WO2011033653A1 - Dispositif de prise en charge d'unification de termes - Google Patents

Dispositif de prise en charge d'unification de termes Download PDF

Info

Publication number
WO2011033653A1
WO2011033653A1 PCT/JP2009/066379 JP2009066379W WO2011033653A1 WO 2011033653 A1 WO2011033653 A1 WO 2011033653A1 JP 2009066379 W JP2009066379 W JP 2009066379W WO 2011033653 A1 WO2011033653 A1 WO 2011033653A1
Authority
WO
WIPO (PCT)
Prior art keywords
property
term
data
target
unit
Prior art date
Application number
PCT/JP2009/066379
Other languages
English (en)
Japanese (ja)
Inventor
康隆 大嶽
Original Assignee
株式会社東芝
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 株式会社東芝 filed Critical 株式会社東芝
Priority to PCT/JP2009/066379 priority Critical patent/WO2011033653A1/fr
Publication of WO2011033653A1 publication Critical patent/WO2011033653A1/fr

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/20—Natural language analysis
    • G06F40/279—Recognition of textual entities

Definitions

  • the present invention relates to a term unification support device.
  • PLIB Parts Library
  • PLIB is a “content” that describes specific product specifications in accordance with a structure and dictionary for defining “properties” that define data items for expressing specification data of each product and “classes” that define product classifications. Is an international standard that defines the structure for exchanging In ISO 1358-42, “attributes” corresponding to detailed items for explaining what a data item means are standardized.
  • a database for managing contents according to PLIB can be regarded as one of hierarchical databases.
  • the product specification document is mainly composed of “table data” and “supplemental description” for the table.
  • supplemental description for the table.
  • One method can be considered.
  • One is a method in which a sentence is parsed to extract a relationship between words, and an inconsistent term is detected in comparison with a “word-related term unification check rule” (see Patent Document 1).
  • One is a method of substituting the most commonly used term based on the frequency of use of the term (see Patent Document 2).
  • the present invention has been made in view of the above, and provides a term unification support device capable of supporting unification of terms used in supplementary descriptions for specification data into recommended words defined in a standard dictionary. Objective.
  • the present invention is a term unification support device, which is data defined by an attribute including a recommended word and a property defined by the attribute, an acquisition unit that acquires dictionary data for each property, and data
  • a first accepting unit that accepts an input of a technical document including specification data representing items in a matrix format and a supplementary description representing an explanation that supplements the data item; and accepting an input of the first term from terms used in the supplementary description
  • a second reception unit ; an extraction unit for extracting a first data item supplemented by using the first term from the specification data; and a first property corresponding to the first data item using the dictionary data
  • a first estimation unit that estimates the first property, and a property corresponding to the first term out of properties having a relationship with the first property using the dictionary data.
  • a second estimation unit for estimating the properties, by using the dictionary data characterized in that it comprises a verification unit for the second property matching with the recommended word and the first term owned.
  • the figure which illustrates the hardware constitutions of a term unification assistance apparatus The figure which illustrates the functional structure of a term unification assistance apparatus.
  • the figure which illustrates a technical document. The flowchart which shows the procedure of the process which registers an index.
  • the flowchart which shows the procedure of the process which estimates a supplement property The figure which shows the correspondence of the column in an specification data, and an attribute.
  • the flowchart which shows the procedure of the process which estimates object term property The flowchart which shows the procedure of the process of step S42.
  • the figure which shows the example of a display of a correction candidate The figure which shows the example of a display of supplementary description omission.
  • the term unification support device 10 includes a control unit 101 such as a CPU (Central Processing Unit) that controls the entire device, a ROM (Read Only Memory) 104 that stores various data and various programs, a RAM (Random Access Memory) 105, and the like.
  • a control unit 101 such as a CPU (Central Processing Unit) that controls the entire device, a ROM (Read Only Memory) 104 that stores various data and various programs, a RAM (Random Access Memory) 105, and the like.
  • the term unification support device 10 includes a display unit 103 that displays information, an operation unit 102 such as a keyboard and a mouse that accepts user instruction input, and a communication I / F (interface) that controls communication with an external device.
