JP2006331001A - Specialist extraction device and dictionary provision device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a specialist extraction device and a dictionary server using a technical term dictionary extensively registered with an unknown term used by a number of users without requiring registration of the unknown term in the technical term dictionary to the user. <P>SOLUTION: This specialist extraction device extracts the unknown term not registered in the technical term dictionary (S101), cuts out a sentence or a clause including the unknown term, and sends it to a dictionary server as unknown term data together with the unknown term (S102). The dictionary server decides whether it is registered in the technical term dictionary (S103) or has to be set as a technical term (S104), and classifies it when it has to be set as the new technical term (S105). They are repeated until processing of all the pieces of the unknown term data is completed (S106). Thereafter, the unknown term set as the technical term is registered in the technical term dictionary or the like and is managed (S107), charging evaluation is performed according to an unknown term data amount (S108), and the new technical term dictionary is sent to the specialist extraction device to perform charging (S109). Thereby, the technical term dictionary extensively storing the unknown term is produced. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to an expert extraction device that extracts a person who is familiar with certain knowledge and information by extracting and analyzing words contained in a document, and a dictionary providing device (dictionary server) that provides a dictionary used in the device. .

In general, knowledge and information obtained by working together with a large number of people are rarely managed intensively and are often held in pieces by individuals and small groups. Here, if the document files created by each individual are stored in a database and can be searched using specific terms related to certain knowledge and information, the owner of the document file, creation / editing, etc. By referring to the person or the like, it should be possible to obtain information on the person who is familiar with the knowledge and information. However, when a large number of document files are accumulated, it is determined whether or not the owner of the document file in which a specific term obtained as a result of the search is described is a person who has a deep knowledge of desired knowledge or information. It is often difficult to do in practice. Therefore, conventionally, an expert extraction device has been proposed that can search a person (expert) who is familiar with desired knowledge and information so that the details can be judged by specifying a word or field. (See Patent Document 1). Conventionally, a savvy person extraction device that automatically extracts persons who are familiar with the contents of a document based on the contents of the document such as an electronic mail has been proposed (see Patent Document 2).
JP 2002-14971 A JP 2002-560001 A

  The conventional expert extraction apparatus as described above is provided with a technical language dictionary in which a large number of specific technical terms related to knowledge and information are recorded, and a search is performed based on this technical language dictionary. Therefore, new technical words (hereinafter referred to as “unknown words”) that are not recorded in this technical language dictionary must be newly registered in the technical language dictionary. However, if this unknown word is automatically registered in the technical language dictionary, there is a disadvantage that it is registered even if it is a typo. Further, if this unknown word is extracted or collectively at a certain point in time, if the user has to perform this registration work, it will be very troublesome for the user. Furthermore, unknown words registered in this manner are frequently used only within a certain organization, and may become a biased technical term dictionary.

  Therefore, the present invention provides an expert extraction device that uses a vocabulary dictionary in which unknown words widely used by many users are registered without registering unknown words in the vocabulary dictionary, and the vocabulary dictionary An object of the present invention is to provide a dictionary providing device that gives

The first invention is an expert extraction device for extracting an expert who is presumed to be familiar with a predetermined technical term based on a predetermined electronic document,
A dictionary in which multiple technical terms are registered,
By referring to the dictionary, the terminology extracting means for extracting the terminology included in the electronic document and extracting an unknown word not registered in the dictionary;
A totaling unit that counts the association frequency including at least one of the number of extractions of the technical terms extracted by the technical term extraction unit and a value corresponding to the number of extractions in association with the person related to the electronic document;
An input means for accepting a technical term or specialized field designated by a predetermined user;
The expert is searched by searching for a person who is associated with the highest relevance frequency by the tabulating means with respect to the technical word accepted by the input means or the technical term corresponding to the specialized field accepted by the input means. A search means for extracting
And an unknown word providing means for providing the unknown word extracted by the technical word extracting means to a predetermined dictionary providing device outside the apparatus.

According to a second invention, in the first invention,
The technical term extraction means extracts a plurality of character strings including the unknown word,
The unknown word providing means provides a part or all of the plurality of character strings extracted by the technical word extracting means to the dictionary providing apparatus.

According to a third invention, in the second invention,
It further comprises delimiter setting means for setting a range to be given to the dictionary providing device among the plurality of character strings based on a delimiter method that is alternatively selected by the user from a plurality of predetermined delimiter methods. It is characterized by.

According to a fourth invention, in the second invention,
Character string selection means is further provided for allowing the user to select a character string to be given to the dictionary providing device among the plurality of character strings.

According to a fifth invention, in the first invention,
The technical term extraction means extracts an already extracted unknown word included in the electronic document as the technical term by referring to the already extracted unknown word.

According to a sixth invention, in the first invention,
When a new dictionary to replace the dictionary is received from the dictionary providing device and the unknown word is registered in the new dictionary, the degree of association of the unknown word and the associated person are determined as the new dictionary. And a dictionary receiving means for taking over as the association degree of the technical word corresponding to the unknown word registered in the above and the associated person.

A seventh invention is the sixth invention, wherein
Further comprising specialized field data in which the correspondence between the specialized field and the technical term is registered;
The dictionary receiving means receives new specialized field data to be substituted for the specialized field data from the dictionary providing device, and when the unknown word is registered in the new specialized field data, the association frequency of the unknown word And the associated person is taken over as the related degree of the technical term corresponding to the unknown word corresponding to the specialized field registered in the new specialized field data and the associated person.

In an eighth aspect based on the first aspect,
The unknown word providing means includes encryption processing means for encrypting data including the unknown word to be given to the dictionary providing apparatus.

In a ninth aspect based on the eighth aspect,
The encryption processing means encrypts the data based on a predetermined public key corresponding to the dictionary providing device by using a public key cryptosystem.

In a tenth aspect of the present invention, a dictionary in which a plurality of technical terms are registered, and the technical terms included in a predetermined electronic document are extracted by referring to the dictionary, and unknown words that are not registered in the dictionary are extracted. The unknown word is received from an expert extraction device comprising an expert word extraction means and an unknown word providing means for giving the unknown word extracted by the expert word extraction means to the outside of the device, and a newly created dictionary is converted into the specialized dictionary. A dictionary providing device for a home extraction device,
An unknown word receiving means for receiving the unknown word from the expert extraction device;
A dictionary creating means for creating a new dictionary including part or all of the received unknown words;
And a dictionary providing means for providing the expert extracting device with a new dictionary created by the dictionary creating means.

In an eleventh aspect based on the tenth aspect,
The dictionary creating means determines whether or not the unknown word received by the unknown word receiving means is a technical word, and only when it is determined that the unknown word is a technical word, specialized word analyzing means for registering in the new dictionary It is characterized by including.

In a twelfth aspect based on the eleventh aspect,
The apparatus further comprises customer data evaluation means for performing a discount evaluation on a fee for providing the new dictionary according to an amount corresponding to the unknown word received by the unknown word receiving means.

In a thirteenth aspect based on the tenth aspect,
The unknown word receiving means receives a plurality of character strings including the unknown word,
The customer data evaluation unit performs the discount evaluation according to an amount corresponding to the plurality of character strings.

In a fourteenth aspect based on the twelfth or thirteenth aspect,
The customer data evaluation means performs the discount evaluation according to an amount corresponding to an unknown word determined to be a technical word by the dictionary creation means among the unknown words received by the unknown word receiving means. Features.

A fifteenth aspect of the invention is an expert extraction device for extracting an expert who is presumed to be familiar with a predetermined technical term based on a predetermined electronic document, and a dictionary in which a plurality of technical terms are registered, and the dictionary Extracting the technical terms contained in the electronic document by referring to the technical terms extracting means for extracting unknown words not registered in the dictionary, and extracting the technical terms extracted by the technical terms extracting means A counting means for counting the association frequency including at least one of the number of times and a value corresponding to the number of extractions in association with a person related to the electronic document, and a technical term or specialized field designated by a predetermined user. For the input means to be accepted and the technical terms accepted by the input means or the technical terms corresponding to the specialized field accepted by the input means, Search means for extracting the expert by searching for persons associated with the frequency, and the unknown word extracted by the technical word extraction means to a predetermined dictionary providing device outside the device via a predetermined communication line An expert extraction device comprising an unknown word providing means
A dictionary in which a plurality of technical terms are registered, and a technical term extraction unit that extracts the technical terms contained in a predetermined electronic document by referring to the dictionary and extracts unknown words that are not registered in the dictionary; A dictionary that receives the unknown word from an expert extraction device including an unknown word providing unit that supplies the unknown word extracted by the technical word extraction unit to the outside of the device, and gives a newly created dictionary to the expert extraction device An apparatus for providing an unknown word receiving means for receiving the unknown word from the expert extraction device via the communication line and a dictionary creation for creating a new dictionary including a part or all of the received unknown word And a dictionary providing device comprising: a dictionary providing device for providing the new dictionary created by the dictionary creating device to the expert extracting device via the communication line. And wherein the door is a dictionary updating system.

