JP4283038B2 - Document registration device, document search device, program, and storage medium - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a document registration apparatus for registering document data in order to search a document that matches an input search expression from a plurality of registered document data, and a document that matches the input search expression is registered. The present invention relates to a document retrieval apparatus that retrieves a plurality of document data, a program that realizes these apparatuses, and a storage medium that stores the program.
[0002]
[Prior art]
As a technique for searching in multiple languages, for example, there is one disclosed in Patent Document 1. In such a technique, a bilingual dictionary between languages is used to implement cross-language search.
[0003]
However, even if a full-fledged cross-language search function is not provided, it is fully conceivable that documents in a plurality of languages are stored in the document database. For example, in a document database used in Japan, it is not uncommon to store English text or English mixed text.
[0004]
On the other hand, in document search technology, for example, normalization that absorbs fluctuations of notations such as “computer” and “computer” and singular / plural forms such as “chair” and “chairs” in order to prevent omission of search. Technology exists. These normalizations should be performed for each document to be registered in the language used in the document, and just because Japanese document databases are used in Japan, only Japanese normalization is performed in the document database. This leads to omission of search for documents written in languages other than Japanese. In order to solve such a problem, some search engines on the WWW are provided with an interface for specifying a language.
[0005]
[Patent Document 1]
JP-A-8-212229 gazette
[Problems to be solved by the invention]
However, in document database technology, the language of the country of sale is the main language, and other languages are sometimes mixed in the document described in the main language. Or, even in a document written in a certain language, there are cases where English, which is generally widely used, is sometimes mixed, and in these cases, it is not possible to specify a language other than the main language. Therefore, there is a problem that sufficient normalization cannot be performed and there is a possibility of a search omission.
[0007]
An object of the present invention is to perform a search with few leaks on a document described in a plurality of languages.
[0008]
According to the first aspect of the present invention, in the document registration apparatus for registering the document data in order to search a document that matches the input search formula from a plurality of registered document data, the registration is attempted. A document data receiving unit that receives input of document data, a language designation receiving unit that receives designation of a language related to the one document to be received, and a plurality of languages including the received designated language each having a different character code area. The document data is normalized for one language among the plurality of languages to create normalized data, and the normalized data is normalized for one other language. By creating new normalized data and creating new normalized data in the same way for all the other languages, Create data, create an index based on the one normalized data, and normalize the document data for each of the plurality of languages when the plurality of languages hold the same character code area To create a plurality of normalized data having the same number as the number of the plurality of languages, and to create an index based on the plurality of normalized data, the index and the received document data And a registration unit for registering the document.
[0009]
Therefore, it is possible to normalize one document in a plurality of languages and execute a search with few omissions on a document described in a plurality of languages.
[0010]
According to a second aspect of the present invention, in the document registration apparatus according to the first aspect, the index creating means designates a plurality of languages by designating a predetermined language in addition to the designated language accepted by the language designation accepting means. And performing the normalization.
[0011]
Therefore, it is possible to reduce the trouble of specifying a plurality of languages for one document on the user side, and to prevent a search omission due to forgetting to specify a language.
[0012]
According to a third aspect of the present invention, in the document registration apparatus according to the second aspect, the index creating means designates English as the predetermined language.
[0013]
Therefore, it is possible to prevent omission of search for documents mixed with English by setting the default language for English which is likely to appear suddenly in a document in a certain language without specifying a language.
[0018]
According to a fourth aspect of the present invention, there is provided a document retrieval apparatus for retrieving a document that matches an input retrieval formula from a plurality of document data registered by the document registration apparatus according to the first aspect . A request accepting means; a language designation accepting means for accepting a language specification relating to the accepted one search request; and normalizing the accepted search request in a plurality of languages including the accepted designated language, and the document data in a plurality of languages. And a search means for executing a search of the plurality of document data using an index created by normalizing the document.
[0019]
Therefore, normalization can be performed in a plurality of languages with respect to one search request, and a search with less omission can be performed on a document described in a plurality of languages.
[0020]
According to a fifth aspect of the present invention, in the document retrieval apparatus according to the fourth aspect , the retrieval unit designates a predetermined language in addition to the designated language accepted by the language designation acceptance unit, and uses the plurality of languages. The normalization is performed.
[0021]
Accordingly, it is possible to reduce the trouble of designating a plurality of languages for one search request on the user side, and to prevent a search omission due to forgetting to designate a language.
[0022]
The invention described in claim 6 is the document search apparatus according to claim 5 , wherein the search means specifies English as the predetermined language.
[0023]
Therefore, it is possible to prevent omission of search for documents mixed with English by setting the default language for English which is likely to appear suddenly in a document in a certain language without specifying a language.