  • the term unification support apparatus 10 edits a technical document including specification data represented in a matrix format for each data item including at least one element and supplementary description representing a description supplementing the data item. It is connected to the document editing apparatus via a communication I / F.
  • FIG. 2 is a diagram illustrating a functional configuration of the term unification support apparatus 10.
  • the term unification support apparatus 10 includes a document reception unit 21, a target term reception unit 22, a supplement target extraction unit 23, a classification reception unit 24, a supplement target property estimation unit 25, a target term property estimation unit 26, and a vocabulary.
  • a collation unit 27, a collation result output unit 28, a supplementary description omission output unit 29, and a dictionary storage unit 30 are included.
  • the collation result output unit 28 and the supplementary description omission output unit 29 are generated on the main storage unit such as the RAM 105 when the control unit 101 executes the program.
  • the dictionary storage unit 30 is stored in the auxiliary storage unit 107, for example.
  • the dictionary storage unit 30 has a dictionary DB and an index table.
  • the dictionary DB stores dictionary data.
  • the dictionary data is defined by a class indicating the classification and hierarchy of products and a property that characterizes the class.
  • a property is defined by at least one attribute.
  • Products are provided to the market to satisfy users' needs and desires, and include not only tangible items but also services, experiences, events, people, places, assets, organizations, information, ideas, etc. Included in the product.
  • FIG. 3 is a diagram illustrating a data structure of the dictionary DB. In the figure, the dictionary DB has a hierarchical structure centered on classification (class).
  • the classification "COMPONENT” is the root class, "Semiconductor” and “General electronic components” exist as its child classes, and "General-purpose logic IC” as its child class in “Semiconductor” And “transistor”. That is, it is shown that “semiconductor” is subdivided into two categories “general-purpose logic IC” and “transistor”, and “transistor” is further subdivided into two categories “MOSFET” and “IGBT”. .
  • the hierarchical structure may have any number of layers.
  • the symbol written in parentheses (in this example, Cn, n is a number) is a code representing a class uniquely assigned in the dictionary storage unit 30, and is an identifier called “class BSU code” in the PLIB standard. It corresponds to.
  • each class defining a hierarchical structure has a unique property (a part surrounded by a dotted line in FIG. 3).
  • Each class inherits the properties defined in the higher class of the class.
  • the properties of class “MOSFET” are the properties defined in that class (“P3”, “P4”, “P5”, “P6”) and the properties defined in its upper class “transistor” (in this example, simplified) That is, the property defined by the class “semiconductor” (there is no simplification in this example) and the property defined by the class “COMPONENT” (“P0”).
  • a code (“property BSU code” or simply “BSU”) representing a property uniquely assigned in the dictionary DB is assigned to the property.
  • BSU code simply “BSU”
  • both the code representing the class and the code representing the property are handled as the BSU code.
  • FIG. 4 is a diagram illustrating dictionary data for each property. Each row except the top row is a record corresponding to each dictionary data for each property.
  • Each dictionary data includes a BSU code for identifying each property, a BSU code for identifying a class in which the property is defined (referred to as a definition class), and an attribute that defines the property of the property.
  • the attributes include, for example, “definition class” “BSU” “recommended name” “dependent property” “synonym” “symbol” “unit” “definition”.
  • the “dependent property” indicates a dependency relationship between the properties, and indicates that the property value is affected by the value of the property on the dependent side.
  • P3 is defined as a dependent property of P4, which indicates that “P3: DC drain current” is a value determined depending on “P4: channel temperature”. This also indicates that the property P3 and the property P6 have the same dependency.
  • attributes such as “data type” of a property and “source” indicating the source of definition of the property itself also exist in the PLIB.
  • the dictionary structure such as class parent / child structure, property and attribute type, dependency between properties, data type, etc.
  • PLIB Parts Library
  • the index table stores an index for each word element of each property or class.
  • FIG. 5 is a diagram illustrating an index table.
  • the index table includes “class / property classification (C / P)”, “BSU code”, “word element”, “attribute type”, and “weight” as an index for each word element for each property.
  • Class / property classification (C / P)” is for distinguishing between classes and properties.
  • the “word element” is a character string represented as a value of a specific attribute such as “recommended name”, “definition”, “synonym”, etc., for each word.
  • the “weight” is given according to the type of attribute.