A sixteenth aspect of the invention is a method for controlling an expert extraction apparatus that extracts an expert who is presumed to be familiar with a predetermined technical term based on a predetermined electronic document,
A technical term extraction step of extracting the technical terms contained in the electronic document by referring to a dictionary in which a plurality of technical terms are registered, and extracting unknown words that are not registered in the dictionary;
A tabulation step of tabulating the number of associations including at least one of the number of extractions of the technical terms extracted in the technical term extraction step and a value corresponding to the number of extractions in association with a person related to the electronic document;
An input step for accepting a technical term or specialized field designated by a predetermined user;
The expert is searched by searching for a person who is associated with the highest relevance frequency by the aggregation step with respect to the technical word accepted in the input step or the technical term corresponding to the technical field accepted in the input step. A search step to extract
An unknown word providing step of providing the unknown word extracted in the technical word extracting step to a predetermined dictionary providing device outside the device.

According to a seventeenth aspect of the present invention, there is provided an expert extraction apparatus for extracting an expert who is presumed to be familiar with a predetermined technical term based on a predetermined electronic document.
A technical term extraction step of extracting the technical terms contained in the electronic document by referring to a dictionary in which a plurality of technical terms are registered, and extracting unknown words that are not registered in the dictionary;
A tabulation step of tabulating the number of associations including at least one of the number of extractions of the technical terms extracted in the technical term extraction step and a value corresponding to the number of extractions in association with a person related to the electronic document;
An input step for accepting a technical term or specialized field designated by a predetermined user;
The expert is searched by searching for a person who is associated with the highest relevance frequency by the aggregation step with respect to the technical word accepted in the input step or the technical term corresponding to the technical field accepted in the input step. A search step to extract
A control program for executing an unknown word providing step of supplying the unknown word extracted in the technical word extraction step to a predetermined dictionary providing device outside the device.

According to an eighteenth aspect of the present invention, there is provided an expert extraction apparatus for extracting an expert who is presumed to be familiar with a predetermined technical term based on a predetermined electronic document.
A technical term extraction step of extracting the technical terms contained in the electronic document by referring to a dictionary in which a plurality of technical terms are registered, and extracting unknown words that are not registered in the dictionary;
A tabulation step of tabulating the number of associations including at least one of the number of extractions of the technical terms extracted in the technical term extraction step and a value corresponding to the number of extractions in association with a person related to the electronic document;
An input step for accepting a technical term or specialized field designated by a predetermined user;
The expert is searched by searching for a person who is associated with the highest relevance frequency by the aggregation step with respect to the technical word accepted in the input step or the technical term corresponding to the technical field accepted in the input step. A search step to extract
A computer-readable recording medium recording a program for executing an unknown word providing step of supplying the unknown word extracted in the technical word extraction step to a predetermined dictionary providing device outside the device.

  According to the first invention, the user provides the unknown word extracted by the technical word extraction means to the dictionary providing device outside the apparatus by the unknown word providing means without registering the unknown word in the technical word dictionary. It is possible to use a specialized word dictionary in which unknown words used by many users created by the dictionary providing apparatus are widely registered.

  According to the second invention, part or all of a plurality of character strings including the unknown word extracted by the technical word extraction means is given to the dictionary providing apparatus by the unknown word providing means. An accurate vocabulary dictionary can be created by referring to the character string received at.

  According to the third aspect, the delimiter setting means sets the range to be given to the dictionary providing device based on the delimiter method selected by the user, so that the dictionary is provided among the character strings including unknown words. A character string that should be allowed to be given to the device can be easily selected by the user of the device, thereby preventing, for example, a sentence requiring confidentiality from being given to the dictionary providing device outside the device. it can.

  According to the fourth aspect of the invention, the character string selection means allows the user to select a character string to be given to the dictionary providing device, so that a character string that should be allowed to be given to the dictionary providing device among character strings including unknown words. The column can be easily selected by the user of the apparatus, and thereby, for example, a sentence requiring confidentiality can be prevented from being given to the dictionary providing apparatus outside the apparatus.

  According to the fifth aspect, the unknown word is extracted as a technical term by referring to the already extracted unknown word, so that the unknown word is handled in the same way as the technical term without waiting for the update of the technical term dictionary. be able to. That is, for example, the unknown words can be aggregated by the aggregation means and searched by the search means.

  According to the sixth aspect, when an unknown word is registered in a new dictionary, the degree of relevance of the unknown word and the associated person are indicated as technical terms corresponding to the unknown word registered in the new dictionary. Therefore, it is possible to prevent duplication of registration and utilize the already-calculated unknown word association degrees and associated person information without waste.

  According to the seventh aspect, when an unknown word is registered in new specialized field data, the degree of association of the unknown word and the associated person are changed to the specialized field registered in the new specialized field data. Since the related frequency of the technical terms corresponding to the corresponding unknown word and the associated person are taken over, duplicate registration is prevented, and the related frequency of the unknown words already collected and the information of the associated person are efficiently used Can be used.

  According to the eighth aspect, data including an unknown word to be given to the dictionary providing device is encrypted, so that a sentence requiring confidentiality can be decoded by a (third party) device other than the dictionary providing device. It is not given in a form, and information leakage can be prevented.

  According to the ninth aspect of the invention, by using the public key cryptosystem, information can be transmitted in an extremely safe manner with a simple method. It is sufficient to set the same public key in advance without setting a different public key for each expert extraction device, and the key in the dictionary providing device. Management of the system can be made easy.

  According to the tenth aspect of the present invention, a user can create a vocabulary dictionary in which unknown words used by many users including these unknown words are widely registered without registering unknown words in the vocabulary dictionary. Can be fed into the expert extraction device.

  According to the eleventh aspect of the invention, only when it is determined that the received unknown word is a technical term, it is registered in a new dictionary, so wrong terms due to typographical errors or misunderstandings, or only within a very narrow range. It is possible to prevent an unsuitable term such as an invalid term from being registered in a new dictionary.

  According to the twelfth aspect of the invention, the discount evaluation for the fee for providing a new dictionary is performed according to the amount corresponding to the received unknown word, so that the customer who sent many unknown words (and related sentences) Big billing discounts will be received. Therefore, transmission of these unknown words from the expert extracting device can be promoted, and thereby, a specialized word dictionary in which many specialized words are registered can be provided from the dictionary providing device.

  According to the thirteenth aspect of the invention, discount evaluation is performed according to the amount corresponding to a plurality of character strings including unknown words (for example, before and after that), so that the transmission of these character strings from the expert extraction device is facilitated. Therefore, it is possible to provide a vocabulary dictionary in which many accurate technical terms are registered from the dictionary providing device.

  According to the fourteenth aspect, since the discount evaluation is performed according to the amount corresponding to the unknown word determined to be the technical word among the received unknown words, an erroneous term due to a typographical error or misunderstanding, It is possible to avoid unnecessary discount evaluation for customers who have sent a large amount of unknown words that are not suitable as technical terms (such as terms that can only be used in a very narrow range) (useful for updating the dictionary). In addition, since a large number of registered unknown words that have already been determined to be technical words in the new dictionary are sent from the expert extraction device having the old (version) dictionary, such a dictionary update is performed. It is possible to avoid unnecessary discount evaluation for customers who neglect (purchase).

  According to the fifteenth aspect of the present invention, the unknown word extracted by the expert extraction device is given to the dictionary providing device without the user registering the unknown word in the technical word dictionary. It is possible to provide a dictionary update system in which an expert word dictionary in which unknown words used by a person are registered is created and the expert word dictionary is updated by providing the expert word dictionary to the expert extraction device.

  According to the sixteenth aspect of the present invention, it is possible to realize a method for controlling the expert extraction apparatus that exhibits the same effect as that of the first aspect of the invention.

  According to the seventeenth aspect, it is possible to realize a control program for an expert extraction apparatus that exhibits the same effect as that of the first aspect.

  According to the eighteenth aspect of the invention, a computer-readable recording medium that exhibits the same effect as that of the first aspect of the invention can be realized.

  Hereinafter, an embodiment of the present invention will be described with reference to the accompanying drawings.

<1. Configuration and operation of expert extraction device>
FIG. 1 is a block diagram showing a configuration of an expert extraction device and a dictionary server (dictionary providing device) according to an embodiment of the present invention. The expert extraction device 5 and the dictionary server 6 constitute a dictionary update system. In the dictionary update system, first, the configuration of the expert extraction device 5 will be described.