[0024]
The invention according to claim 7 is readable by a computer that causes a computer to execute processing for registering the document data in order to search a document that matches the input search formula from a plurality of registered document data. In the program, the document data receiving process for receiving the input of the document data to be registered, the language specifying receiving process for receiving the specification of the language relating to the received one document, and the plurality of languages including the received specified language are different. When the character code area is held, the document data is normalized with respect to one language out of the plurality of languages to create normalized data, and the normalized data is Normalize one language and create new normalized data, and for all the other languages, When creating one normalized data by creating new normalized data, creating an index based on the one normalized data, and when the plurality of languages hold the same character code area In addition, the document data is normalized for each of the plurality of languages to create a plurality of normalized data having the same number as the number of the plurality of languages, and an index is created based on the plurality of normalized data And a registration process for registering the index and the received document data.
[0025]
Therefore, it is possible to normalize one document in a plurality of languages and execute a search with few omissions on a document described in a plurality of languages.
[0026]
The invention described in claim 8 is a computer-readable program that causes a computer to execute a process of searching a plurality of document data registered by the program of claim 7 for a document that matches an input search expression. A search request reception process for receiving a search request for search; a language specification reception process for receiving a language specification relating to this one search request to be received; and normalizing the received search request in a plurality of languages including the received specified language; A program that causes a computer to execute search processing for searching for the plurality of document data using an index created by normalizing the document data in a plurality of languages.
[0027]
Therefore, normalization can be performed in a plurality of languages with respect to one search request, and a search with less omission can be performed on a document described in a plurality of languages.
[0028]
The invention according to claim 9 is a storage medium storing a computer-readable program, wherein the program is the program according to any one of claims 7 and 8. It is a medium.
[0029]
Therefore, the stored program produces the same operations and effects as the invention according to claim 7 or 8 .
[0030]
DETAILED DESCRIPTION OF THE INVENTION
An embodiment of the present invention will be described.
[0031]
FIG. 1 is a block diagram of electrical connection of the document search system 1 according to the present embodiment. The document search system 1 is an apparatus that implements the document registration apparatus and the document search apparatus of the present invention. As shown in FIG. 1, a CPU 11 that performs various operations and controls each part of the document search system 1 in a centralized manner. A memory 12 composed of a ROM and a RAM is connected by a bus 13.
[0032]
The bus 13 is connected to a magnetic storage device 14 such as a hard disk, an input device 15 such as a keyboard and a mouse, a display device 16 and a storage medium reader 18 that reads a storage medium 17 such as an optical disk via a predetermined interface. And a predetermined communication interface 19 for communicating with the network 2 is connected. As the storage medium 17, various media such as an optical disk such as a CD and a DVD, a magneto-optical disk, and a flexible disk can be used. As the storage medium reading device 18, specifically, an optical disk device, a magneto-optical disk device, a flexible disk device, or the like is used according to the type of the storage medium 17.
[0033]
The document retrieval system 1 reads the program 20 that implements the program of the present invention from the storage medium 17 that implements the storage medium of the present invention, and installs it in the magnetic storage device 14. These programs may be downloaded and installed via the network 2 such as the Internet. As a result of this installation, the document search system 1 becomes ready to execute a predetermined process described later. The program 20 may operate on a predetermined OS.
[0034]
Next, processing executed by the document search system 1 will be described. FIG. 2 is a functional block diagram of functions realized by the document search system 1 based on the program 20. 3 and 4 are flowcharts for explaining processing executed by the document search system 1.
[0035]
First, processing when document data is registered by the document search system 1 will be described with reference to FIGS. First, when storing the document data in the document search system 1, the user designates a specific language type and inputs the document data to be registered (document data receiving means, language designation receiving means, document data). Reception process, language designation reception process) (Y in step S1). This designation (language designation) is sent to the language designation unit 21 (step S2). The language designation unit 21 adds a language designation of a specific language, “English” in this example, to the language designation by the user. It is set as language information (step S3).
[0036]
The document data storage unit 22 receives the document data and language information, and first refers to a language designation / character code area correspondence table 31 as shown in FIG. 5 (step S4). In this table 31, a language designation / character code area correspondence table as shown in FIG. 5 is recorded. In this example, various languages 32 such as Japanese and English are registered in association with character code ranges (character code area 33) used in the language.
[0037]
Then, two languages designated by the language information are searched from the language 32, and it is determined whether or not the character code areas 33 corresponding to the two languages overlap each other (step S5). For example, when the plurality of languages included in the language information are “Japanese” and “English”, the character code area 33 does not overlap (N in step S5). In such a case, the document data storage unit 22 normalizes the user specified language (in this example, “Japanese”) on the received document data (step S6), and specifies the language for the result. Normalization of the specified language (in this example, “English”) of the unit 21 is performed (step S7), and using the result, the document data storage unit 23 (magnetic storage device 14) serving as a database in which the document data group is stored. Are created (stored in an index table 41 of FIG. 5 described later) (step S8), and document data is stored (step S9).