  • This index is used for property estimation performed by the supplement target property estimation unit 25 described later. In this example, it is assumed that only properties are registered in the index table.
  • FIG. 6 is a diagram illustrating a technical document.
  • the technical document includes specification data expressed in a matrix format for each of a plurality of data items including elements such as “item name”, “symbol”, “rating”, and “unit”, and “Note 1”. ”And“ Note 2 ”, and supplementary description representing a supplementary explanation of the data item is included.
  • the top line of the specification data is a heading. “Note 1” in the supplementary description is a supplement to the explanation of the data item because “DC drain current” is linked to the data item of the specification data.
  • the target term accepting unit 22 accepts input of terms (referred to as target terms) that are used in supplementary descriptions included in the target document and that are designated and transmitted by the user in the document editing apparatus 11 and are to be collated. .
  • the classification receiving unit 24 receives an input of a product class related to the target document, which is designated and transmitted by the user in the document editing apparatus 11.
  • the supplementary target extracting unit 23 uses the target terms received by the target term accepting unit 22 in the supplementary description in the specification data included in the technical document accepted by the document accepting unit 21 to supplement the explanation.
  • the target data item (referred to as supplementary target item) is extracted.
  • the supplement target extraction unit 23 extracts a data item associated with the explanation in which the target term is used in the supplement description.
  • the supplement target property estimation unit 25 uses the supplement target items extracted by the supplement target extraction unit 23, the class that the classification reception unit 24 has received input, and dictionary data stored in the dictionary DB of the dictionary storage unit 30. Thus, it is estimated which dictionary data the supplement target item corresponds to in the dictionary DB, that is, which property the supplement target item corresponds to in the dictionary DB.
  • the target term property estimation unit 26 uses the property estimated by the supplement target property estimation unit 25 (referred to as a supplement target property) and the class received by the classification reception unit 24, and the target term reception unit 22 receives the input. Estimate the properties for the target terms (called target term properties).
  • the supplementary target property estimation unit 25 narrows down the target term property to a property that is related to the supplementary target property in the PLIB structure.
  • Search for There are the following two methods for searching for the target term property.
  • Method 1 When the property to be supplemented has no dependency relationship with other properties 1) Specify the class to which the property to be supplemented belongs 2) The class specified in 1) and the class having a vertical relationship with the class, that is, 1 ), A set of properties belonging to any of the class specified in (1), all classes below the class, and all classes above the class is defined as the search target Q. 3) Among the search targets Q, the “target term” is a keyword.
  • a property having a dependency relationship with the property to be supplemented that is, a property for which the property to be supplemented is defined as a “dependent property” and a property to be supplemented
  • the supplement target property estimation unit 25 that performs a similar search of dictionary data for each property using “target term” as a keyword in the search target R, using the index table as the search target R.
  • the property obtained as a result of the similar search executed by the method is output with ranking. The higher the ranking, the higher the degree of the target term property is estimated.
  • the vocabulary collating unit 27 collates the recommended word included in the target term property estimated by the target term property estimating unit 26 with the target term received by the target term receiving unit 22.
  • the collation result output unit 28 transmits the collation result of the vocabulary collation unit 27 to the document editing apparatus 11. Specifically, if the recommended word does not match the target term as a result of the collation by the vocabulary collating unit 27, the collation result output unit 28 displays the first collation result indicating the recommended word as a correction candidate for the target term. It is transmitted to the document editing apparatus 11.
  • all or part of the attributes related to the name are used in the supplementary description for the supplementary target item.
  • a property that is not included (referred to as a supplemental missing property) is searched, and a second matching result indicating a supplementary missing property recommendation word as a supplementary description missing is transmitted to the document editing apparatus 11 according to the search result.
  • the term unification support apparatus 10 stores each dictionary data illustrated in FIG. 3 as dicElem (i, j) in the main storage unit (step S1).
  • i represents a dictionary data record corresponding to each row except the top row
