept. The attribute value extraction means specifies the concept of the phrase-analyzed document by selecting a concept dictionary having a part-of-speech type composition pattern that matches the part-of-speech type composition pattern of the word-parsed document from various concept dictionaries. . Then, the attribute value related to the identified concept is obtained by extracting the attribute value using the selected concept dictionary.
[0029]
Therefore, the claims 2 and 4 According to the invention described in Represents the semantic content of a word string Select a concept dictionary that has a part-of-speech type composition pattern that matches the part-of-speech type composition pattern of the word-analyzed document from the concept dictionary, and extract attribute values using the selected concept dictionary to determine which document contents Even with such a concept, the attribute value can be reliably extracted.
[0030]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the drawings.
[0031]
FIG. 1 is a block diagram showing a configuration of a structured document generation apparatus according to an embodiment of the present invention.
[0032]
The structured document generating apparatus includes a keyboard 1 for inputting character information by key operation, a mouse 2 as a pointing device, an input / output port 3 for inputting / outputting information to / from the outside, and a structured document creation. Hard disk drive (Hard Disc Drive) 4 that stores an application program for executing the arithmetic processing of the above, a microcomputer 5 that executes arithmetic processing for creating a structured document, and an operation result of the microcomputer 5, for example, a structured document LCD (Liquid Crystal Display) 6 is displayed.
[0033]
The microcomputer 5 includes a RAM (Random Access Memory) that is a data work area, a ROM (Read Only Memory) in which a predetermined control program is stored, a CPU (Central Processing Unit) that executes arithmetic processing, and the like. Yes. The microcomputer 5 creates a structured document based on a text document input from the keyboard 1 or a text document input from the outside via the input / output port 3. Note that a structured document may be created from a text document already recorded on a recording medium (not shown).
[0034]
FIG. 2 is a block diagram showing a functional configuration of the microcomputer 5. The microcomputer 5 includes an XML document creation unit 10 that creates an XML document from a text document, a phrase analysis unit 20 that performs phrase analysis, and a meaning extraction unit 30 that extracts a meaning and creates a structured document. Yes.
[0035]
The XML document creation unit 10 creates an XML document by using a natural text document and information such as the title and creation date of the text document, and supplies the created XML document to the phrase analysis unit 20. The text document may be any of a document generated by operating the keyboard 1, a document input from the input / output port 3, or a document stored in advance in the HDD 4 or other storage medium.
[0036]
FIG. 3 is a block diagram illustrating a functional configuration of the phrase analysis unit 20.
[0037]
The phrase analysis unit 20 includes a phrase dictionary 21 that indicates correspondence between words and parts of speech, a content character string separation unit 22 that separates a structure tag of an XML document and a content character string that is a natural sentence, a specific word Specific word separation / replacement unit 23 that separates and replaces, specific word part-of-speech determination unit 24 that determines the part-of-speech of the specific word, morpheme analysis unit 25 that performs morpheme analysis of the content character string, and substitution word replacement with the specific word Alternative word replacement unit 26 and a compound word or phrase from a plurality of words A phrase string A compound phrase generation unit 27 for generating A phrase that is a word, compound word or phrase And a part-of-speech tag assigning unit 28 for attaching a part-of-speech tag to each other.
[0038]
FIG. 4 is a diagram showing the configuration of the word dictionary 21. The phrase dictionary 21 includes a dictionary related to words and a dictionary related to phrases (compound words and phrases).
[0039]
The phrase dictionary 21 is composed of words, part-of-speech types, and alternative words with respect to words. Although the alternative word will be described in detail later, it is used only in the case of a specific word.
[0040]
According to FIG. 4, for example, for the word “amplification”, the part of speech type is “noun-sa change connection”. This means that “amplification” is a noun and is connected to a sub-variable. For the word “problem”, the part of speech type is “noun-Nay adjective stem”. This means that "problem" is a noun and becomes the stem of the adjective "... not". Further, for the word “time”, the part of speech type is “noun-suffix-adverbial possible”. This indicates that “time” is a noun, becomes a suffix “... time”, and can be used as an adverb. For the word “ga”, the part of speech type is “particle-case particle-general”. This indicates that “ga” is a general case particle. In this way, the part of speech type represents not only the type of part of speech of the corresponding word but also the attribute of the corresponding word.