[0038]
For example, when the plurality of languages included in the language information are “French” and “English”, there is an overlap in the character code area 33 (Y in step S5). In this case, the document data storage unit 22 normalizes the user specified language (in this example, “French”) to the received document data (step S10), and uses the result to store the document data in the document data storage unit 23. (Step S11), the language specification unit 21 normalizes the specified language (in this example, “English”) (step S12), and uses the result to create an index in the document data storage unit. Create (step S13) and store document data (step S14). Steps S6 to S8 and S10 to S13 realize index creation means and index creation processing, and Steps S9 and S14 realize document registration means and document registration processing.
[0039]
The index created by such processing is stored in the index table 41 of FIG. Below, the index of this index table 41 is demonstrated concretely. In the example of FIG. 5, an example of the index table 41 when the document 1 “messaging manager” is stored in “Japanese” and “English” and the document 2 having the same contents is stored in “French” and “English”. Indicates.
[0040]
In the document 1, since different normalization is applied to different character code portions, only the number of words constituting the document is created. That is, “Japanese” (0x3000-0x30ff, 0x3200-0x33ff, 0x4e00-0x9fff, 0xf900-0xfaff, 0xff00-0xff9f in the character code area 33) and “English” (0x0020-0x00ff in the character code area 33) are Since there is no overlapping character code range, the document 1 is first normalized in Japanese, and then the result is subjected to English normalization. The result of normalizing “Japanese” on document 1 is “messaging manager”, and the result of normalizing “English” is the normalization rule for English because there is a normalization rule related to the ing form. , "Message manager". As a result, the two words “message” and “manager” constituting the normalized document 1 are registered as indexes to the document 1. Therefore, in the index table 41, two words “message” and “manager” are registered in the column of the word notation 42, and the document 1 is registered in the column of the document 43 in association with these.
[0041]
On the other hand, in document 2, in order to create an index for normalization of different languages, an index for variations generated by normalization is created. That is, when the document 2 “messaging manager” is stored in “French” and “English”, according to the language designation-character code area correspondence table 31, “French” (0x0020-0x00ff in the character code area 33). Has the same character code range as “English”, a normalized document obtained by normalizing document 2 with “French” and a normalized document obtained by normalizing with “English” are created. In the case of normalization with “French”, since “messaging → message” does not exist in the French normalization rule, a normalized document “messaging manager” is generated. In addition, when normalized in “English”, a normalized message “message manager” is generated. Different words are extracted from these two normalized documents, and three words “message”, “messaging”, and “manager” are registered as an index to the document 2. Therefore, in the index table 41, three words “message”, “messaging”, and “manager” are registered in the column of the word notation 42, and the document 2 is registered in the column of the document 43 in association with each of them. .
[0042]
Next, a case where document data is searched by the document search system 1 will be described with reference to FIGS. In the following, the word “windows” will be described as an example. First, the user inputs a search request to the request input unit 24 (search request reception means, search request reception processing) (Y in step S21), and further specifies a language (for example, “Japanese”, “English”, “ "French") (Language designation accepting means, language designation accepting process) (Y in step S22). The language designation unit 21 normally adds “English” as a specific language to the language designation from the user (step S23). In this example, “English” is already included in the language designation. Nothing is added.
[0043]
The multilingual expansion unit 25 normalizes the search request word with respect to the specified language, in this example, “Japanese”, “English”, “French” (step S24). As a result, in this example, “windows”, Get “window”, “window”. The request input unit 24 receives this, and in this example, the overlapping “windows” are combined into one, and as a result of normalization, “windows” and “window” in this example are sent to the search unit 26 (step S25).
[0044]
As a result of normalization, the search unit 26 performs a search using “index” in the index table 43 with “windows” or “window” in this example (step S26), and outputs the result as a search result (step S27). ). Search means and search processing are realized by steps S23 to S26.
[0045]
In this example, if a document containing the description “windows” is registered as “Japanese”, the word “windows” is registered as “English” or “French” or “Japanese” or “English”. If the search is performed, the word “window” is hit in the search, and a search without omission can be realized.
[0046]
【The invention's effect】
The inventions according to claims 1, 6, 9, 10, and 11 normalize one document and one search request in a plurality of languages, and perform a search with few leaks for a document described in a plurality of languages. Execution can be performed.
[0047]
The inventions described in claims 2 and 7 are the inventions described in claims 1 and 6, reducing the trouble of specifying multiple languages for one document and one search request on the user side, and forgetting to specify a language. Search omission due to can be prevented.
