JP2005327316A - Integrated retrieval system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an integrated retrieval device allowing a user to cheaply and easily use it without considering structures of individual retrieval devices, and capable of properly retrieving data demanded by the user. <P>SOLUTION: When a keyword is inputted by an input means 2a, a keyword expansion means 2c expands the inputted keyword to a keyword group related thereto by using a keyword expansion function of a retrieval device having the keyword expansion function; and a data registration control means 2d registers the keyword based on the expanded keyword group to a retrieval device of a registration object by relating it with data. Thereby, even when data are retrieved by the retrieval device without having a keyword expansion function, the data can be properly retrieved like the retrieval device having the keyword expansion function. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an integrated search device that uses a plurality of search devices to search for data based on a specified keyword.

2. Description of the Related Art Conventionally, a search device (database system) is known in which data such as text data is associated with a keyword and registered in a database, and data is searched based on a specified keyword.
Such a search device is provided according to the purpose or for each user. For example, a plurality of search devices may exist in the same processing terminal device or in the same network.

For the purpose of effectively using the plurality of search devices described above, a plurality of search devices are integrated.
For example, the invention described in Japanese Patent Publication No. 60-58492 discloses a database search method for searching data in a plurality of databases of the same type. The invention described in Japanese Patent Application Laid-Open No. 60-195653 discloses a database integration method for integrating and managing a plurality of distributed databases.

  Here, in the database of the search device, by defining the definition of logical structure (so-called schema), the expression of attributes related to data is unified, the data types of data belonging to the same attribute are unified, and the same By representing meaningful data with the same data values, data equivalence is ensured. For example, in the “unit price” as an attribute representing the amount of money, the data is defined as numerical data, and when the unit of data in the attribute is “yen”, the “unit price” attribute includes , Data of units such as “dollar” and “thousand yen” are never mixed.

  However, text data has a uniform definition as character type or character string type data, but there are multiple terms that have the same meaning in text data, and the expressions are different. Are not always represented by the same data value (word). For example, when a foreign word is written in katakana, there may be a plurality of words having the same meaning depending on the way of expression.

  Here, the keywords associated with the text data by the search device include a case in which the words in the text data are intact or a case in which the words are given in relation to the text data, and the search is performed by the search device. In this case, it is determined whether or not the keyword specified by the user or the like matches the keyword associated with the text data.

  Therefore, when searching for text data as described above, if a phrase having the same meaning as the keyword specified by the user or the like is associated as a keyword of the text data but does not match, the text data is set as a search target. There is a problem with data values that is not detected and the data required by the user cannot be detected properly.

  As a method for solving this problem, there are a first method of handling a plurality of words having the same meaning in a single word (control word) and a method for expanding a specified keyword (for example, synonym expansion). Two methods are known.

  The first method uses the controlled / registered controlled words as keywords that correspond to the text data, and uses the controlled words as keywords specified by the user when searching for data. Can be detected. On the other hand, when registering or retrieving text data, a controlled word to be a keyword must be selected from the controlled word set.

  In the second method, when registering text data, the user inputs a keyword, expands the input keyword into a plurality of synonyms, and associates the plurality of synonyms with the text data. Thus, the keyword designated by the user thereafter and the keyword associated with the data are effectively matched. In addition, when searching, the user can freely select a keyword, expand the selected keyword into a plurality of synonymous keywords, and search for text data according to the expanded keywords, so that the user needs Can be detected. In addition, in order to implement | achieve the keyword expansion | deployment function using a synonym in a search device, the function to search with respect to a synonym dictionary and a synonym dictionary is required.

  By applying either of these two methods to a plurality of search devices, the above-described problem relating to data values can be solved. However, when the first method is applied, it is necessary to unify controlled word sets used in a plurality of search devices. In addition, when the second method is applied, it is necessary to provide a keyword expansion function for each of the plurality of search devices, which increases costs.

  Further, when processing such as registration or search is performed for a plurality of search devices, the purposes such as the data to be handled and the situation to be handled are different in each search device. However, even if the database configuration is the same, the schema is different even if the database configuration is the same, and there is a problem that each search device cannot be used in the same procedure.

  As a method for solving this difference in the structure of the database, there is a method of converting a search formula designated by the user into a search formula or operation command string suitable for each search device. For example, in the invention described in Japanese Patent Publication No. 2-67682, a structured database can be searched by converting a search expression written by a user in SQL (Structured Query Language) into DML (Database Operation Language). I am doing so. That is, by constructing a relational database system on the structural database system, the user can access the structural database in the same procedure as the relational database system.

  In addition, when the configuration of a plurality of search devices is a relational database system, there is a federal database as a method for resolving the system difference related to the schema difference. It is known as a technique for enabling search as a relational database according to one schema.