[0041]
Here, the “word” includes a specific word indicating a word newly defined by the phrase dictionary 21 in addition to general words such as nouns and particles.
[0042]
For example, “2S [AD] [0-9] +” shown in FIG. 4 starts with 2S, the next is one of letters A to D, and the last is composed of one or more numbers. Represents a word. A word corresponding to the above condition represents a transistor part number and is not in a general dictionary. Therefore, the phrase dictionary 21 defines words that meet the above conditions as specific words.
[0043]
Here, for the word (specific word) “2S [AD] [0-9] +”, the part of speech type is “noun—proper noun—identifier—part number—transistor”, and the alternative term is “substitution—transistor part number”. It is. This indicates that the specific word is a proper noun, an identifier, and a transistor product number. Further, since the specific word itself does not understand the meaning, it represents that “transistor part number” exists as an alternative word.
[0044]
The phrase dictionary 21 is a phrase Column (Compound words and phrases) pattern ”And“ part of speech ”and also functions as a dictionary for creating compound words and phrases.
[0045]
"Phrase string pattern "Represents a pattern to be established as a compound word or phrase. For example, according to FIG. 4, “(noun and not (noun-suffix-adverb possible) or symbol) * [2, ∞]” is a noun (except for noun-suffix-adverb possible) or a symbol, Two or more continuous patterns are represented, and the “part of speech type” of this pattern is a noun phrase. In other words, this of The pattern of the phrase string that meets such a condition indicates that it is handled as a noun phrase.
[0046]
5 and 6 are diagrams for explaining a document analyzed by the phrase analysis unit 20. Hereinafter, a case where the XML document shown in FIG. 5A is input to the phrase analysis unit 20 will be described.
[0047]
When the XML document shown in FIG. 5A is supplied from the XML document creation unit 10, the content character string separation unit 22 separates the structure tag of the XML document from the content character string that is a natural sentence. Then, the content character string separation unit 22 supplies the XML document structure tag shown in FIG. 5B to the part-of-speech tag adding unit 28, and the content character string of the XML document shown in FIG. This is supplied to the replacement unit 23.
[0048]
The specific word separation / replacement unit 23 collates the content character string supplied from the content character string separation unit 22 with the phrase dictionary 21, replaces the specific word included in the content character string with an alternative word, and converts the specific word To separate. Then, the separated specific word is supplied to the specific word part-of-speech determination unit 24, and the content character string in which the specific word is replaced with the alternative word is supplied to the morpheme analysis unit 25.
[0049]
Specifically, the specific word separation / replacement unit 23 replaces the specific word “2SC7777” included in the content character string with “substitution-transistor part number” based on the phrase dictionary 21. Then, the specific word “2SC7777” shown in FIG. 5D is separated and supplied to the specific word part-of-speech determination unit 24, and A in the replaced content character string “(substitution-transistor part number) shown in FIG. "Heat generation is a problem during class amplification" is supplied to the morphological analysis unit 25.
[0050]
The specific word part of speech determination unit 24 determines the part of speech of the specific word based on the phrase dictionary 21 and supplies the specific word and the part of speech type information to the alternative word replacement unit 26. Specifically, the part of speech of “2SC7777” is determined, and as shown in FIG. 5 (F), “2SC7777-noun-proper noun-identifier-article-transistor” as specific word and part-of-speech type information is replaced with an alternative word. To the unit 26.