  • j represents an attribute corresponding to each column.
  • Step S4 Each of the character strings dic (i, j) represented as attribute values is analyzed for words (Step S4).
  • the character string is “junction temperature”. In this case, it is decomposed into “junction” and “temperature” by word analysis.
  • the term unification support apparatus 10 performs weighting according to the type of the attribute for all the word elements thus decomposed, and assigns the weight value (w (j)) for each property and each word element.
  • w (j) weight value for each property and each word element.
  • each index illustrated in FIG. 5 is registered in the index table for each word element.
  • a word element having a weight of “0” is not registered.
  • the index table in which the index is registered in this way is used in the term unification support process described below.
  • the data structure of the dictionary DB included in the dictionary storage unit 30 is as illustrated in FIGS.
  • the user designates a technical document (target document), designates a term (target term) to be collated among terms included in the supplementary description contained in the target document, and designates a class.
  • the document editing apparatus 11 transmits data including the designated target document, target term, and class to the term unification support apparatus 10.
  • the technical document illustrated in FIG. 5 is designated, “junction temperature” is designated as a target term, and “transistor” is designated as a class.
  • the class designation may be performed by, for example, a BSU code or a class name.
  • the term unification support device 10 When the term unification support device 10 receives the data transmitted from the document editing device 11, the term unification support device 10 receives the input of the target document by the function of the document reception unit 21, and inputs the target term by the function of the target term reception unit 22. A class input is received by the function of the reception and classification reception unit 24 (step S20). Next, the term unification support device 10 uses the target term whose input is accepted in step S20 among the specification data included in the subject document whose input is accepted in step S20 by the function of the supplementary target extraction unit 23. A target data item (supplementation target item) for which explanation is supplemented is extracted (step S21).
  • the supplement target item since the data item associated with “Note 1” for which explanation is supplemented by using “junction temperature” is “DC drain current”, this is the supplement target item. That is, the third line of the specification data included in the technical document illustrated in FIG. 6 is extracted as a supplement target item.
  • the term unification support apparatus 10 uses the function of the supplement target property estimation unit 25 to store the supplement target items extracted in step S2, the class accepted in step S1, and the dictionary DB of the dictionary storage unit 30. Using the stored dictionary data, it is estimated which dictionary data the supplement target item corresponds to in the dictionary DB, that is, which property the supplement target item corresponds to in the dictionary DB (step S22).
  • the term unification support apparatus 10 estimates which property in the dictionary DB corresponds to the supplement target item as follows. First, the term unification support apparatus 10 selects a property to be compared in the dictionary DB (step S30). Next, the term unification support device 10 applies the following rule, and calculates the score as the similarity for each column in which the attribute is estimated among the columns belonging to the supplement target item (step S31). (a) When the character string exactly matches the attribute in the dictionary data: 5 points (b) When the character string partially matches the attribute in the dictionary data: 1 point (c) No distinction is made regarding “name” (“recommended name” “synonym”)
  • step S32 When the similarity calculation is completed for all the columns (step S32: YES), the term unification support apparatus 10 selects the property having the highest total score and determines this as the similar property (step S33). This similar property is estimated as a property corresponding to the supplement target item (supplement target property).
  • the term unification support apparatus 10 uses the supplement target property estimated in step S22 and the class whose input is accepted in step S20 by the function of the target term property estimation unit 26, and in step S20.
  • a property (target term property) for the target term for which the input has been accepted is estimated (step S23).
  • the term unification support apparatus 10 refers to the “dependent property” in each dictionary data stored in the dictionary DB, and determines whether or not the supplement target property estimated in step S22 has a dependency relationship with another property. Judgment is made (step S40). That is, the term unification support apparatus 10 determines whether or not a property whose supplement target property is defined as “dependent property” is stored in the dictionary DB as dictionary data. If the determination result is affirmative (step S40: YES), the term unification support apparatus 10 first performs the processing according to [Method 2] described above, and first, the supplement target property and the property having a dependency relationship with the supplement target property. Is set as the search target R (step S41).