  The expert extraction device 5 is realized by a computer having a general configuration such as a user interface (user I / F) device 1, an arithmetic processing unit 2, a storage unit 3, and a communication unit 4. The arithmetic processing unit 2 in the expert extraction device 5 which is a computer, according to control by a predetermined program, includes a technical word extraction unit 21, a user identification unit 22, a totaling unit 23, a search unit 24, a character recognition unit 25, a network monitor. 26, the document management server 27, the unknown word cutout location setting unit 28, the unknown word optimization processing unit 29, and the encryption processing unit 41.

  Further, the storage unit 3 includes a user-specific word database (user-specific word DB) 31, a specialized field database (specialized field DB) 32, and a document database (document DB) 33, and a general word dictionary. 34, technical term dictionary 35, unknown word data 36, and the like are stored. The storage unit 3 is typically a hard disk drive device. The storage unit 3 is not limited to this, but is preferably a storage device that can operate at a relatively large capacity and high speed.

  The user I / F device 1 is, for example, a keyboard, a mouse, a display, a printer, and the like, and is used by a user of the device to operate the expert extraction device 5 that is a computer and check the processing result. .

  The arithmetic processing unit 2 is a microprocessor that operates as a CPU of a computer, for example. The microprocessor is connected to a memory used for various arithmetic processes. A program for realizing the above functions in the arithmetic processing unit 2 is read into the memory from a hard disk (or other storage device) constituting the storage unit 3 to control the microprocessor. When the program is introduced into the hard disk, a computer-readable recording medium such as a CD-ROM or DVD-ROM (not shown) is used. For example, the program stored in the CD-ROM in a compressed state is introduced into the hard disk in an executable state by operating a setup program stored in the executable state in the CD-ROM. (Installed).

  The communication unit 4 includes a dictionary server 6 and a plurality of other computers (typically a device that creates and prints a document file that is an electronic document, hereinafter referred to as a “document creation device”), a network printer (not shown), and the like A NIC (Network Interface Card) used for communication is connected to a LAN (Local Area Network) 9 so as to be able to communicate with each other. Here, the document file created in the document creation apparatus which is another computer is accumulated in the storage unit 3 via the communication unit 4 and managed by the document management server 27.

  It should be noted that another known communication network (for example, the Internet or a dedicated communication line) may be used in place of the LAN 9, and data such as a dictionary can be read by a computer such as a CD-R without using the communication network. Data may be exchanged between the expert extraction device 5, the dictionary server 6, or another computer, such as a document creation device, via a medium or the like. Thus, the communication unit 4 functions as a data providing unit and a receiving unit.

  The encryption processing unit 41 transmits unknown word data, which will be described later, to the dictionary server 6 via the communication unit 4 and the LAN 9, or the latest technical language dictionary and specialized field database (hereinafter referred to as “technical language dictionary, etc.”) from the dictionary server 6. ”), A predetermined encryption process or decryption process is performed. Details will be described later. As described above, data exchange with the dictionary server 6 and the like may be performed via various recording media and the like. The encryption processing unit 41 also functions as data providing means and receiving means.

  The network monitor 26 acquires an electronic document such as a document file that flows through the LAN 9 that is a network to which the communication unit 4 is connected and an image file of the document. When the packet acquired by the communication unit 4 is a packet related to a document file, the network monitor 26 extracts a character code from the data load unit of each packet related to the document file, and the corresponding document file (that is, information on the document data itself or The file including the accompanying information of the document data including information representing the creator / editor of the file is reconstructed. The reconstructed document file is output to the technical term extraction unit 21, the user identification unit 22, the document management server 27, and the like. Packets relating to the document file are, for example, packets that flow to the LAN 9 when trying to store the document file created by the document creation device in the storage unit 3 of the expert extraction device, or spool the created mail to the mail server. The packet that flows when

  When the packet acquired by the communication unit 4 is a packet related to the image file of the document, the network monitor 26 extracts the image data of the document from the data load unit of each packet related to the image file, and the corresponding image file (that is, The image data itself and a file including accompanying information such as information representing the creator / editor of the image data are reconstructed. The reconstructed image file is output to the character recognition unit 25 and the user identification unit 22. Packets relating to image files are not illustrated, for example, packets that flow when an image file created by the document creation device is stored in the storage unit 3 of the expert extraction device 5 or image files created by the document creation device. Packets that flow when trying to print by a network printer, packets that flow when trying to store an image file read by a network scanner (not shown) in the storage unit 3 of the expert extraction device 5, and the like.

  The character recognition unit 25 includes a predetermined character extraction module and a character recognition dictionary, and has a known configuration in which character recognition is performed by a predetermined character recognition engine. When an image file of a document is supplied to the character recognition unit 25, the character extraction module extracts a portion assumed to be a character from the image file as a character image. The character recognition engine converts the image data into a character code by performing character recognition on the character image cut out by the character cut-out module with reference to the character recognition dictionary. Thus, a document file is constructed from the image file. The constructed document file is supplied to the technical term extraction unit 21 and the document management server 27. The image file supplied to the character recognition unit 25 is acquired from the network monitor 26, but may be acquired from the scanner 51 or the like connected (locally) to the expert extraction device 5.

  The technical term extraction unit 21 refers to the general word dictionary 34 and the technical term dictionary 35 registered in the storage unit 3 for technical terms included in the document file supplied from the network monitor 26 or the character recognition unit 25. Extract. The document file may be created by the expert extraction device 5 and may not necessarily be obtained via the network monitor 26, the character recognition unit 25, or the like. For example, when the expert extraction device 5 is not connected to a network or the like, the document file is created in the expert extraction device 5 or is given from another document creation device via a predetermined storage medium.

  The general word dictionary 34 and the technical word dictionary 35 referred to by the technical word extraction unit 21 are registered with a predetermined headword, its part of speech, etc., with a unique key (Key) for identifying each word. ing. In this general word dictionary 34, for example, words that are used on a daily basis such as “further”, “disc”, “home appliance”, “kanji”, “expression”, and the like are registered. Since these words are words that are not difficult to understand if there is knowledge necessary for social life, the general word dictionary 34 is a collection of words that are not technical terms, but may include technical terms. . These words also include conjunctions and the like necessary for composing sentences. Next, the technical language dictionary 35 will be described.

  FIG. 2 is a diagram for explaining an example in which a certain unknown word is registered in the technical word dictionary. More specifically, FIG. 2A shows the state of the technical word dictionary before the unknown word is registered. FIG. 2B is a diagram illustrating a part, and FIG. 2B is a diagram illustrating a jargon dictionary after unknown words are registered with respect to the jargon shown in FIG.

  As shown in FIG. 2A, the technical term dictionary 35 includes, for example, “EUC”, “SED / sed”, “disk array”, “Unicode”, “information home appliance”, “regular expression”, and the like. Technical terms are registered. These technical terms are words that are used on the premise of specialized knowledge that is rarely used on a daily basis, such as technical terms such as computers and semiconductors, and technical terms such as painting and music. This technical term is set in advance in consideration of specialized knowledge desired by the user of the apparatus. Therefore, the technical term dictionary 35 may include general words that are useful for extracting specialists, and there is no particular limitation on the contents thereof.

  The technical term extraction unit 21 extracts a technical term registered in the technical term dictionary 35 from the document file by performing a well-known morphological analysis with reference to the general word dictionary 34 and the technical term dictionary 35 as described above. Alternatively, unknown words that are not registered in either the general word dictionary 34 or the technical word dictionary 35 are extracted.

  The technical term extraction unit 21 registers the extracted unknown words in the technical term dictionary 35 and a user-specific word database 31 described later. That is, the unknown word is automatically regarded as a technical term. This reduces the work burden on the user who registers a newly born word as a technical term.

  FIG. 2B shows a technical term dictionary in which an unknown word “third level” is newly registered for the technical term shown in FIG. As shown in this figure, unknown words are registered in the same manner as technical words registered in advance. The unknown word may be registered in a data format other than the professional dictionary (for example, a format such as an unknown word dictionary) as long as it is a format that is treated as a technical word, that is, a format that is extracted as a technical word and is a target of aggregation. .

  Further, the technical word extraction unit 21 separates the unknown word from the technical word dictionary 35 as unknown word data 36 together with a sentence including the unknown word (here, a character string of a whole paragraph including a sentence including the unknown word). It is stored in the storage unit 3 and transmitted to the dictionary server 6 at a predetermined timing by the unknown word cut-out location setting unit 28. In addition, if the sentence containing the unknown word memorize | stored as mentioned above can be used in order to analyze the meaning of an unknown word, there will be no limitation in the range.

  The unknown word cut-out location setting unit 28 sets a part of the sentence (character string) included in the unknown word data 36 that is permitted to be transmitted to the dictionary server 6 by the user's selection and the unknown word as a new unknown. The word data 36 is cut out and the unknown word data 36 cut out is transmitted to the dictionary server 6. As described above, the unknown word cutout location setting unit 28 functions as a data providing unit together with the communication unit 4. Detailed operation of the unknown word cutout location setting unit 28 will be described later.