  In other words, it was introduced in Section 3.1 Methodology for Developing a Federated Database System (pp215-218) in `` Federated Database Systems for Managing Distributed, Heterogeneous, and Autonomous Databases '', ACM Computing surveys, Vol.22, No.3, September 1990. Multiple schemas can be shown to the user as if they were a single database. Note that in order to use schema integration, it is necessary to prepare information (data dictionary) for associating different schemas in each database with one schema and a function for converting based on the information.

By combining the two methods described above, system differences such as database configuration differences and schema differences can be absorbed, and the user can use a plurality of search devices in the same procedure.
For example, integration of search devices using schema integration will be described with reference to FIG. The figure shows that a search device a having a schema such as attributes “product name” and “type” by a schema integration and a search device b having a schema such as attributes “product name” and “type” have attributes “product” and “type”. The attribute “product name” of the search device a and the attribute “product name” of the search device b are integrated into the attribute “product”. Similarly, the attribute “type” of the search devices a and b is "Is integrated into" Type ".
Therefore, when the user performs a search, the search devices a and b can be searched by performing a search according to the integrated schema.

Even if a plurality of search devices can be integrated and the search procedure can be made the same, if each search device does not have a keyword expansion function etc., the above-mentioned problem regarding the data value still remains. Remains.
That is, a normal term (for example, “Relational Database Management System” or the like) is used as the value of the attribute “type” of the search device a, and an abbreviated term (for example, “RDBMS” is used as the value of the attribute “type” of the search device b. ”, Etc.), even if the corresponding product is searched by specifying“ type = RDBMS ”, the“ relational database management system ”in the search device a having the same meaning as“ RDBMS ” Is not detected because the data values do not match, and the user cannot obtain a product such as “Ox relational” corresponding to the “relational database management system”. Therefore, necessary data cannot be properly searched unless an accurate keyword is set for each search device.

  The invention described in Japanese Patent Application Laid-Open No. 6-35784 discloses a system that outputs arbitrary data as significant information to a plurality of terminals from databases having different data structures.

As described above, in the integrated search device including a plurality of search devices, as a method for appropriately searching for data required by the user, there is a method having a keyword expansion function for each search device. Requires a keyword expansion function for a plurality of search devices, which is costly.
As another method, there is a method of using controlled words in each search device, but in this method, a set of controlled words used for each search device may be different. There has been a problem that the user has to specify a controlled word set for each target search device as a keyword at the time of search.

  The present invention has been made in view of the above-described conventional circumstances, and can easily use the configuration of each search device without consideration of the cost without considering the user and appropriately search for the data required by the user. An object of the present invention is to provide an integrated search device and a keyword expansion method that can be used.

  In order to achieve the above object, an integrated search apparatus according to the present invention associates data and keywords with each other and registers them in a database, and includes an integrated search including a plurality of search apparatuses that search data based on specified keywords. In the apparatus, an input means for inputting a keyword for data to be registered, a keyword expansion means for expanding the input keyword into a related keyword, and a keyword based on the expanded keyword is associated with the data to a predetermined search apparatus. And a data registration control means for registration.

Here, expanding an input keyword into a related keyword is, for example,
A synonym of a keyword, a phrase that shares a concept with a keyword, a phrase closely related to a keyword, a procedure for converting an initial letter to uppercase or lowercase if a keyword is represented in alphabet, or If the keyword to be entered is in katakana notation, a word / phrase related to the input keyword is extracted based on a knowledge base (for example, a dictionary) holding knowledge such as a procedure for converting to another katakana notation that represents the same meaning. That means.
Also, to associate keywords based on expanded keywords with data,
This includes not only associating at least one of the expanded keywords with the data but also associating the input keyword with the data in addition to at least one of the expanded keywords.

  In the integrated search device described above, when a keyword is input by the input unit, the keyword expansion unit expands the input keyword into a related keyword, and the data registration control unit converts the keyword based on the expanded keyword into data. Corresponding information is registered in a predetermined search device.

  Therefore, when storing data in the search device, keywords related to the input keyword can be associated with the data, so that a plurality of keywords related to the input keyword can be associated with the data. Necessary data can be searched appropriately. In addition, even when registering data in a search device that uses controlled words, keywords that correspond to the controlled words can be obtained by expanding the input keywords, and the user can use the search device in the storage location. This saves you the trouble of checking the controlled words associated with the data.

  An integrated search device according to the present invention relates to data to be searched in an integrated search device including a plurality of search devices that search data based on a specified keyword by registering data and keywords in association with each other in a database. An input means for inputting a keyword, a keyword expansion means for expanding the input keyword into related keywords, and a data search control means for causing a predetermined search device to search for data based on the expanded keyword. It is characterized by that.