  • the term unification support apparatus 10 determines that the supplement target property estimated in step S22 has a dependency relationship with other properties, and therefore the supplement target property “P3” and the property “P4” that has a dependency relationship with the supplement target property. ”,“ P6 ”(“ P3 ”,“ P4 ”,“ P6 ”) are defined as the search target R.
  • the term unification support apparatus 10 estimates the property for the target term by performing a similar search of the property included in the search target R using the target term for which the input has been accepted in step S20 as a keyword, using the index table. (Step S42).
  • the detailed procedure of the process of step S42 is demonstrated using FIG.
  • the term unification support device 10 performs word analysis on the target term (step S50), and searches the index table for the property including the word element obtained as a result in the index table (step S51). Specifically, the target term “junction temperature” is broken down into word elements of “junction” and “temperature”. In the index table illustrated in FIG. 5, indexes IDX3 to IDX5 including these word elements are obtained as a result of the search.
  • the term unification support apparatus 10 groups the indexes obtained as a result of the search for each property, and calculates the sum of weights (score) (step S52).
  • the term unification support apparatus 10 ranks the calculation results for each property in descending order and outputs each property with a ranking (step S53). The higher the ranking, the higher the degree of the target term property is estimated.
  • the property “P4” is output as the first place, and this is estimated as the target term property.
  • the property “P1” of the dictionary data illustrated in FIG. 4 is not included even though the recommended name “junction temperature” is included as an attribute. It is in.
  • the property “P1” belongs to the class “general-purpose logic IC”.
  • the property corresponding to the supplement target item “DC drain current” for which explanation is supplemented using “junction temperature” is “P3”, and the class to which the property belongs is “MOFSET”. Therefore, the property “P4” belonging to the same class is estimated as the target term property, and the property “P1” is excluded from the target, so that the target term property is correctly estimated.
  • step S40 when the supplementary target property estimated in step S22 does not have a dependency relationship with other properties (step S40: NO), the term unification support apparatus 10 performs processing according to the above [Method 1].
  • the class to which the supplement target property belongs is specified (step S43).
  • the term unification support apparatus 10 determines a set of properties belonging to any of the class identified in step S43, all classes lower than the class, and all classes higher than the class as the search target Q (step S44). ).
  • the term unification support apparatus 10 estimates the target term property by performing a similar search of the property included in the search target Q using the index term using the target term that has been accepted in step S20 as a keyword ( Step S45).
  • the detailed procedure of step S45 is substantially the same as step S42, and is substantially the same as that described with reference to FIG.
  • the term unification support apparatus 10 collates the recommended word included in the target term property estimated in step S23 with the target term accepted in step S1 by the function of the vocabulary matching unit 27. (Step S24). If they do not match (step S25: NO), the term unification support apparatus 10 gives the document editing apparatus 11 a first matching result indicating the recommended word as a correction candidate for the target term by the function of the matching result output unit 28. Transmit (step S26).
  • the recommended word is “channel temperature”. This is not consistent with the subject term “junction temperature”.
  • the first collation result indicated by this “channel temperature” as a correction candidate is transmitted to the document editing apparatus 11. Then, when receiving the first collation result, the document editing apparatus 11 uses the first collation result to display “channel temperature” as a correction candidate for “junction temperature” as illustrated in FIG. To display.
  • the term unification support apparatus 10 uses the function of the supplementary description omission output unit 29 to execute the supplementary target item among the other properties.
  • a property (supplementary omission property) in which all or part of the attribute (“recommended name”, “synonym”, “definition”, etc.) relating to the name is not used is searched. If there is a supplemental missing property as a result of the search (step S27: YES), the term unification support apparatus 10 transmits a second collation result indicating the recommended word of the supplemental missing property as a supplemental description missing to the document editing apparatus 11. (Step S28).