  The unknown word optimization processing unit 29 receives the latest technical word dictionary and the like from the dictionary server 6, and based on the received technical word dictionary and the like, an unknown word (new technical word) and a total value associated with the unknown word Is optimized in the word database 31 for each user. As described above, the unknown word optimization processing unit 29 functions as a data receiving unit together with the communication unit 4. Detailed operation of the unknown word optimization processing unit 29 will also be described later.

  The user specifying unit 22 is a person related to the file, such as a creator, an editor, or an owner of the file from the document file from which the technical word is extracted by the technical word extraction unit 21 or the image file of the corresponding document. That is, a person who is recognized to be related to the technical term extracted by the technical term extraction unit 21 (hereinafter, this person is referred to as “user of technical term” or simply “user”) is specified. Although the user of the technical term is specified based on the accompanying information of the file, it may be performed by detecting the user of the document creation apparatus that created the file, It may be performed by analyzing. Thus, the user (of technical terms) specified by the user specifying unit 22 and the technical terms extracted by the technical term extraction unit 21 are registered in the storage unit 3 in a state of being associated with each other. .

  The totaling unit 23 identifies each technical term extracted by the technical term extraction unit 21 by the user specifying unit 22 while referring to the correspondence relationship between the technical term registered in the storage unit 3 and its user. The user-specific word database 31 is constructed in the storage unit 3 by counting the number of times (number of extractions) associated with each user.

  In addition, the totaling unit 23 may total the ratio (or frequency) related to each user specified by the user specifying unit 22, or attach an index or rank indicating the degree of relevance. Also good. Such a totaling unit 23 is a numerical value indicating the relevance between a technical term and its user, such as the number of extractions, the extraction ratio, the index, and the ranking, that is, how many experts the technical term user has for the technical term. What is necessary is just to determine the numerical value (henceforth "relevance frequency") which shows or guesses.

  FIG. 3 is a diagram illustrating a specific example of the user-specific word database 31 constructed in the storage unit 3 by the counting unit 23. As shown in FIG. 3, the headings of the user-specific word database 31 are given by technical terms or equivalent words or symbols. For each technical term, for each user such as “Tanaka”, “Takahashi”, and “Nakamura”, the number of times each technical term and each user are associated by the document file or image file (number of extractions) It is tabulated. For example, the term “mirror disk” is used 180 times by “Tanaka”, 18 times by “Takahashi”, and 8 times by “Nakamura”. Further, it is shown that the technical term “EUC” is used 8 times by “Tanaka”, 149 times by “Takahashi”, and 100 times by “Nakamura”. Furthermore, the term “regular expression” is used 6 times by “Tanaka”, 18 times by “Takahashi”, and 116 times by “Nakamura”.

  When the user-specific word database 31 is constructed, the above-described correspondence between the technical term stored in the storage unit 3 and the user is, for example, “Tanaka” as the technical language user. In the document file, the technical term “mirror disk” is extracted four times and the technical term “RAID” is extracted five times.

  Based on this correspondence, the counting unit 23 increases the number of users “Tanaka” corresponding to the technical term “mirror disk” by four, and the user “Tanaka” corresponding to the technical term “RAID”. The user-specific word database 31 is constructed by performing a totaling operation such as increasing the number of times of. Similarly, even if a related frequency other than the number of extractions is used, it can be easily tabulated by performing a predetermined calculation. As described above, the user-specific word database 31 includes unknown words that are newly handled as technical terms. The handling of unknown words when the dictionary is updated (upgraded) will be described later.

  When the technical term extraction unit 21 and the user specifying unit 22 determine the correspondence between a technical term for a certain document file and its user, the document management server 27 stores the data of the document file and the image file or the The document database 33 is constructed by creating a source pointer indicating the location of the document file or image file included in the accompanying information. When there is no need to correspond to the original document, it is not necessary to create the origin pointer, and the document management server 27 may be omitted.

  FIG. 4 is a diagram for explaining the configuration of the specialized field database 32. As shown in FIG. 4, in this specialized field database 32, a predetermined correspondence between a specialized field indicating the type of specialized knowledge and the like and a headline of the technical term is registered. For example, the technical field “HDD distributed multiplexing” is classified in association with technical terms such as “RAID”, “disk array”, and “mirror disk”. In addition, in the specialized field of “character code standardization”, technical terms such as “EUC”, “Shift JIS”, “Unicode”, “External character registration”, and “Third level” are classified.

  Here, the totaling unit 23 constructs the user-specific word database 31 by performing totaling not only for the technical terms but also for each specialized field registered in the specialized field database 32. In the specialized field database 32, as described above, the technical terms such as “RAID” and “disk array” are classified into the specialized field “HDD distributed multiplexing”, and each specialized field such as “RAID” and “disk array”. Since the number of times of extraction is known for each user corresponding to a word, by extracting the sum of the number of times of extraction for each technical word classified into each specialized field, extraction for each user corresponding to each specialized field is performed. The number of times can be determined.

  The user-specific word database 31 and the document database 33 are preferably updated relatively frequently. As the number of updates increases, the relationship between the degree of detail of technical terms and the number of extractions (relevance frequency) becomes clearer as specialists with deeper expertise and more experience, and new people are added. This is because it is possible to cope with an increasing number of unknown words. Compared to this, the frequency of updating the specialized field database 32 may be small. This is because the correspondence between specialized fields and technical terms is stable to some extent. For example, this update may be performed when an unknown word appears or the usage changes.

<2. Search operation of expert extraction device>
Next, the operation of the expert extraction device 5 when a specialized word or a specialized field is designated by the user of the device and a search is performed will be described. When the designated technical word is an unknown word, this unknown word is transmitted from the expert extraction device 5 to the dictionary server 6 at a predetermined timing. This operation will be described later. First, the user of the device uses a keyboard, a mouse, or the like included in the user I / F device 1 to input a jargon (or a specialized field) corresponding to information regarding desired specialized knowledge and experience. In this case, the user I / F device 1 functions as an input unit.

  The technical terms input by the user of the apparatus are supplied to the search unit 24. When the technical language is supplied, the search unit 24 searches the user-specific word database 31. That is, in the word database 31 for each user, when a technical term that matches or is similar to the supplied technical term is searched and a technical term that matches or is similar to the technical term is searched, the use associated with the technical term is used. Extract the person. For example, when the term “PERL” is input by the user of the apparatus, the search unit 24 reads “..., user“ Tanaka ”0 times,..., User“ Takahashi ”15 times, user“ Nakamura ”163. Search results like " When such search results are obtained, the search unit 24 sorts the search results in descending order, for example, and displays the sorted search results on the display of the user I / F device 1. In the above example, the search results displayed on the display of the user I / F device 1 are “user“ Nakamura ”163 times, user“ Takahashi ”15 times,..., User“ Tanaka ”0 times,. Is like that. By referring to this result, the user of the device can easily determine that the user “Nakamura” is highly likely to have deep expertise and experience in the input technical term. it can. Therefore, the user of the apparatus can obtain desired knowledge and information with a high probability by directly asking this user “Nakamura” later.

  However, it is possible that the user of the device may not be able to ask this user directly due to the circumstances of the user specified as above, and not only the user but also a document in which the user uses the technical language. You may want to know the file itself. In this case, when the user of the apparatus inputs a technical term, the user I / F apparatus 1 can be configured to specify that the user wants to know the location of the document file itself. For example, an object screen that can be selected by the user of the apparatus as to whether or not the user wants to know the location of the document file itself is displayed on a search screen or the like. When a technical term is input in this way and the user of the device specifies that the user wants to know the location of the document file itself, the search unit is based on the user identified by the technical term and the technical term. 24 searches the document database 33. The document database 33 stores original pointers indicating the locations of users and document files themselves in association with technical terms. Therefore, the search unit 24 can extract the corresponding source pointer based on the identified user and the technical term. Then, for example, it is displayed on the display of the user I / F device 1 that the original of the document file in which the user “Tanaka” uses the technical term “disk array” is “weekly report 1999/02/18”. By referring to the original pointer, the user of the apparatus can easily grasp the location of the document file itself, and the necessity of directly asking knowledge and information from the specified user is reduced.

  In addition, when the user of the apparatus does not know so much technical terms, it is possible to perform a search using a specialized field. The concept of a specialized field is wider than each specialized term, and the number of specialized fields is considerably smaller than the number of specialized terms. Therefore, it is difficult to let the device user input specialized fields such as “HDD distributed multiplexing”. It is also easy to display a plurality of specialized fields as options and allow the user of the apparatus to select one of them. In this case, it is possible to designate a specialized field close to desired knowledge and information from the meaning of the word without knowing a specific term that defines the specialized field. For example, in this case, the user of the device uses a keyboard, a mouse, or the like included in the user I / F device 1 to select a specialized field corresponding to information on the desired expertise or experience from the displayed options. .