  Here, to search based on the expanded keyword is not only to search using at least one of the expanded keywords, but also to be input in addition to at least one of the expanded keywords. This also includes letting you search using other keywords.

  In the above integrated search device, when a keyword related to data to be searched is input by the input means, the keyword expansion means expands the input keyword into related keywords, and the data search control means based on the expanded keyword. A predetermined retrieval device is made to retrieve data.

  Therefore, when searching for a search device, because a search is performed using keywords related to the input keyword, it is possible to search using a plurality of keyword groups related to the input keyword, and necessary data Can be searched appropriately. Also, when searching for a search device that uses controlled words, keywords that correspond to the controlled words can be obtained by expanding the input keywords, and the controlled words of the target data for the user to search for can be obtained. Can be saved.

  An integrated search device according to the present invention relates to data to be searched in an integrated search device including a plurality of search devices that search data based on a specified keyword by registering data and keywords in association with each other in a database. An input means for inputting a keyword, a preliminary search control means for causing the search device to search for data based on the input keyword, and an extraction for extracting a keyword related to the keyword input from the search result according to a preset rule And a data search control means for causing a predetermined search device to search for data based on the extracted keyword.

  Here, extracting in accordance with a preset rule means, for example, extracting a keyword assigned to search result data, extracting a keyword in the same sentence as an input keyword in the search result data, Extract keywords that are close to the keyword entered in the search result data, extract keywords that have the same type of document structure as the keyword entered in the search result data, input in the search result data This means that keywords having a grammatical or semantic relationship such as a relationship between a keyword and a subject predicate, a modified modification relationship, a parallel relationship, and the like are extracted.

  The search based on the extracted keyword is not only a search using at least one of the extracted keywords, but also an input keyword in addition to at least one of the extracted keywords. It also includes making it search using.

  In the integrated search device described above, when a keyword related to data to be searched is input by the input means, the preliminary search control means causes the search device to search for data based on the input keyword, and the extraction means inputs from the search result. A keyword related to the extracted keyword is extracted according to a preset rule, and based on the extracted keyword, the data search control means causes a predetermined search device to search for data.

  Therefore, since it is possible to obtain a keyword related to the input keyword without having a knowledge base, the necessary data can be appropriately searched by performing a search based on the obtained keyword. . In addition, even when a knowledge base is not provided, a keyword corresponding to a controlled word can be obtained by obtaining a keyword related to the input keyword when searching for a search device using the controlled word. It is possible to save the trouble of checking the controlled word of the data.

  An integrated search device according to the present invention registers data and keywords in a database in association with each other, and in an integrated search device provided with a plurality of search devices that search for data based on a specified keyword. An input means for inputting a keyword, a preliminary search control means for causing the search device to search for data based on the input keyword, and an extraction for extracting a keyword related to the keyword input from the search result according to a preset rule Means and data registration control means for registering a keyword based on the extracted keyword in a predetermined search device in association with the data.

  Here, associating a keyword based on the extracted keyword with data means not only associating at least one of the extracted keywords with data, but also in addition to at least one of the extracted keywords. It also includes associating keywords with data.

  In the integrated search device described above, when a keyword is input by the input means, the preliminary search control means causes the search device to search for data based on the input keyword, and the extraction means adds the keyword input from the search result. Related keywords are extracted according to preset rules, and based on the extracted keywords, the data search control means causes a predetermined search device to search for data.

Therefore, even if a knowledge base is not provided, when data is stored in the search device, keywords related to the input keyword can be associated with the data, and data necessary for the search can be appropriately searched. Even if a knowledge base is not provided, when data is registered in a search device that uses controlled words, keywords corresponding to the controlled words can be obtained by expanding the input keywords, and the user can obtain data It is possible to save the trouble of checking the controlled word of the search device to be stored that is associated with.
Further, as is apparent from the description of each claim above, the present invention is also a method for expanding keywords used for database search into related keywords, and the keywords are expanded into related keywords by a search device having a keyword expansion function. A method of expanding a keyword for database search using the provided keyword for database search such as keyword registration or keyword search in the other search device by providing the expanded keyword to another search device It is.

  As described above, according to the integrated search device of the present invention, when data is registered in the search device, the keyword input from the user is expanded, and the keyword based on the expanded keyword is associated with the data. Since the registration is performed, a plurality of keywords related to the input keyword can be made to correspond to the data, and even if the search device does not have the function of expanding the keyword at the time of search, the keyword is expanded. It is possible to search for appropriate data similar to a search device having a function. In addition, even when data is registered in a search device that uses controlled words, the user does not need to be aware of the controlled words associated with the data, and the time and effort for examining the controlled words can be saved.