[0051]
The morpheme analysis unit 25 performs morpheme analysis on the content character string supplied from the specific word separation / replacement unit 23 with reference to the phrase dictionary 21. Specifically, the content character string is decomposed into individual words, and each decomposed word (including a specific word) is associated with a part of speech type. Then, the morpheme analyzer 25 supplies each word shown in FIG. 5G and the corresponding part of speech type to the substitute word replacement unit 26. In FIG. 5G, the words “hour”, “ni”, “fever”, “ga”, and “problem” are omitted, but morphological analysis is similarly performed for these words. Then, each word and the corresponding part of speech type are supplied to the alternative word replacement unit 26.
[0052]
The alternative word replacement unit 26 selects an alternative word from each word subjected to morphological analysis based on the information supplied from the specific word part-of-speech determination unit 24 and the morpheme analysis unit 25, and specifies the selected alternative word as the original specification. Replace with a word. Specifically, “alternative-transistor” is replaced with “2SC7777”. 6A is supplied to the compound word / phrase generation unit 27. In FIG. 6A, the words “hour”, “ni”, “fever”, “ga”, and “problem” are not shown. These words are also omitted from FIG. 6B described later.
[0053]
The compound word / phrase generating unit 27 determines whether there is a partial sequence of words corresponding to the compound word / phrase in the phrase dictionary 21 among the continuous words supplied from the substitute word replacing unit 26, and the corresponding word If there is a substring, replace all applicable substrings with compound words / phrases.
[0054]
Specifically, the word substrings “A”, “class”, and “amplification” correspond to the word string patterns of the noun phrases in the word dictionary 21 shown in FIG. Therefore, the compound phrase generation unit 27 generates “class A amplification” that is a noun phrase from the word subsequences “A”, “class”, and “amplification”, and each of them is shown in FIG. phrase And the corresponding part-of-speech type are supplied to the part-of-speech tag assignment unit 28. The compound phrase generation unit 27 supplies the information supplied from the alternative word replacement unit 26 to the part-of-speech tag assignment unit 28 as it is when there is no partial sequence of words corresponding to the compound phrase in the phrase dictionary 21.
[0055]
The part-of-speech tag giving unit 28 receives each of the parts supplied from the compound phrase generation unit 27 phrase A part-of-speech tag indicating each part-of-speech type is given to (including compound words and phrases). And each phrase And each part-of-speech tag are embedded as elements of the structure tag <content> supplied from the content character string separation unit 22. As a result, the part-of-speech tag assigning unit 28 generates and outputs an XML document that has been subjected to phrase analysis, as shown in FIG.
[0056]
In this way, the phrase analysis unit 20 separates the structure tag and the content character string (natural sentence) from a general XML document, performs morphological analysis on the content character string, and performs each of the content character strings. phrase A part-of-speech tag to each phrase And the part-of-speech tag are embedded in the original XML document. That is, the phrase analysis unit 20 performs morphological analysis on the natural sentence, and each of the natural sentences constituting the natural sentence. phrase By adding a part-of-speech tag, it is possible to clarify the document structure of a natural sentence and output a document whose meaning is easy to extract.
[0057]
In addition, the phrase analysis unit 20 registers the newly defined specific word and its alternative word in the phrase dictionary 21, so that the alternative word is used instead of the specific word in the morphological analysis, and after the morphological analysis, By replacing the substitute word with the original specific word, it is possible to accurately perform morphological analysis even for technical terms and technical terms that are not in a general dictionary.
[0058]
The phrase dictionary 21 is not limited to the configuration shown in FIG. 4 and can store other words and parts of speech types. The phrase dictionary 21 can store not only patterns for creating noun phrases from a plurality of nouns, but also patterns for generating other compound words / phrases such as adjective phrases and adverb phrases. Of course.
[0059]
FIG. 7 is a block diagram illustrating a functional configuration of the meaning extraction unit 30. The meaning extraction unit 30 extracts the meaning from the XML document that has been subjected to the phrase analysis, and generates an XML document in which the extracted meaning is structured.
[0060]
The meaning extraction unit 30 selects an optimal item concept dictionary from a plurality of item concept dictionaries by performing pattern matching and a concept dictionary 31 that stores attribute value extraction patterns and attribute value output formats for various concepts. A pattern matching unit 32; an attribute value extracting unit 33 that extracts an attribute value using the selected item concept dictionary; and a shaping unit 34 that formats and outputs the tag corresponding to the extracted attribute value into an XML document. ing.