  On the LAN 9, the same document may be communicated many times. If you repeat the process of counting technical terms for these same documents, there will be a discrepancy between the number of times the user actually used the technical terms and the number of times the totals were processed. There is a high possibility that the user of the apparatus will make a mistake in the determination of the person who has. For example, a mail is sent from a document creation apparatus and spooled to a mail server (not shown), and the same spooled mail is received by another document creation apparatus. Therefore, for example, the network monitor 26 acquires information (for example, an IP address) of a document file or an image file, and when the transmission source is a server such as a mail server, Not acquired as a document file. If the document file is acquired only when the document file is transmitted from the client, it is possible to prevent the same document from being repeatedly counted and to further increase the reliability of the search result.

  As described above, the network monitor 26 obtains a document file or the like flowing through the LAN 9. If this document file contains an unknown word that is not in the technical term dictionary 35, the unknown word is extracted by the technical term extraction unit 21. And stored as unknown word data 36 and transmitted to the dictionary server 6 by the unknown word cutout location setting unit 28 at a predetermined timing. The unknown word transmitted from the expert extraction device 5 is received by the dictionary server 6, and the dictionary server 6 performs a predetermined process on the unknown word and then creates a new version of the specialized word dictionary and the like. It transmits to the extraction device 5. The configuration of the dictionary server 6 will be described again with reference to FIG.

<3. Configuration and operation of dictionary server>
The dictionary server 6 that is the dictionary providing apparatus shown in FIG. 1 is realized by a computer having a general configuration such as an arithmetic processing unit, a storage unit 7, and a communication unit 8. The arithmetic processing unit of the dictionary server 6 as a computer functions as a customer data evaluation unit 61, a technical term analysis unit 62, a data management unit 63, a data distribution unit 64, and an encryption processing unit 81 according to control of a predetermined program. . The dictionary server 6 is not limited to a server in the client server system as long as it is a device having a function of providing a dictionary.

  This arithmetic processing unit is, for example, a microprocessor that operates as a CPU of a computer, like the arithmetic processing unit 2 in the expert extraction device 5, and a memory used for various arithmetic processing is connected to the microprocessor. . A program for realizing the above functions in the arithmetic processing unit is read into a memory from a hard disk (or other storage device) constituting the storage unit 7 and controls the microprocessor. In order to introduce the program into the hard disk, a computer-readable recording medium such as a CD-ROM or DVD-ROM (not shown) is used as in the case of the expert extraction device 5.

  The storage unit 7 includes a customer-specific charging database (customer-based charging DB) 72, a specialized field database (specialized field DB) 74, and a version database (version DB) 75, and a reference charging rate. Data 71 and a technical language dictionary 73 are stored. The storage unit 7 is typically a hard disk drive device, but is not limited thereto. However, the storage unit 7 is preferably a storage device that has a relatively large capacity and can operate at high speed.

  In the dictionary server 6 of the present embodiment, user I / F devices such as a keyboard, a mouse, a display, and a printer are omitted, but may be provided for users of the dictionary server 6.

  The communication unit 8 is a NIC that is used to communicate with the expert extraction device 5 and other document creation devices such as a plurality of computers, and is connected to the LAN 9 so as to be capable of mutual communication. FIG. 1 shows a configuration in which only one expert extraction device 5 is connected to the dictionary server 6, but actually a plurality of expert extraction devices 5 are connected to the dictionary server 6. And

  The cryptographic processing unit 81 receives a predetermined unknown word or the like from the expert extracting device 5 or transmits the latest technical word dictionary or the like to the expert extracting device 5 via the communication unit 8 and the LAN 9. Perform encryption or decryption. The encryption process and the decryption process are performed by a so-called public key cryptosystem. This public key cryptosystem uses a pair of keys, a public key that is widely disclosed to others and a private key that is stored only by the principal, thereby enabling communication between a specific sender and receiver. Provide an environment in which communication can be performed safely without being known to a person. For example, data encrypted using a public key corresponding to the expert extraction device 5 registered in a general public key server or dictionary server 6 (here, the latest technical language dictionary, etc.) Since the decryption can be performed only by using the secret key that is strictly stored in the extracting device 5, the latest technical language dictionary or the like is not illegally used by a third party who has not paid the price.

  Similarly, the public key cryptosystem is used in the cryptographic processor 41 in the expert extraction device 5. For example, data encrypted using a public key corresponding to a dictionary server 6 registered in a general public key server or expert extraction device 5 (here, all of unknown words and sentences including the unknown words or Can be decrypted only by using a secret key that is strictly stored in the dictionary server 6, so that an unknown word or a matter that requires confidentiality contained in a sentence containing the unknown word can be decrypted by a third party. Never known. Further, in this case, the same public key (corresponding to the dictionary server 6) can be used in a plurality of other expert extraction devices other than the expert extraction device 5, so that it depends on each expert extraction device. It is sufficient to set the same public key in advance without setting different public keys individually, and the key management in the dictionary server 6 is easy.

  Note that data encrypted using a secret key that is strictly stored in the expert extraction device 5 or the dictionary server 6 can be decrypted only with the corresponding public key. Thus, authentication (so-called digital signature) can be performed to the effect that the sender of the data is indeed the person. As a result, it is impossible for another person to impersonate the person himself, so that the safety of communication between the expert extraction device 5 and the dictionary server 6 can be further enhanced.

  The technical word analysis unit 62 determines whether or not the unknown word is a technical word based on the unknown word decrypted by the encryption processing unit 81 and all or part of the sentence including the unknown word. If it is a word, classify it into the corresponding specialized field. Detailed operation of the technical term analysis unit 62 will be described later.

  The data management unit 63 registers (merges) the unknown word determined and classified by the technical language analysis unit 62 as a new technical term in the technical language dictionary 73. For example, a new key, headline, part of speech, etc. are registered in the same manner as when an unknown word “third level” is newly registered in the technical term dictionary 35 shown in FIGS. 2 (a) and 2 (b). The Further, the data management unit 63 registers the technical term in association with the corresponding specialized field in the specialized field database 74 according to the classification result of the specialized word performed by the specialized word analysis unit 62. As described above, the technical term analysis unit 62 and the data management unit 63 function as a dictionary creation unit.

  Furthermore, the data management unit 63 manages versions of the merged technical term dictionary 73 and specialized field database 74. For example, when the data management unit 63 collectively registers a plurality of unknown words, the technical word dictionary 73 (and the specialized field database 74) before the batch registration and the specialized word dictionary 73 (and the specialized field database 74) after the batch registration. Is stored in the version database 75 together with a unique version number. A unique version number may be assigned to each registration, and all data of the technical language dictionary 73 (and the specialized field database 74) may be stored.

  The customer data evaluation unit 61 performs billing discount evaluation of the user of the expert extraction apparatus 5 that has transmitted the unknown word determined as a new technical word by the technical language analysis unit 62, that is, the customer, and the discount rate based on this evaluation Is registered in the customer-specific charging database 72. Detailed operation of the customer data evaluation unit 61 will be described later.

  The data distribution unit 64 gives the latest technical language dictionary 73 and the specialized field database 74 to the cryptographic processing unit 81 at predetermined intervals or according to the customer's request, and the cryptographic processing unit 81 encrypts them and extracts the expert extraction device. To 5. The latest technical language dictionary 73 and specialized field database 74 are charged, and the charge amount is calculated according to the amount of updated data, for example, and is given to the customer data evaluation unit 61. The customer data evaluation unit 61 refers to the customer-specific charging database 72, and when the discount rate is registered, the final charge amount obtained by discounting the charge amount received from the data distribution unit 64 with the discount rate Is registered in the customer-specific charging database 72.

  As described above, the unknown word data 36 extracted by the technical word extraction unit 21 in the expert extraction device 5 is transmitted to the dictionary server 6 by the unknown word cutout location setting unit 28 at a predetermined timing. By registering this unknown word in a technical language dictionary or the like in a predetermined case, a new version of the technical language dictionary or the like is created and transmitted to the expert extraction device 5. A series of operations related to unknown words in the expert extraction device 5 and the dictionary server 6 will be described with reference to the drawings.

<5. Unknown word related operations in expert extraction device and dictionary server>
FIG. 5 is a flowchart showing a flow of processing related to an unknown word in the expert extraction device and the dictionary server. In step S101 shown in FIG. 5, the technical term extraction unit 21 in the expert extraction device 5 extracts an unknown word that is not registered in the technical term dictionary 35 and registers it in the technical term dictionary 35. A sentence including an unknown word (for example, a character string including a plurality of words before and after the unknown word) is stored in the storage unit 3 as the unknown word data 36 separately from the technical word dictionary 35.