  Further, according to the integrated search device according to the present invention, when searching for data from the search device, the keyword input from the user is expanded, and the search is performed based on the expanded keyword. Even in a search device that does not have a function, it is possible to search for appropriate data in the same manner as a search device that has a keyword search function. In addition, even when data is searched from a search device using controlled words, the user can perform an appropriate search without being aware of the controlled words, and the user can check the controlled words of the search device to be searched. Can be saved.

  Furthermore, according to the integrated search device according to the present invention, the search device is preliminarily searched based on the keyword input from the user, and the keyword related to the keyword input by the user is extracted from the search result. Even without a knowledge base, keywords can be expanded, and the expanded keywords can be used for registering data in the search device or searching for data from the search device.

An integrated search apparatus according to a first embodiment of the present invention will be described with reference to FIGS.
This integrated search device includes search devices A and B, a registration mediation unit 1, a data registration unit 2, and a profile storage unit 3.

The search device A has a database that stores data and keywords related to the contents of the data. When registering data, the search device A registers the data and keywords related to the data in association with each other, while searching for data. In this case, the data associated with the specified keyword is searched and detected.
Further, the search device A has a function of expanding keywords based on a synonym table (dictionary) as shown in FIG. 2, and when receiving a keyword to be expanded, expands the keyword into a synonym. Output the expansion result.

  The search device B is a search device that handles controlled words as keywords, and has a database that stores data and controlled words related to the contents of the data. Are registered in association with each other, and when searching for data, the data associated with the specified controlled word is searched for and detected. The search device B has a set of controlled words as a controlled word table as shown in FIG.

The registration intermediary means 1 includes information for accessing the connected search devices A and B, and uses the connection procedure and access procedure specific to the search device to connect the data registration means 2 to each search device A and B. Mediate communication between them.
The data registration unit 2 includes an input unit 2a that inputs data to be registered, a keyword associated with the data, and a registration destination search unit, a search unit selection unit 2b that selects a search unit having a keyword expansion function, Keyword expansion means 2c for expanding keywords using the keyword expansion function, data registration control means 2d for registering the input data and the keywords passed from the keyword expansion means 2c in association with the search device to be stored ,have.

The input unit 2a has an input device such as a mouse and a keyboard, and accepts input from a user such as data to be registered, a keyword associated with the data, a search device for a storage destination, and the like. In the present embodiment, when a storage destination search device is input, the search device registered in the profile storage means 3 can be selected.
Here, the input of data to be registered includes not only inputting data directly from the input device but also inputting a name for identifying data from the input device and inputting the data body from the storage device. . Further, the keyword input from the user is not limited to the keyword directly input by the user, but may be a keyword extracted from the data input by the user by an independent word dictionary or the like.

  When expanding the keyword, the search device selection unit 2b selects a search device having a keyword expansion function from the profile stored in the profile storage unit 3, and notifies the keyword expansion unit 2c. Note that when a plurality of search devices have a keyword expansion function, they can be arbitrarily selected.

  The keyword expansion unit 2c acquires from the profile storage unit 3 a command for using the keyword expansion function of the search device notified from the search device selection unit 2b. Then, keyword expansion is performed by passing the input keyword together with the acquired command to the search device, and the expanded keyword group is acquired from the search device and transferred to the data registration control means 2d.

The data registration control unit 2d registers the keyword group received from the keyword expansion unit 2c in the registration target search apparatus in association with the input data.
Note that it is detected from the profile storage means 3 whether or not the search device to be registered is a search device that handles controlled words. In the case of a search device that handles controlled words, it is received from the keyword expansion means 2c. A keyword that is a controlled word is extracted from the keyword group based on the controlled word table possessed by the search device, and the extracted keyword is associated with the input data and registered in the search device to be registered. You can also.

The profile storage unit 3 includes a storage device such as a magnetic disk, and stores the characteristics of the search devices A and B in the storage device as a searchable profile. As shown in FIG. 3, this profile includes, for each search device, presence / absence of a keyword expansion function, function call method information such as a command when using the keyword expansion function when the keyword expansion function is provided, It has information on whether or not the search device uses controlled words. The profile information is set in advance by the administrator of the integrated search apparatus.
Here, the keyword expansion means described in claim 1 is configured by the above-described keyword expansion means 2c and the keyword expansion function of the search device.

Next, the operation of the integrated search device according to the first embodiment will be described with reference to FIG.
First, the input means 2a accepts input from the user such as data to be stored in the retrieval device and keywords associated with the data (steps S1 and S2), and further accepts selection input to the retrieval device at the storage destination (step S3) and selects Profile information relating to the retrieved search device is obtained from the profile storage means 3 (step S4), and it is detected whether or not the search device at the storage destination has a keyword expansion function (step S5).