  • the supplementary target property “P3” is defined as the “dependent property” of the properties “P4” and “P6”. Therefore, the supplement target property “P3” has a dependency relationship with the properties “P4” and “P6”.
  • the character string that is the value of the attribute related to the name such as the recommended word is not included in “Note 1” in the supplementary description. For this reason, the term unification support apparatus 10 transmits to the document editing apparatus 11 the second collation result indicating the recommended word of the property “P6” as the supplementary description omission in “Note 1”.
  • the document editing apparatus 11 Upon receiving the second collation result, the document editing apparatus 11 uses this to point out the possibility of a supplementary description omission in “P6: gate leakage current” in “Note 1” as illustrated in FIG.
  • the message is displayed on the display device.
  • the property that corresponds to the target data item (supplement target item) for which the term (target term) included in the supplementary description for the specification data supplements the explanation is estimated.
  • a property (target term property) corresponding to the target term is estimated, and correction candidates for the target term are presented.
  • various programs executed by the term unification support apparatus 10 may be stored on a computer connected to a network such as the Internet and provided by being downloaded via the network.
  • the various programs are recorded in a computer-readable recording medium such as a CD-ROM, a flexible disk (FD), a CD-R, a DVD (Digital Versatile Disk), etc. in a file that can be installed or executed.
  • the computer program product may be provided.
  • the term unification support device 10 is configured to have the dictionary DB.
  • the term unification support device 10 is not limited to this, and is configured such that other information processing devices have the dictionary DB. May access the dictionary DB via the other information processing apparatus to acquire the corresponding dictionary data.
  • the dictionary DB conforms to the PLIB.
  • the present invention is not limited to this, and the dictionary DB may be applied to a database other than the PLIB.
  • the direction corresponding to the attribute is assumed to be the column direction and the direction corresponding to the property is the row direction, but these may be reversed.
  • the term unification support apparatus 10 receives an input of a designated class by the function of the classification receiving unit 24, and estimates a supplement target property only for properties belonging to the class.
  • the class may not be specified, and in this case, the supplement target property may be estimated using all the dictionary data stored in the dictionary DB as a processing target.
  • the term unification support apparatus 10 determines whether or not the supplement target property has a dependency relationship with other properties when estimating the property for the target term by the function of the target term property estimation unit 26.
  • the process was divided into [Method 1] and [Method 2]. However, the processing may be performed only by [Method 1] without considering whether or not the supplement target property has a dependency relationship with other properties.
  • the term unification support apparatus 10 does not perform the processes of steps S40 to S41 but performs the processes of steps S43 to S45.
  • step S43 for the supplementary property “P3: DC drain current”, in the example of FIG. 4, the class is specified as “MOFSET” in step S43.
  • step S44 the corresponding properties are “P1”, “P3”, “P4”, “P5”, “P6”, and the search target Q is the properties “P1”, “P3”, “P4”, “P5”. , Defined as a set of “P6”.
  • step S45 after step S50 of FIG. 12, in step S51, indexes IDX1 to IDX5 including the word elements “junction” and “temperature” in the index table illustrated in FIG. 5 are obtained as a result of the search. It is done.
  • step S53 the property “P4” is output as the first and the property “P1” is output as the second, and these are estimated as candidates for the target term property.
  • step S24 of FIG. 8 each recommended word included in each target term property candidate estimated in step S23 (step S53 in FIG. 12) is compared with the target term received in step S1. If they do not match, in step S26, a first collation result indicating each recommended word as a correction candidate for the target term is transmitted to the document editing apparatus 11.
  • the first collation result recommended words that do not match the target term for the ranked properties of the plurality of ranks can be indicated as correction candidates.
  • the document editing apparatus 11 Upon receiving the first collation result, the document editing apparatus 11 uses the first collation result as the correction candidate 1 in the order of ranking, as shown in FIG. And “junction temperature” for the property “P1” as correction candidate 2 is displayed on the display device.
  • the threshold value of the score calculated in step S52 may be determined in advance, and the recommended word of the score property that is equal to or lower than the threshold value may not be a correction candidate even if it does not match the target term.