[0061]
FIG. 8 is a block diagram showing the configuration of the concept dictionary 31. The concept dictionary is composed of, for example, a plurality of item concept dictionaries configured for various concepts (items) such as “current problem” and “solution”. Here, the “current problem” will be described as an example.
[0062]
The item concept dictionary includes an “item” representing an item in the item concept dictionary, a “synonym” representing a synonym of the “item”, an “attribute describing the item”, and an “XML output” representing an XML output format of the extracted attribute value. Format ".
[0063]
The “attribute that describes the item” includes “attribute name” and “tag pattern”. The “attribute name” represents what attribute value is extracted. "Tag pattern" includes attribute values phrase It represents the configuration pattern of each part-of-speech tag in the column and the position of the attribute value included in the configuration pattern.
[0064]
Here, “noun *” represents an arbitrary noun or noun phrase. The combination of “$” and the attribute name following it indicates that the face of the word corresponding to the position of the tag pattern becomes the value of the attribute.
[0065]
For example, in FIG. 8, “problem” indicates that an attribute value that describes what kind of “problem” is extracted from the XML document that has been analyzed. "Tag pattern" of "Problem" includes an attribute value that describes "Problem" phrase It represents the part-of-speech tag composition pattern of the column.
[0066]
The “tag pattern” of “problem” phrase The part of speech tag is a noun or noun phrase phrase The part-of-speech tag can be arbitrary, but the following phrase Indicates that “is”. Here, since there is a “$ problem” in the first part of speech tag element, it corresponds to the first part of speech tag. phrase Becomes the value (attribute value) of the attribute name “problem”.
[0067]
“Target” indicates that an attribute value that describes what kind of “target” is extracted from the XML document that has been analyzed. The “tag pattern” of “target” is the first phrase The part of speech tag is a noun or noun phrase phrase The part-of-speech tag can be arbitrary, but the following phrase ( phrase Column) indicates “in”. In addition, since there is “$ target” in the element of the first part of speech tag, it corresponds to the first part of speech tag. phrase Becomes the value (attribute value) of the attribute name “target”.
[0068]
Further, “problem occurrence situation” indicates that an attribute value that describes what kind of “problem occurrence situation” is extracted from the XML document that has been analyzed. The “tag pattern” in the “problem situation” phrase The part of speech tag is a noun or noun phrase phrase The part-of-speech tag is "noun-suffix-adverbable" (a noun that becomes a suffix that can be used as an adverb) phrase The part of speech tag can be arbitrary, but the last part phrase Indicates “to”. In addition, since the element of the first part-of-speech tag includes “$ problem occurrence”, it corresponds to the first part-of-speech tag. phrase Becomes the value (attribute value) of the attribute name “problem occurrence situation”.
[0069]
The “XML output format” represents outputting an XML document in which each extracted attribute value is replaced with the location of “$ attribute name”.
[0070]
For example, the first line of “XML output format” in FIG. 8 indicates that the output is replaced with the attribute value of “problem”. Note that <target> and <when problem occurs> incorporated as elements of <current problem> indicate that corresponding elements are formed and output together with the structure tag.
[0071]
And the pattern collation part 32, the attribute value extraction part 33, and the shaping part 34 perform the following processes using the concept dictionary 31 comprised as mentioned above.
[0072]
The pattern matching unit 32 collates the phrase analyzed XML document supplied from the phrase analyzing unit 20 with each item concept dictionary of the concept dictionary 31 and selects an item concept dictionary corresponding to the XML document. Specifically, an item concept dictionary in which all tag patterns completely match the constituent pattern of the part of speech tag of the document is selected from each item concept dictionary. Then, the pattern matching unit 32 supplies the “item” of the selected item concept dictionary to the attribute value extraction unit 33 and also supplies the attribute value extraction unit 33 with the XML document corresponding to the item concept dictionary.