  If the search device at the storage destination has a keyword expansion function, even if keyword expansion is performed at the time of registration, there is little effect obtained at the time of search, so keyword expansion is not performed, and data registration control means The keyword 2d input by the user is associated with the data and registered in the search device of the storage destination (step S6).

On the other hand, if the registration destination search device does not have a keyword search function, a keyword expansion process described later is performed (step S7).
In this keyword expansion process, the search device selection means 2b acquires profile information related to keyword expansion of all search devices from the profile storage means 3, and selects a search device having a keyword expansion function from the profile information. When a search device having a keyword expansion function is selected, an access method for using the keyword expansion function of the search device selected by the keyword expansion unit 2c is acquired from the profile storage unit 3, and according to the acquired access method The keyword is expanded by passing the input keyword to the search device, and the expanded keyword group is acquired from the search device and transferred to the data registration control means 2d.

When the keyword expansion process is finished, the keyword expansion means 2c detects whether or not there is another input keyword (step S8). If it exists, the keyword expansion process (step S7) is performed. Is similarly performed for the keyword.
When the keyword expansion process is completed for all keywords input by the user, the data registration control unit 2d registers the keyword group expanded by the keyword expansion process in the search device in association with the data. Control is performed (step S6).

Here, as an example of the above-described operation, a case where data is registered in the search apparatus B will be described with reference to FIG. Note that dotted squares in each search device represent stored data, and words in the data represent keywords associated with the data.
First, in addition to inputting data to be stored, when the user inputs the keyword “document”, the profile information is referred to, and the keyword expansion function of the search apparatus A is used. Is expanded to the keywords “document”, “document”, and “document”. As a result, the storage target search device B selected by the user stores the data to be stored in association with the keywords based on the keywords “document”, “document”, and “document”.

Here, since the search device B is a search device that uses controlled words, the keywords “document” and “document” that are not controlled words are not registered in the data, and the keywords “document” are stored in association with each other. .
In the above-described example, when the storage destination search device does not use the controlled word, a plurality of related keywords such as “document”, “document”, and “document” are associated with the data.

In this way, since a plurality of keywords based on the keywords developed by the user can be registered in correspondence with the data, the user can search without searching for new keywords. Can properly search for the data required by.
Also, if the search device to be registered uses controlled words, even if the user inputs a keyword different from the controlled word, the input keyword is expanded so that the related controlled word is registered in the search device. Therefore, the user does not need to input a keyword while paying attention to the controlled word associated with the data, and can save the trouble of checking the controlled word.
When integrating existing search devices, the above-described effects can be obtained if any one of the plurality of search devices has a keyword expansion function.
In the search device selection means 2b described above, a search device for performing keyword expansion is arbitrarily determined. For example, the number of vocabularies that the keyword expansion function for each search device has in advance or the speed of expansion processing A suitable search device may be selected based on information such as.

An integrated search apparatus according to the second embodiment of the present invention will be described with reference to FIGS.
This integrated search device includes search devices C and D, search mediating means 4, data search means 5, profile storage means 6, and search result merging means 7.

The search devices C and D have a database that stores data and keywords related to the contents of the data. When registering data, the search devices C and D search the data while registering the data and keywords related to the data in association with each other. In doing so, data associated with the specified keyword is searched and detected. Note that the search devices C and D have different search expression formats and search result formats when searching for data.
Further, the search device C has a function of expanding keywords based on a synonym table (dictionary) as shown in FIG. 6, and when receiving a keyword to be expanded, expands the keyword into a synonym. Output the expansion result.

  The search mediating means 4 includes information for accessing the connected search devices C and D, and the data search means 5 and the search devices C and D are connected using the connection procedure and access procedure unique to the search device. Mediate communications.

  The data search means 5 includes an input means 5a for inputting a keyword related to data to be searched as a search expression, a search expression interpretation means 5b for interpreting the input search expression to detect a keyword, and a search device having a keyword expansion function. The search device selection means 5c to be selected, the keyword expansion means 5d to expand the keyword using the keyword expansion function of the search device, the search expression generation means 5e to generate the search expression based on the expanded keyword, Data search control means 5f for allowing the search device to be searched to search for data using the search formula.

The input unit 5a has an input device such as a mouse and a keyboard, and accepts a search expression including a keyword related to data to be searched and an input from a user such as a storage destination search device. In this embodiment, when inputting a search destination search device, the search device registered in the profile storage means 6 can be selected, and a plurality of search devices can also be selected.
The format of the input search formula is unified regardless of the search target search device, such as a simple search formula or a search formula related to a search device that the user is familiar with. It is a format.

  The search expression interpreter 5b detects the keyword by interpreting the search expression designated by the user based on the search expression format information preset in the profile storage means 6. When interpreting a search expression, information on the search expression specified by the user such as the position of the keyword in the search expression, the keyword combination method, the logical expression based on the keyword, etc. is grasped. The expression generation unit 5e is notified.