Abstract

L'invention porte sur un dispositif de prise en charge d'unification de termes qui acquiert des données qui sont définies par des attributs comprenant des termes recommandés et des propriétés définies par les attributs, les données étant des données de dictionnaire en rapport avec chaque propriété, et reçoit en entrée : un document technique contenant des données de spécification exprimant des éléments de données sous une forme matricielle et des descriptions supplémentaires exprimant des explications complétant les éléments de données ; et un premier terme parmi des termes à utiliser dans la description supplémentaire. Ensuite, le dispositif de prise en charge d'unification de termes extrait, à partir des données de spécification, un premier élément de données complété à l'aide du premier terme, estime une première propriété équivalente au premier élément de données à l'aide des données de dictionnaire, estime une seconde propriété correspondant au premier terme parmi les propriétés ayant une relation avec la première propriété à l'aide des données de dictionnaire, et vérifie le mot recommandé possédant la seconde propriété avec le premier terme à l'aide des données de dictionnaire.
PCT/JP2009/066379 2009-09-18 2009-09-18 Dispositif de prise en charge d'unification de termes WO2011033653A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/JP2009/066379 WO2011033653A1 (fr) 2009-09-18 2009-09-18 Dispositif de prise en charge d'unification de termes

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/JP2009/066379 WO2011033653A1 (fr) 2009-09-18 2009-09-18 Dispositif de prise en charge d'unification de termes

Publications (1)

Publication Number Publication Date
WO2011033653A1 true WO2011033653A1 (fr) 2011-03-24

Family

ID=43758279

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2009/066379 WO2011033653A1 (fr) 2009-09-18 2009-09-18 Dispositif de prise en charge d'unification de termes

Country Status (1)

Country Link
WO (1) WO2011033653A1 (fr)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH06176018A (ja) * 1992-12-07 1994-06-24 Fuji Xerox Co Ltd 校正支援装置
JPH07325826A (ja) * 1994-05-31 1995-12-12 Meidensha Corp 日本語処理システム
JPH09259125A (ja) * 1996-03-27 1997-10-03 Hitachi Kokusai Business:Kk 文書作成支援システム及び用語辞書
JP2007026074A (ja) * 2005-07-15 2007-02-01 Toshiba Corp 部品情報表示装置、部品情報表示システム、部品情報表示プログラムおよび部品情報表示方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH06176018A (ja) * 1992-12-07 1994-06-24 Fuji Xerox Co Ltd 校正支援装置
JPH07325826A (ja) * 1994-05-31 1995-12-12 Meidensha Corp 日本語処理システム
JPH09259125A (ja) * 1996-03-27 1997-10-03 Hitachi Kokusai Business:Kk 文書作成支援システム及び用語辞書
JP2007026074A (ja) * 2005-07-15 2007-02-01 Toshiba Corp 部品情報表示装置、部品情報表示システム、部品情報表示プログラムおよび部品情報表示方法

Similar Documents

Publication Publication Date Title
US7996437B2 (en) Program for mapping of data schema
US10296579B2 (en) Generation apparatus, generation method, and program
US7814102B2 (en) Method and system for linking documents with multiple topics to related documents
US8005815B2 (en) Search engine
US9280535B2 (en) Natural language querying with cascaded conditional random fields
Hornung et al. Recommendation based process modeling support: Method and user experience
CN109145110B (zh) 标签查询方法和装置
Bjarnadóttir The database of modern Icelandic inflection (Beygingarlýsing íslensks nútímamáls)
US7853595B2 (en) Method and apparatus for creating a tool for generating an index for a document
JP2016099741A (ja) 情報抽出支援装置、方法およびプログラム
US20120179709A1 (en) Apparatus, method and program product for searching document
CN111143370B (zh) 用于分析多个数据表之间关系的方法、设备和计算机可读存储介质
CN102227723A (zh) 辅助误译的检测的装置及方法
JP2007102309A (ja) 自動分類装置
JP2010134709A (ja) 語彙誤り検出装置及び語彙誤り検出方法
JP2007172260A (ja) 文書ルール作成支援装置および文書ルール作成支援方法並びに文書ルール作成支援プログラム
Tang et al. iASA: learning to annotate the Semantic Web
CN112818005B (zh) 结构化数据的搜索方法、装置、设备及存储介质
WO2011033653A1 (fr) Dispositif de prise en charge d'unification de termes
JP2009043029A (ja) 関連ｄｂ作成装置
US20230177362A1 (en) Risk assessment apparatus, risk assessment method, and program
CN113988057A (zh) 基于概念抽取的标题生成方法、装置、设备及介质
JP4965766B2 (ja) 関係情報抽出装置および属性情報抽出装置
JP2009104475A (ja) 類似文書検索装置、類似文書検索方法およびプログラム
US20160283605A1 (en) Information extraction device, information extraction method, and display control system

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 09849512

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 09849512

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: JP