  Subsequently, in step S102, the unknown word cutout part setting unit 28 determines a part of the sentence (character string) included in the unknown word data 36 that is permitted to be transmitted to the dictionary server 6 at a predetermined timing as a user of the apparatus. The selected portion and the unknown word are cut out as new unknown word data 36 and transmitted to the dictionary server 6. After transmission, the unknown word data 36 is deleted, and new unknown word data 36 based on the unknown word extracted by the technical word extraction unit 21 is formed. Here, it is preferable from the viewpoint of the work efficiency of the user of the apparatus that the predetermined timing is a time when a certain number of unknown words are stored, such as a time when a predetermined number of registered unknown words is reached. However, it may be the time when a predetermined period has passed, or the time when a predetermined condition is satisfied or the time when the user of the apparatus arbitrarily decides.

  FIG. 6 is a diagram illustrating a storage example of unknown word data including a sentence before being cut out by selection of a user of the apparatus. In the left column of the figure, unknown words such as information home appliances, third level, and ubiquitous are stored, and in the right column of the figure, sentences including the corresponding unknown words are stored. As shown in this figure, the meaning of each unknown word can be easily estimated from a sentence including the unknown word. Of these unknown word data, a processing example in step S101 of the unknown word cutout location setting unit 28 for the unknown word data 36 including the unknown word “information home appliance” will be described.

  FIG. 7 is a diagram illustrating an example of a screen display for allowing the user of the apparatus to select a portion that permits transmission to the dictionary server by the unknown word cutout location setting unit. In the upper column shown in the figure, four paragraphs “paragraph”, “sentence”, “sentence”, and “word” are described as separation methods, and only one of these separation methods is described in the apparatus. So-called radio buttons are displayed so that the user can select. In this radio button display method, a plurality of small circular flat buttons are displayed in the screen (window) in the figure, and any one of these buttons is ( (Alternatively) When selected, a filled circle is placed in the circle of the selected button. In FIG. 7, “sentence” is selected as the separation method. Such a radio button functions as a delimiter setting unit that allows a user of the apparatus to easily extract a character string that permits transmission to the dictionary server 6 from a character string that includes an unknown word. In addition to the radio button, this delimitation setting means may be a well-known means that allows an alternative selection of the user of the apparatus for delimiting the character string.

  In the lower column shown in FIG. 7, an unknown word “information home appliance” is displayed on the left side, and on the right side, sentences including the unknown word are displayed divided into phrases. A so-called check box is displayed above each clause so that the user of the apparatus can select the clause. In this check box display method, a small square check box is displayed in the screen (window) in the figure, and this check box is selected when this check box is selected by the user of the device using a mouse or the like. The check box is filled. In FIG. 7, since “sentence” is selected as the delimiting method, a plurality of check boxes corresponding to all the clauses are automatically selected (by the unknown word cutout location setting unit 28). . Further, when one of the selected check boxes is selected, the check box is in a non-selected state, and when a non-selected check box is selected, the check box is in a selected state. Such a check box functions as a character string selection unit that allows a user of the apparatus to select a character string that permits transmission from a character string including an unknown word to the dictionary server 6. In addition to the check box, the character string selection means may be any known means that enables the user of the apparatus to select one or more character strings.

  FIG. 8 is a diagram showing another example of a screen display for allowing the user of the apparatus to select a portion that permits transmission to the dictionary server by the unknown word cutout location setting unit. In FIG. 8, since “sentence” is selected as the delimiting method, only the check box corresponding to the phrase including the unknown word is automatically selected (by the unknown word cutout location setting unit 28). Yes. As shown in FIG. 8, when a phrase is selected, the unknown word cutout location setting unit 28 identifies the unknown word “information home appliance” and the portion from which only the phrase “information home appliance” is cut out as unknown word data 36. However, it is difficult to guess the meaning of the unknown word from the phrase including the unknown word even though it can be easily determined that the unknown word is a noun. However, by transmitting only this phrase, it is possible to prevent a sentence requiring confidentiality from being transmitted to the dictionary server 6.

  FIG. 9 is a diagram showing still another example of a screen display for allowing the user of the apparatus to select a portion that permits transmission to the dictionary server by the unknown word cutout location setting unit. In FIG. 9, “Sentence” is selected as the separation method in the same manner as in FIG. 8, but if “Sentence” is selected, confidentiality that describes the name of the product to be marketed and the time of its introduction, etc., is displayed. A required sentence is transmitted to the dictionary server 6. Therefore, it is preferable that the delimiter method of “sentence” or “word” can be selected as in this embodiment.

  However, as described above, it is difficult to guess the meaning of a single phrase including an unknown word even though it can be easily determined that the unknown word is a noun. Therefore, as shown in FIG. 8 or FIG. 9, even when “clause” is selected as the separation method, for example, several phrases (character strings) before and after the unknown word are transmitted as the unknown word data 36. It is preferred that This example will be described with reference to FIG.

  FIG. 10 is a diagram showing a screen display example in which a plurality of portions permitting transmission to the dictionary server are selected by the user of the apparatus. In FIG. 10, as in FIG. 8 or FIG. 9, “sentence” is selected as the delimiting method, but in addition to the check box corresponding to the phrase including the unknown word “third level” being selected, Check boxes corresponding to the preceding and following phrases “Registered external characters” and the phrase “Kanji” are selected. Therefore, it becomes easy to guess the meaning of the unknown word “third level”, and the check box corresponding to the phrase of the sentence portion requiring confidentiality is not selected, so that the matters requiring confidentiality are stored in the dictionary server. 6 can be easily prevented according to the selection of the user of the apparatus.

  As described above, the processing of steps S101 and S102 shown in FIG. 5 is performed by the expert extraction device 5, but the processing of the next steps S103 to S109 is performed by the dictionary server 6. Hereinafter, it demonstrates in order.

  In step S <b> 103, the technical word analysis unit 62 of the dictionary server 6 registers the unknown word included in the unknown word data 36 transmitted from the unknown word extraction part setting unit 28 of the expert extraction device 5 in the technical language dictionary 73. It is determined whether or not it has been done. Such a determination is made by using the latest terminology dictionary 73 as described later, but the terminology dictionary 35 stored in the storage unit 3 of the expert extraction device 5 is not necessarily the latest. Because. As a result of this determination, if registered (Yes in step S103), the process proceeds to step S106. If the result of this determination is that it is not registered (No in step S103), the process proceeds to step S104.

  In the subsequent step S104, the technical word analysis unit 62 should treat the unknown word included in the unknown word data 36 transmitted from the unknown word cutout location setting unit 28 of the expert extraction device 5 as a technical term. Determine whether. This determination is made because the unknown word may not be suitable as a technical term, such as an erroneous term due to a typo or misunderstanding, or a term that can be used only in a very narrow range. The technical language analysis unit 62 performs the above determination using a technical language discrimination function including a knowledge database including misuse examples and non-technical word examples, a predetermined inference engine, and the like. As a result of this determination, if the unknown word should not be handled as a technical term (No in step S104), the process proceeds to step S106. As a result of this determination, if it should be handled as a technical term (Yes in step S104), the process proceeds to step S105.

  In the subsequent step S105, the technical word analysis unit 62 classifies the unknown words included in the unknown word data 36 into a predetermined specialized field, and the data management unit 63 registers the classification result in the specialized field database 74. The unknown word is registered (merged) in the technical language dictionary 73.

  More specifically, the technical word analysis unit 62 refers to a sentence including an unknown word included in the unknown word data 36 (for example, a character string such as a plurality of clauses), and thereby converts the unknown word into the specialized field database 74. Categorized in a specific field of expertise registered (or newly registered). For example, as shown in FIG. 4, when classifying an unknown word “third level”, the technical word analysis unit 62 includes the phrases “Registered external characters” and “Kanji characters” included in the unknown word data 36 ( It is determined that the technical term “regular character registration” is registered in the specialized field “character code standardization” registered in the specialized field database 74. This discrimination is performed by the technical term discriminating function of the technical term analyzing unit 62. Thus, the technical language analysis unit 62 classifies the unknown word (new technical term) “third level” into the specialized field “character code standardization”. This classification is performed by the technical term classification function of the technical term analysis unit 62. The above classification results are as shown in FIG. Next, the data management unit 63 registers the classification result in the specialized field database 74 and, for example, a new “third level” unknown in the technical term dictionary 35 shown in FIGS. 2 (a) and 2 (b). Similar to the case where a word is registered, the unknown word is registered (merged) in the technical term dictionary 73. This registration is performed by the merge function of the data management unit 63.

  Next, in step S106, the technical word analysis unit 62 determines whether or not the analysis of all unknown word data transmitted by the unknown word cutout location setting unit 28 in step S102 has been completed. As a result of the determination, if the processing is not completed (No in step S106), the processing returns to step S103, and the above processing is repeated until all the analysis is completed (S106 → S103 →... → S106). If all the analyzes are completed (Yes in step S106), the process proceeds to step S107.