  When performing keyword expansion, the search device selection unit 5c selects a search device having a keyword expansion function from the profile information in the profile storage unit 3, and notifies the keyword expansion unit 5d of the search device. Note that when a plurality of search devices have a keyword expansion function, they can be arbitrarily selected.

The keyword expansion unit 5d acquires from the profile storage unit 6 a command for using the keyword expansion function of the search device notified from the search device selection unit 5c. Then, along with the acquired command, the keyword detected by the search expression interpretation means 5b is expanded to the keyword by passing it to the search apparatus, and the expanded keyword group is acquired from the search apparatus and transferred to the search expression generation means 5e.
For example, when the keyword “document” is expanded using the keyword expansion function (synonym expansion function) of the search device C, a command “find document” is generated and passed to the search device C.

The search formula generation unit 5e creates a search formula suitable for the search target search apparatus by using the expanded keyword group, information on the format of the search formula in the profile storage unit 6, and the search formula input by the user.
For example, when the search expression input by the user is the keyword “document” and the expanded keyword group is “document”, “document”, “document”, a search expression suitable for the search device D is created. A search expression representing a logical sum of “document or document or document” is created by referring to a keyword operator that is information relating to the format from the profile information.

  The data search control unit 5f uses the search formula created by the search formula generation unit 5e to cause the search target search device to search for data corresponding to the search formula.

  The profile storage means 6 includes a storage device such as a magnetic disk, and stores the characteristics of the search devices C and D as a searchable profile in the storage device. As shown in FIG. 7, this profile includes, for each search device, presence / absence of a keyword expansion function, and function call method information such as a command when using the keyword expansion function when having a keyword expansion function, It has information on the format of the search expression and information on the output format of the search result.

Information related to the search format includes information such as the type and meaning of the search formula, the position of the keyword in the search formula, and the like, and is used for interpretation and conversion of the search formula.
For example, when a search expression of a certain search device expresses keywords side by side and has a meaning of searching for data including any of the specified keywords, the search expression means a logical sum of the keywords. Therefore, a keyword delimiter is set as an operator of the logical sum of information related to the search format. In addition, when a certain search device can specify a logical sum or logical product of keywords as a search expression, a keyword expression using a data expression such as a tree structure is used as a sub-expression representing the logical sum or logical product of keywords. The combination method with the operator is set. Note that the profile information of the present embodiment is set when the logical sum or logical product format of the retrieval formula for each retrieval device is the same, and the logical sum operator and logical product operator are set. ing.

The information related to the output format of the search result is a pattern such as a conversion command string, a conversion script, a regular expression, etc., for example, an identifier or name representing data in the search result, or a keyword related to the data There is information, etc., which is used to interpret search results.
The profile information is set in advance by the administrator of the integrated search apparatus.

When the user designates a search for a plurality of search devices, the search result merging unit 7 receives each search result from each search device and, based on the profile information in the profile storage unit 6, for each search device. Interpret search results with different formats and combine them into a unified search result. In this embodiment, when a search result is received from each search device, an identifier representing data is extracted from the search results of each search device based on the search result extraction pattern in the profile information of each search device and collected. Yes.
Here, the keyword expansion means according to claim 2 is configured by the above-described keyword expansion means 5d and the keyword expansion function of the search device.

Next, the operation of the integrated search device according to the second embodiment will be described with reference to FIG.
First, when the input unit 5a receives an input of a search expression including a keyword related to desired data from the user (step S9), the input unit 5a passes the search expression to the search expression interpretation unit 5b, and the search interpretation unit 5b receives the search expression. Are detected and keywords are detected (steps S10 and S11).

Next, the search device selection unit 5c selects a search device for performing keyword expansion based on the profile information in the profile storage unit 6 (step S12).
When a search device for keyword expansion is selected, an access method for using the keyword expansion function of the selected search device is acquired from the profile storage unit 6 by the keyword expansion means 5d, and the search is performed according to the acquired access method. The keyword detected by the expression interpreter 5b is expanded to a keyword by passing it to the search device, and the expanded keyword group is acquired from the search device and transferred to the search expression generator 5e (step S13).
For example, when a predetermined number of keywords cannot be obtained by keyword expansion, the keyword expansion is insufficient, and thus step S13 described above is repeated again for a search device having another keyword expansion function. (Step S14).

  When the keyword group is received from the keyword expansion means 5d, the search expression generating means 5e searches the keyword group, the search expression input by the user, and the search expression of the search device in the profile information of the profile storage means 6. Based on the information related to the format, a search formula generation process, which will be described later, is performed to generate a search formula for each search device to be searched (step S15).