[0073]
For example, here, the composition pattern of the part-of-speech tag of the document shown in FIG. 6C matches the “tag pattern” of the item concept dictionary “current problem” shown in FIG. Accordingly, the pattern matching unit 32 supplies the item name “current problem” and the element of <content> in FIG.
[0074]
Note that when there are a plurality of item concept dictionaries in which the component pattern of part-of-speech tags of the document supplied from the phrase analysis unit 20 and all the tag patterns completely match, the condition of the tag pattern of the item concept dictionary is further limited. That's fine. For example, instead of “noun *” representing an arbitrary noun or noun phrase, “noun-proper noun” representing a proper noun may be used, or other conditions may be limited.
[0075]
The attribute value extraction unit 33 reads the item concept dictionary indicated by the “item” supplied from the pattern matching unit 32 from the concept dictionary 31, and converts the item concept dictionary from the document supplied from the pattern matching unit 32 to the tag pattern of the item concept dictionary. Applicable phrase Find a column. And phrase Corresponds to "$ attribute value" of the tag pattern from the column phrase Is extracted as an attribute value.
[0076]
Specifically, the attribute value extraction unit 33 corresponds to each of the three tag patterns shown in FIG. 8 from the <content> elements in FIG. phrase Find a column. And each found out phrase Corresponds to the “$ attribute value” of each tag pattern from the column phrase “Heat generation”, “2SC7777”, “Class A amplification” are extracted and supplied to the shaping unit 34 as attribute values.
[0077]
The shaping unit 34 shapes each attribute value supplied from the attribute value extraction unit 33 according to the output format of the concept dictionary 31 to generate an XML document, and outputs the XML document to the outside.
[0078]
Specifically, an attribute value “fever” for explaining “problem”, an attribute value “2SC7777” for explaining “target”, and an attribute value “class A amplification” for explaining “problem occurrence” are output as “XML output”. An element is formed by substituting the corresponding “$ attribute value” of “format”, and the formed element and structure tag are output in an XML document format.
[0079]
FIG. 9 is a diagram illustrating the XML document having the meaning extracted from the meaning extracting unit 30. Conventionally, a structured document can be generated from a predetermined document. However, when a natural sentence is included in the predetermined document, the natural sentence is output as it is. That is, the conventional structured document includes a natural sentence in the element of the structure tag. On the other hand, the meaning-extracted XML document shown in FIG. 9 is a document in which the meaning of the natural sentence contained in the element of the structure tag is extracted and the extracted meaning is structured. .
[0080]
As described above, the structured document generation apparatus according to the present embodiment performs morphological analysis on a natural sentence included in an XML document and performs each morphological analysis. phrase Part-of-speech types are assigned to and the composition pattern of these part-of-speech types and attribute values phrase Matches the tag pattern representing the part-of-speech type composition of the column and matches phrase Extract attribute values from columns. Then, by outputting the extracted attribute values in accordance with a predetermined XML output format, an XML document in which a natural sentence is structured can be generated.
[0081]
In particular, the structured document generation apparatus prepares a tag pattern for extracting each attribute value for each of various concepts (items) in advance. A concept dictionary can be specified, and attribute values for explaining the concept can be extracted from a natural sentence.
[0082]
Further, the structured document generation apparatus prepares in advance an XML output format of attribute values and structure tags necessary for explaining the concept for each of various concepts (items). By simply outputting in accordance with the XML output format, an XML document in which the extracted attribute values are structured can be easily generated.
[0083]
As shown in FIG. 9, the XML document generated by the structured document generation apparatus is a structured document in which the concept of a natural sentence is described by a structured tag and an attribute value. It is a document that is easy to respond to inquiries. Hereinafter, application examples of the XML document will be described.
[0084]
(Application 1)
A case will be described in which a document describing that “the weight of the product is reduced” is searched from a database storing a large number of documents.