[0048]
The inventions of claims 3 and 8 are the inventions of claims 2 and 7, in which the default language is used for English which is likely to appear suddenly in a document in a language, without particularly specifying the language. By doing so, it is possible to prevent omission of search for documents mixed with English.
[0049]
The invention according to claim 4 is the invention according to any one of claims 1 to 3, wherein the index is created by performing a plurality of normalizations as document registration processing when a plurality of languages are designated. As a result, even if the size of the index is increased, it is possible to deal with a plurality of languages in one document without depending on the characters used by the language.
[0050]
The invention according to claim 5 is the invention according to any one of claims 1 to 4, and the effect of normalization performed in a plurality of languages as a document registration process when a plurality of languages are designated. When there is no influence on each other, normalization of one language and normalization of another language are simultaneously performed on one document, so that the size of an index to be created can be reduced while preventing a search omission.
[Brief description of the drawings]
FIG. 1 is a block diagram of electrical connection of a document search system according to an embodiment of the present invention.
FIG. 2 is a functional block diagram of a document search system.
FIG. 3 is a flowchart of processing when registering document data.
FIG. 4 is a flowchart of processing when retrieving document data.
FIG. 5 is an explanatory diagram of a language designation / character code area correspondence table;
FIG. 6 is an explanatory diagram of an index table.
[Explanation of symbols]
1 Document Registration Device, Document Search Device 17 Storage Medium 20 Program

Claims (9)

In a document registration apparatus for registering the document data in order to search a document that matches the input search formula from a plurality of registered document data,
Document data receiving means for receiving input of document data to be registered;
Language designation accepting means for accepting designation of a language relating to the one document to be accepted;
When the plurality of languages including the received designated language hold different character code areas, the document data is normalized with respect to one language out of the plurality of languages to obtain normalized data. Create new normalized data by normalizing the normalized data for one other language, and create new normalized data for all the other languages in the same way By creating one normalized data, creating an index based on the one normalized data, when the plurality of languages hold the same character code area, for the document data, wherein the plurality of each language performs normalization to create as many plurality of normalized data of the plurality of languages, and indexing means of making indexes based on the plurality of normalized data
Registration means for registering the index and the received document data;
A document registration apparatus comprising:   2. The document registration according to claim 1, wherein the index creation unit designates a predetermined language in addition to the designated language accepted by the language designation acceptance unit and performs the normalization in the plurality of languages. apparatus.   The document registration apparatus according to claim 2, wherein the index creating unit designates English as the predetermined language. In a document search apparatus that searches a plurality of document data registered by the document registration apparatus according to claim 1 for a document that matches the input search formula,
Search request accepting means for accepting the search request for the search;
Language designation accepting means for accepting the designation of the language related to the one search request to be accepted;
Retrieval means for normalizing the accepted search request in a plurality of languages including the accepted designated language and performing a search of the plurality of document data using an index created by normalizing the document data in a plurality of languages When,
A document retrieval apparatus comprising:   5. The document search apparatus according to claim 4, wherein the search unit specifies a predetermined language in addition to the specified language received by the language specification reception unit and performs the normalization in the plurality of languages. .   The document search apparatus according to claim 5, wherein the search unit specifies English as the predetermined language. In a computer-readable program for causing a computer to execute a process of registering the document data in order to search a document that matches the input search formula from a plurality of registered document data,
Document data reception processing for receiving input of document data to be registered;
A language designation accepting process for accepting designation of a language relating to the one document to be accepted;
When the plurality of languages including the received designated language hold different character code areas, the document data is normalized with respect to one language out of the plurality of languages to obtain normalized data. Create new normalized data by normalizing the normalized data for one other language, and create new normalized data for all the other languages in the same way By creating one normalized data, creating an index based on the one normalized data, when the plurality of languages hold the same character code area, for the document data, the normalization to make creating a plurality of normalized data as many of said plurality of languages for each of a plurality of languages, the indexing process for creating an index based on the plurality of normalized data
A registration process for registering this index and the received document data;
A program that causes a computer to execute. In a computer-readable program for causing a computer to execute processing for searching a plurality of document data registered by the program of claim 7 for a document that matches an input search expression,
A search request receiving process for receiving a search request for the search;
A language designation accepting process for accepting designation of a language related to the one search request to be accepted;
A search process for normalizing the received search request in a plurality of languages including the specified language that has been received, and executing a search for the plurality of document data using an index created by normalizing the document data in a plurality of languages When,
A program that causes a computer to execute. In a storage medium storing a computer-readable program,
A storage medium, wherein the program is the program according to any one of claims 7 and 8.

Images