  When the search formula is generated, the data search control means 5f causes each search device to search for the corresponding data using the search formula for each search device via the search mediation means 4 (step S16). Then, the search result merging means 7 receives the search results from the respective search devices, refers to the profile information in the profile storage means 6, identifies the searched data in the search results, and searches all search target search devices. The data retrieved by is collected (step S17).

Next, search expression generation processing by the search expression generation means 5e will be described with reference to FIG.
The search formula generation unit 5e receives information such as the configuration of the search formula specified by the user from the search formula interpretation unit 5b (step S18), and information such as the configuration and format of the search formula of the search target search device from the profile storage unit 6. Is acquired (step S19).
Then, it is detected whether or not the search expression in the search device to be searched can express the logical sum of keywords (step S20).

  If the logical sum can be expressed, the keyword group passed from the keyword expansion means 5d can be collected as one check expression. Therefore, first, the keyword group is summarized by logical sum (step S21), and the logical sum is summarized. The retrieved keyword group is replaced with the keyword part of the search expression input by the user (step S22), and the replaced search expression is searched for the search target search apparatus based on the acquired search formula information of the search target search apparatus. Conversion into a search expression (step S23).

On the other hand, when the search expression of the target search device cannot express the logical sum of keywords, the keyword group cannot be combined into one search expression, so the following processing is performed for each keyword in the keyword group.
That is, each keyword in the keyword group is replaced with the keyword part of the search expression input by the user (step S24), and the replaced search expression is based on the acquired information on the format of the search expression in the search target search device. Is converted into a search expression for the search device to be searched (step S25). The above processing (steps S24 and S25) is repeated for the remaining keywords in the same manner (step S26).

Here, in the integrated search device according to the second embodiment, a case where data is searched from the search devices C and D will be described as an example with reference to FIG. Note that dotted squares in the search device represent stored data, and words in the data represent keywords associated with the data.
First, when a search expression including the keyword “document” is input by the user, the keywords “document”, “document”, and “document” are acquired by keyword expansion, and each search expression using the keyword group as a logical sum is searched. It converts into the search formula of the format for every apparatus, searches each search apparatus using the said search formula, and integrates a search result.
Therefore, even if the keyword input by the user is “document”, the search is performed by the search expression including the expanded keyword group such as “document”, “document”, and “document”, which is conventionally detected from the search device D. Data associated with keywords such as “document” and “document” that could not be detected can also be detected.

Thus, even a search device that does not have a keyword expansion function can perform the same search as a search device that has a keyword expansion function. Further, even when a plurality of search devices are searched simultaneously, data can be easily searched without distinguishing the features of each search device. In addition, even when data is searched from a search device using controlled words, the user can perform an appropriate search without being aware of the controlled words, and the user can check the controlled words of the search device to be searched. Can be saved.
When integrating existing search devices, the above-described effects can be obtained if any one of the plurality of search devices has a keyword expansion function.

In the above-described search device selection means 5c, the search device for performing keyword expansion is arbitrarily determined. For example, the number of vocabularies or expansion processing that the keyword expansion function for each search device has in advance is provided. A suitable search device may be selected based on information such as the speed of the user.
In the first embodiment and the second embodiment described above, the keyword expansion function of the search device is used. However, the keyword expansion function is provided as a separate configuration from the search device. In this case, all the search devices may not have the keyword expansion function.

Next, an integrated search device according to a third embodiment of the present invention will be described with reference to FIG. In addition, the same number is attached | subjected to the part similar to above-mentioned 2nd Example, and the overlapping description is abbreviate | omitted.
This integrated search device includes search devices E and F, search mediation means 4, data search means 8, profile storage means 9, and search result merging means 7.

  The search devices E and F have a database that stores data and keywords related to the contents of the data. When registering data, the search devices E and F search the data while registering the data and keywords related to the data in association with each other. In doing so, data associated with the specified keyword is searched and detected. The search devices E and F differ in the format of the search expression and the format of the search result when searching for data.

  The data search means 8 includes an input means 5a for inputting a keyword related to data to be searched as a search expression, a search expression interpretation means 5b for interpreting the input search expression to detect a keyword, and information necessary for keyword expansion. Search device selection means 8a for selecting a search device to be used for acquisition, keyword expansion means 8b for expanding a keyword based on the information of the search device, and search expression generation means for generating a search expression based on the keyword 5e, and data search control means 5f for causing the search device to be searched to search for data using a search expression.

  The search device selection means 8a selects a search device that can acquire a keyword group related to the data from the searched data based on the profile of the profile storage means 9, for example, a search device that does not use a controlled word, The selected search device is notified to the keyword expansion means 8b. If a plurality of search devices can be selected, any search device can be selected.