[0085]
FIG. 10A shows a conventional natural sentence document stored in the database, FIG. 10B shows a conventional search result, and FIG. 10C shows the XML document of the present application stored in the database. FIG. The XML document of the present application refers to an XML document generated by the above-described structured document generation apparatus from the conventional natural sentence document shown in FIG. That is, the documents (A) and (C) have the same content, suggesting that the weight of the product increases.
[0086]
Here, when a document describing that “the weight of the product is reduced” is searched by a general method, “product”, “weight”, and “decrease” are usually used as search keywords. When the conventional document is stored in the database, as shown in (B), the conventional document including the words “product”, “weight”, and “decrease” is erroneously searched.
[0087]
On the other hand, when the XML document of the present application is stored in the database, the structure tag may be followed sequentially. Here, a structure tag relating to “product weight” is first searched, and a structure tag indicating whether the weight of the product has increased or decreased is searched from the elements of the structure tag. Specifically, first, <product weight> is searched, and <direction> representing a change in weight is searched from the elements. In the XML document of the present application, the element of <direction> is “increase”, so that it is not erroneously searched.
[0088]
Therefore, the structured document generation apparatus can generate an XML document that does not cause an erroneous search from the document even if the document is erroneously searched by a normal keyword search.
[0089]
(Application example 2)
A case will be described in which an inquiry “What is the current countermeasure against processing defects?” Is made to one document in the database.
[0090]
FIG. 11A shows a conventional natural sentence document stored in the database, and FIG. 11B shows an XML document of the present application stored in the database. The documents (A) and (B) have the same contents.
[0091]
When a conventional document is stored in the database, no response can be made even if an inquiry is made "What is the current countermeasure against processing defects?"
[0092]
On the other hand, when the XML document of the present application is stored in the database, the structure tag related to the inquiry may be followed. Here, first, a structure tag related to “countermeasure” is searched, and elements of the structure tag (lower structure tag and its elements) are extracted. Then, when the extracted structure tag and element are combined, it is possible to answer that the thickness has been increased by 2 mm in response to the above inquiry.
[0093]
Therefore, the structured document generation apparatus can generate an XML document that can easily respond to an inquiry by extracting and structuring the meaning of a natural sentence.
[0094]
The present invention is not limited to the above-described embodiments, and various design changes can be made within the scope described in the claims.
[0095]
For example, the concept dictionary 31 is not limited to the configuration shown in FIG. In the present embodiment, the case where there are three tag patterns in one item concept dictionary has been described as an example, but as many tag patterns as the number of attribute values to be extracted can be provided.
[0096]
In the present embodiment, an XML document has been described as an example. However, for example, an SGML document may be used. At this time, the “XML output format” of the item concept dictionary may be changed to “SGML output format”. In consideration of creating a structured document from a natural sentence, the document format is not particularly limited.
[0098]
【The invention's effect】
Book A structured document generation apparatus and a structured document generation program according to the invention are provided for a phrase-analyzed document. phrase Contains the composition pattern of the part of speech type of the column and the attribute value to be extracted phrase By extracting attribute values by collating with the part-of-speech type composition pattern that represents the composition of the part-of-speech type of the column, and generating the structured document by structuring the extracted attribute value based on the output format information, A structured document from which meaning is extracted can be obtained.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a configuration of a structured document generation apparatus according to an embodiment of the present invention.
FIG. 2 is a block diagram showing a functional configuration of a microcomputer.
FIG. 3 is a block diagram showing a functional configuration of a phrase analysis unit.
FIG. 4 is a diagram illustrating a configuration of a word dictionary.
FIG. 5 is a diagram illustrating a document being analyzed by a phrase analysis unit.
FIG. 6 is a diagram illustrating a document that is analyzed by a phrase analysis unit.
FIG. 7 is a block diagram showing a functional configuration of a meaning extraction unit.
FIG. 8 is a block diagram illustrating a configuration of a concept dictionary.
FIG. 9 is a diagram illustrating a meaning-extracted XML document output from a meaning extraction unit;
10A is a diagram showing a conventional natural sentence document stored in a database, FIG. 10B is a diagram showing a conventional search result, and FIG. 10C is an XML document of the present application stored in the database. FIG.