  In subsequent step S107, the data management unit 63 manages the merged technical term dictionary 73 and specialized field database 74 versions. For example, every time the above series of processing (S101 to S106) is completed, the data management unit 63 performs the pre-processing terminology dictionary 73 (and the specialized field database 74) and the post-processing terminology dictionary 73 (and the specialized term dictionary 73). Difference data from the field database 74) is stored in the version database 75 together with a unique version number. Such management is performed by the version management function of the data management unit 63.

  Next, in step S108, the customer data evaluation unit 61 performs an evaluation process of data related to the charging discount evaluation of each customer. More specifically, the customer data evaluation unit 61 determines the number of unknown words for each customer who is a user of the expert extraction device 5 that has transmitted an unknown word registered as a new technical term by the technical term analysis unit 62. And the number of words (or the number of phrases) of sentences included in unknown word data is totaled. Only the number of newly registered unknown words is counted. The unknown words that have already been registered and the unknown words that cannot be handled as technical terms are not useful for upgrading the technical term dictionary 73 and are subject to discount evaluation. It is because it is not possible. Also, counting the number of words (or the number of clauses) in a sentence included in unknown word data is based on the judgment whether or not the unknown word can be handled as a technical term, and in the specialized field corresponding to the unknown word. This is because it is very useful for judgment in classification, and should be subject to discount evaluation. Note that instead of counting the number of words (or the number of phrases), the length of the character string is totaled, or the character string delimiting method (here, paragraph, sentence, phrase, and word) is supported. The amount corresponding to the character string may be aggregated, for example, by summing up predetermined evaluation values.

  Further, the customer data evaluation unit 61 refers to the reference billing rate data 71 to calculate a billing rate corresponding to the total result (for example, when there are 10 or more unknown words and the total number of words is 50 or more). The obtained charging rate is registered in the customer-specific charging database 72. These evaluation processes are performed by the totaling function of the customer data evaluation unit 61. Note that the discount rate (or billing rate) calculated for billing discount may be a discount amount, or may be based on other discount methods (for example, giving points for purchase).

  In this way, a customer who has sent many unknown words and related sentences can receive a large discount, so that transmission of these unknown word data from the expert extraction device 5 can be promoted. Therefore, the dictionary server 6 (operator, for example, a software maker that implements an expert extraction device) can collect more unknown words and related sentences, so it is more accurate and registers more technical words. The specialized vocabulary dictionary 73 can be provided.

  Subsequently, in step S109, the data distribution unit 64 provides the latest technical language dictionary 73 and the specialized field database 74 to the cryptographic processing unit 81 at predetermined intervals or according to the customer's request, and the cryptographic processing unit 81 provides them. It encrypts and transmits to the expert extraction apparatus 5. Typically, the data distribution unit 64 determines the versions of the technical term dictionary 35 and the specialized field database 32 stored in the storage unit 3 of the expert extraction device 5 (for example, the version number is obtained from the expert extraction device 5). The difference data between these and the latest technical language dictionary 73 and the specialized field database 74 is read from the version database 75 and transmitted to the expert extracting device 5. These latest technical language dictionary 73 and specialized field database 74 are charged, and the charge amount is calculated according to, for example, the version number and the difference data amount, and is given to the customer data evaluation unit 61. The customer data evaluation unit 61 refers to the customer-specific charging database 72. If a discount amount is registered, the customer data evaluation unit 61 uses, as a final charging amount, an amount obtained by discounting the discount amount from the charging amount received from the data distribution unit 64. Register in the customer-specific billing database 72. This billing process is performed by the billing function of the customer data evaluation unit 61.

  As described above, the processing in steps S103 to S109 shown in FIG. 5 is performed in the dictionary server 6, but the processing in the next step S110 is performed in the expert extraction device 5. This will be described below. Note that the processing of steps S103 to S107 (or all processing of the dictionary server 6) may be performed by a human (for example, an expert such as a dictionary creator). The processing (work) in this case is performed, for example, via a user I / F device (not shown) of the dictionary server 6, and a new technical language dictionary or the like is created by a dictionary creating means realized by a predetermined program (editing tool or the like). Is done.

  In step S110, the unknown word optimization processing unit 29 of the expert extraction device 5 receives the latest technical language dictionary or the like from the dictionary server 6 via the communication unit 4 and the encryption processing unit 41, and receives the received technical language dictionary. Based on the above, an optimization process is performed to reflect the unknown word (new technical term) and the total value related to the unknown word in the word database 31 for each user. As described above, the user-specific word database 31 stores the technical terms (and specialized fields) and the number of times related to each user specified by the user specifying unit 22. The optimization processing unit 29 reflects the contents of the latest technical term dictionary and the like on these. For example, the unknown word optimization processing unit 29 newly registers specialized words and specialized fields newly registered in the latest technical term dictionary or the like in the user-specific word database 31, and these unknown words (new technical terms) ) Is already registered in the user-specific word database 31, the corresponding total value is reflected in the newly registered technical terms and specialized fields (that is, after taking over), and is already registered. Delete the unknown word.

  Here, if an unknown word that has already been registered in the user-specific word database 31 as a new technical term is not newly registered in the latest technical term dictionary or the like, the unknown word optimization processing unit 29 determines the unknown word. Is automatically deleted. This is because the unknown words that are not handled as technical words in the dictionary server 6 are mostly miswritten or misused, and are preferably deleted. However, there are cases where the user of the expert extraction device 5 wants to use the unknown word for a special reason (for example, because it is a new term related to unprecedented research). In consideration of this, the unknown word optimization processing unit 29 does not automatically delete the unknown word, but makes an inquiry to the device user on the display screen, for example, and the device user permits the deletion. It is preferable to delete the unknown word only if

  Through the above processing, the unknown word extracted by the expert extraction device 5 is registered (or not registered) in the latest expert word dictionary or the like, and finally optimized by the expert extraction device 5. These processes are repeated each time a new unknown word is extracted by the expert extraction device 5 and transmitted to the dictionary server 6.

<6. Effect>
As described above, in the present embodiment, the user automatically registers the unknown word without registering it in the technical word dictionary by the technical word extraction unit 21 of the expert extracting device 5, and includes many of these unknown words. It is possible to provide a dictionary server that can create a technical word dictionary in which unknown words used by users are widely registered in the dictionary server 6 and give them to the expert extraction device 5.

  Further, the unknown word cutout location setting unit 28 of the expert extraction device 5 can easily cut out or select a character string that permits transmission from the character string including the unknown word to the dictionary server 6 for the user of the device. Therefore (delimitation setting means or character string selection means), it is possible to prevent a sentence requiring confidentiality from being transmitted to the dictionary server 6.

  Furthermore, since the customer data evaluation unit 61 of the dictionary server 6 receives a larger discount for customers who send more unknown words and related sentences, the transmission of these unknown word data from the expert extraction device 5 is promoted. be able to. Therefore, it is possible to provide the technical term dictionary 73 in which many technical terms are registered accurately from the dictionary server 6.

It is a block diagram which shows the structure of the expert extraction apparatus and dictionary server which concern on one Embodiment in this invention. In the said embodiment, it is a figure for demonstrating the example in which a certain unknown word is registered with respect to a technical language dictionary. It is a figure which shows the specific example of the word database classified by user constructed | assembled in a memory | storage part by the total part in the said embodiment. It is a figure for demonstrating the structure of the specialized field database in the said embodiment. It is a flowchart which shows the flow of the process relevant to the unknown word in the expert extraction apparatus and dictionary server of the said embodiment. In the said embodiment, it is a figure which shows the example of storage of the unknown word data containing the text before being cut out by selection of the user of an apparatus. It is a figure which shows an example of the screen display for making the user of an apparatus select the part which permits the transmission to a dictionary server by the unknown word cutout location setting part in the said embodiment. It is a figure which shows another example of the screen display for making the user of an apparatus select the part which permits the transmission to a dictionary server by the unknown word cutout location setting part in the said embodiment. It is a figure which shows the further another example of the screen display for making the user of an apparatus select the part which permits the transmission to a dictionary server by the unknown word clipping location setting part in the said embodiment. It is a figure which shows the example of a screen display in which the part which permits the transmission to the dictionary server in the said embodiment was selected in multiple numbers by the user of the apparatus.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... User I / F apparatus 2 ... Arithmetic processing part 3 ... Memory | storage part 4 ... Communication part 5 ... Expert extraction apparatus 6 ... Dictionary server 7 ... Memory | storage part 8 ... Communication part 21 ... Technical term extraction part 22 ... User identification part DESCRIPTION OF SYMBOLS 23 ... Aggregation part 24 ... Search part 25 ... Character recognition part 26 ... Network monitor 27 ... Document management server 28 ... Unknown word extraction part setting part 29 ... Unknown word optimization process part 31 ... User-specific word database 32 ... Special field database 33 ... Document database 34 ... General word dictionary 35 ... Technical word dictionary 36 ... Unknown word data 41 ... Encryption processing part 51 ... Scanner 61 ... Customer data evaluation part 62 ... Technical word analysis part 63 ... Data management part 64 ... Data distribution part 71 ... Billing rate data for reference 72 ... Billing database by customer 73 ... Technical dictionary 74 ... Specialty field database 75 ... Barge Database 81 ... Cryptographic processing part