11A is a diagram showing a conventional natural text document stored in a database, and FIG. 11B is a diagram showing an XML document of the present application stored in the database.
[Explanation of symbols]
20 Phrase analysis part
21 phrase dictionary
22 Content string separator
23 Specific word separation / replacement section
24 Part-of-speech determination part
25 Morphological analyzer
26 Alternative word replacement
27 Compound Word / Phrase Generator
28 Part-of-speech tagging section
30 Semantic extraction part
31 concept dictionary
32 Pattern matching part
33 Attribute value extraction unit
34 Orthopedics

Claims (4)

A document input means for inputting a document including a natural sentence;
A phrase analysis unit that performs phrase analysis on a document including a natural sentence input by the document input unit, and outputs a phrase-analyzed document in which words and parts of speech constituting the document are associated with each other;
A part-of-speech type composition pattern that is provided for each concept representing the meaning content of the word string and that is a part-of-speech type composition of a word string that includes an attribute value to be extracted and represents a pattern corresponding to each concept, and the attribute Output format information for structuring values and corresponding tags and outputting structured elements for each concept, concept dictionary storage means for storing a concept dictionary,
The phrase string composition pattern of the phrase string of the word-parsed document is compared with the part-of-speech type composition pattern of the concept dictionary stored in the concept dictionary storage means, and the phrase string is extracted from the concept dictionary storage means. An attribute value extracting means for selecting a concept dictionary corresponding to the semantic content of the text, and using the selected concept dictionary to extract an attribute value from a phrase string of the phrase analyzed document corresponding to the part of speech type configuration pattern;
Based on the attribute value extracted by the attribute value extracting means and the output format information of the selected concept dictionary, a structured document representing an element structured about the concept corresponding to the selected concept dictionary is generated. Document generating means for
A structured document generation apparatus comprising: Before Symbol attribute value extracting means selects the concept dictionary from among a plurality of concept dictionary stored in the storage means, concept dictionary having a part-of-speech type arrangement pattern matching the part of speech type of configuration patterns of the word analysis document Extract attribute values using the selected concept dictionary,
2. The structured document is generated based on the attribute value extracted by the attribute value extracting unit and the output format information of the concept dictionary selected by the attribute value extracting unit. Structured document generator. Computer
A document input means for inputting a document including a natural sentence;
A phrase analysis unit that performs phrase analysis on a document including a natural sentence input by the document input unit, and outputs a phrase-analyzed document in which words and parts of speech constituting the document are associated with each other;
A part-of-speech type composition pattern that is provided for each concept representing the semantic content of the word string and that is a part-of-speech type composition of a word string that includes an attribute value to be extracted, and that represents a pattern corresponding to the concept, and the attribute Output format information for structuring values and corresponding tags and outputting structured elements for the concept, concept dictionary storage means for storing a concept dictionary,
The phrase string composition pattern of the phrase string of the word-parsed document is compared with the part- of- speech type composition pattern of the concept dictionary stored in the concept dictionary storage means, and the phrase string is extracted from the concept dictionary storage means. An attribute value extracting means for selecting a concept dictionary corresponding to the semantic content of the text, and using the selected concept dictionary to extract an attribute value from a phrase string of the phrase analyzed document corresponding to the part of speech type configuration pattern;
Based on the attribute value extracted by the attribute value extracting means and the output format information of the selected concept dictionary, a structured document representing an element structured about the concept corresponding to the selected concept dictionary is generated. Document generating means for
Structured document generation program to function as The attribute value extraction unit selects a concept dictionary having a part-of-speech type configuration pattern that matches a part-of-speech type configuration pattern of the phrase-analyzed document from a plurality of concept dictionaries stored in the concept dictionary storage unit. , Extract attribute values using the selected concept dictionary,
The document creation unit generates a structured document based on the attribute value extracted by the attribute value extraction unit and the output format information of the concept dictionary selected by the attribute value extraction unit. Structured document generator.

Images