Claims (18)

An expert extraction device that extracts an expert who is presumed to be familiar with a predetermined technical term based on a predetermined electronic document,
A dictionary in which multiple technical terms are registered,
By referring to the dictionary, the terminology extracting means for extracting the terminology included in the electronic document and extracting an unknown word not registered in the dictionary;
A totaling unit that counts the association frequency including at least one of the number of extractions of the technical terms extracted by the technical term extraction unit and a value corresponding to the number of extractions in association with the person related to the electronic document;
An input means for accepting a technical term or specialized field designated by a predetermined user;
The expert is searched by searching for a person who is associated with the highest relevance frequency by the tabulating means with respect to the technical word accepted by the input means or the technical term corresponding to the specialized field accepted by the input means. A search means for extracting
An expert extraction device comprising: an unknown word providing unit that supplies the unknown word extracted by the technical word extraction unit to a predetermined dictionary providing device outside the device. The technical term extraction means extracts a plurality of character strings including the unknown word,
2. The expert extraction according to claim 1, wherein the unknown word providing unit provides a part or all of the plurality of character strings extracted by the technical word extracting unit to the dictionary providing apparatus. apparatus.   It further comprises delimiter setting means for setting a range to be given to the dictionary providing device among the plurality of character strings based on a delimiter method that is alternatively selected by the user from a plurality of predetermined delimiter methods. The expert extraction device according to claim 2, wherein:   The expert extraction device according to claim 2, further comprising a character string selection unit that allows the user to select a character string to be given to the dictionary providing device among the plurality of character strings.   The technical term extraction unit extracts an unknown word that has already been extracted included in the electronic document as the technical term by referring to the unknown word that has already been extracted. Expert extraction equipment.   When a new dictionary to replace the dictionary is received from the dictionary providing device and the unknown word is registered in the new dictionary, the degree of association of the unknown word and the associated person are determined as the new dictionary. The expert extraction device according to claim 1, further comprising: a dictionary receiving unit that takes over as the association degree of the technical word corresponding to the unknown word registered in and the person who is associated with the unknown word. Further comprising specialized field data in which the correspondence between the specialized field and the technical term is registered;
The dictionary receiving means receives new specialized field data to be substituted for the specialized field data from the dictionary providing device, and when the unknown word is registered in the new specialized field data, the association frequency of the unknown word And the associated person is taken over as the related frequency of the technical term corresponding to the unknown word corresponding to the specialized field registered in the new specialized field data and the associated person. Item 7. The expert extraction device according to Item 6.   The expert extraction device according to claim 1, wherein the unknown word providing unit includes an encryption processing unit that encrypts data including the unknown word to be given to the dictionary providing device.   9. The expert extraction according to claim 8, wherein the cryptographic processing means encrypts the data based on a predetermined public key corresponding to the dictionary providing device by using a public key cryptosystem. apparatus. A dictionary in which a plurality of technical terms are registered, and a technical term extraction unit that extracts the technical terms contained in a predetermined electronic document by referring to the dictionary and extracts unknown words that are not registered in the dictionary; A dictionary that receives the unknown word from an expert extraction device including an unknown word providing unit that supplies the unknown word extracted by the technical word extraction unit to the outside of the device, and gives a newly created dictionary to the expert extraction device A providing device,
An unknown word receiving means for receiving the unknown word from the expert extraction device;
A dictionary creating means for creating a new dictionary including part or all of the received unknown words;
A dictionary providing device, comprising: a dictionary providing unit that provides the expert extracting device with a new dictionary created by the dictionary creating unit.   The dictionary creating means determines whether or not the unknown word received by the unknown word receiving means is a technical word, and only when it is determined that the unknown word is a technical word, specialized word analyzing means for registering in the new dictionary The dictionary providing apparatus according to claim 10, comprising:   The customer data evaluation unit according to claim 10, further comprising a customer data evaluation unit that performs a discount evaluation on a fee for providing the new dictionary according to an amount corresponding to an unknown word received by the unknown word reception unit. Dictionary providing device. The unknown word receiving means receives a plurality of character strings including the unknown word,
The dictionary providing apparatus according to claim 12, wherein the customer data evaluation unit performs the discount evaluation according to an amount corresponding to the plurality of character strings.   The customer data evaluation means performs the discount evaluation according to an amount corresponding to an unknown word determined to be a technical word by the dictionary creation means among the unknown words received by the unknown word receiving means. The dictionary providing apparatus according to claim 12 or 13, characterized in that it is characterized in that: An expert extraction device that extracts an expert who is presumed to be familiar with a predetermined technical term based on a predetermined electronic document, wherein a dictionary in which a plurality of technical terms are registered, and the dictionary is referred to, Extracting the technical terms contained in the electronic document, extracting technical terms that are not registered in the dictionary, and extracting the technical terms extracted by the technical terms extracting means A totaling means for counting the association frequency including at least one of the corresponding values in association with a person related to the electronic document, an input means for receiving a technical term or a specialized field designated by a predetermined user, Corresponding to the technical terms accepted by the input means or the technical terms corresponding to the specialized field accepted by the input means with the highest relevance frequency by the aggregation means A search means for extracting the expert by searching for a registered person, and an unknown word that gives the unknown word extracted by the technical word extraction means to a predetermined dictionary providing apparatus outside the apparatus via a predetermined communication line An expert extraction device comprising providing means;
A dictionary in which a plurality of technical terms are registered, and a technical term extraction unit that extracts the technical terms contained in a predetermined electronic document by referring to the dictionary and extracts unknown words that are not registered in the dictionary; A dictionary that receives the unknown word from an expert extraction device including an unknown word providing unit that supplies the unknown word extracted by the technical word extraction unit to the outside of the device, and gives a newly created dictionary to the expert extraction device An apparatus for providing an unknown word receiving means for receiving the unknown word from the expert extraction device via the communication line and a dictionary creation for creating a new dictionary including a part or all of the received unknown word And a dictionary providing device comprising: a dictionary providing device for providing the new dictionary created by the dictionary creating device to the expert extracting device via the communication line. And wherein the door, dictionary update system. A method for controlling an expert extraction device that extracts an expert who is presumed to be familiar with a predetermined technical term based on a predetermined electronic document,
A technical term extraction step of extracting the technical terms contained in the electronic document by referring to a dictionary in which a plurality of technical terms are registered, and extracting unknown words that are not registered in the dictionary;
A tabulation step of tabulating the number of associations including at least one of the number of extractions of the technical terms extracted in the technical term extraction step and a value corresponding to the number of extractions in association with a person related to the electronic document;
An input step for accepting a technical term or specialized field designated by a predetermined user;
The expert is searched by searching for a person who is associated with the highest relevance frequency by the aggregation step with respect to the technical word accepted in the input step or the technical term corresponding to the technical field accepted in the input step. A search step to extract
A control method comprising: an unknown word providing step of providing the unknown word extracted in the technical word extraction step to a predetermined dictionary providing device outside the device. To an expert extraction device that extracts experts who are assumed to be familiar with a given technical term based on a given electronic document,
A technical term extraction step of extracting the technical terms contained in the electronic document by referring to a dictionary in which a plurality of technical terms are registered, and extracting unknown words that are not registered in the dictionary;
A tabulation step of tabulating the number of associations including at least one of the number of extractions of the technical terms extracted in the technical term extraction step and a value corresponding to the number of extractions in association with a person related to the electronic document;
An input step for accepting a technical term or specialized field designated by a predetermined user;
The expert is searched by searching for a person who is associated with the highest relevance frequency by the aggregation step with respect to the technical word accepted in the input step or the technical term corresponding to the technical field accepted in the input step. A search step to extract
A control program for executing an unknown word providing step of providing the unknown word extracted in the technical word extraction step to a predetermined dictionary providing device outside the device. To an expert extraction device that extracts experts who are assumed to be familiar with a given technical term based on a given electronic document,
A technical term extraction step of extracting the technical terms contained in the electronic document by referring to a dictionary in which a plurality of technical terms are registered, and extracting unknown words that are not registered in the dictionary;
A tabulation step of tabulating the number of associations including at least one of the number of extractions of the technical terms extracted in the technical term extraction step and a value corresponding to the number of extractions in association with a person related to the electronic document;
An input step for accepting a technical term or specialized field designated by a predetermined user;
The expert is searched by searching for a person who is associated with the highest relevance frequency by the aggregation step with respect to the technical word accepted in the input step or the technical term corresponding to the technical field accepted in the input step. A search step to extract
A computer-readable recording medium recording a program for executing an unknown word providing step of supplying the unknown word extracted in the technical word extraction step to a predetermined dictionary providing device outside the device.

Images