The keyword expansion means 8b includes a preliminary search control means 8c for acquiring data related to the keyword in the search formula input by the user from the search device selected by the search device selection means 8a, and a preliminary search. And a keyword selection unit 8d as an extraction unit that performs keyword expansion by extracting a keyword related to the keyword from the data acquired by the control unit 8c.
The preliminary search control means 8c passes the keyword detected by the search expression interpretation means 5b to the search expression generation means 5e, and is associated with the keyword in the search device selected by the search expression selection means 8a to the search expression generation means 5e. A search expression for searching the received data is generated, the search apparatus is used to search using the search expression, and the search result is received. Next, the received search result is passed to the keyword selection means 8d.

  The keyword selection unit 8d extracts a keyword candidate group from the received search result according to the keyword extraction pattern of the profile information in the profile storage unit 9. Then, a keyword that is closely related to the keyword (keyword input by the user) detected by the search expression detection unit 5b from the extracted keyword candidate group is extracted as a keyword.

Here, as a criterion for determining whether or not the keyword is deeply related, for example, the relationship between the grammar, the face, the position, etc. in the search result for the keyword inputted by the user, the keyword in the search result There are the appearance frequency, the relationship between the ratio of keywords in the entire search device and the ratio of keywords in the search results.
When using the relationship between the ratio of keywords in the entire search device and the ratio of keywords in the search results as a reference, the lower the ratio included in the entire search device, the higher the ratio included in the search results, the more related Is treated as a deep keyword.

  The profile storage means 9 includes a storage device such as a magnetic disk, and stores the characteristics of the search devices E and F as a searchable profile in the storage device. As shown in FIG. 11, this profile includes information about whether or not a keyword group related to data can be acquired for each search device, information about a search expression format, information about an output format of a search result, Information on keyword information of the search result. The profile information is set in advance by the administrator of the integrated search apparatus.

Next, the operation of the integrated search device according to the third embodiment will be described with reference to the drawings.
Here, the operation until the keyword is expanded, which is different from the second embodiment described above, will be described with reference to FIG.
First, when the data input means 5a receives an input of a search expression including a keyword related to desired data from the user, the input means 5a passes the search expression to the search expression interpretation means 5b, and the search interpretation means 5b interprets the search expression. To detect keywords.
Based on the profile information stored in the profile storage unit 9, a search device as a preliminary search target that can acquire a keyword group related to the data from the data searched by the search device selection unit 8 a is selected.

On the other hand, the preliminary search control unit 8c passes the keyword detected by the search formula interpretation unit 5b to the search formula generation unit 5e, and the search formula generation unit 5e to which the keyword is passed is information on the search formula in the search device targeted for preliminary search. Is acquired from the profile storage means 9 (step S27), and a search expression is generated from the keyword based on the information related to the search expression (step S28).
Next, the preliminary search control means 8c uses the generated search formula to cause the search device that is the target of the preliminary search to perform a search, obtains the search result, and passes it to the keyword selection means 8d (step S29).
When the search result is received, the keyword selecting unit 8d interprets it according to the keyword extraction pattern of the profile information in the profile storage unit 9 (step S30), and extracts a keyword candidate group (step S31). Next, a keyword group that is closely related to the keyword input by the user is extracted from the extracted keyword candidate group (step S33).
As a result, the keywords included in the search formula by the user can be expanded into related keyword groups. The subsequent processing is the same as the operation after keyword expansion in the second embodiment.

  Here, a case where a search is performed from the search devices E and F will be described with reference to FIG. Note that dotted squares in the search device represent stored data, and words in the data represent keywords associated with the data.

  First, when the user designates the keyword “document” in the search formula, the search formula is analyzed, the keyword “document” is detected, and the search device E is selected as the search device for the preliminary search based on the profile information. Is done. Then, the data of the search device E is searched using the keyword “document”, and data associated with keywords such as “document, document”, “document, management”, “document, document”, etc. is obtained as a search result. It is done. Next, based on the profile information, keyword candidates such as “document”, “document”, and “management” associated with the obtained data are extracted, and “document”, “ Document "is extracted. In this embodiment, since the keyword candidate having the appearance count of 1 is not used as a keyword when performing a search, “management” is not extracted as a keyword.

  Then, a search expression including the extracted keyword group for the search devices E and F to be searched is generated based on the profile information. For example, “document | document” is generated as a search expression for the search apparatus E, and “document or document” is generated as a search expression for the search apparatus F. Next, each search device is searched using the generated search formula, and search results such as data identifiers, names, and the like are extracted from the search results of each search device based on profile information. The results are collected in a unified format and output to an output device such as a display device.

  By performing the above-described search operation, it is possible to obtain data that could not be obtained when the search device F was searched by specifying “document” as a search formula. That is, data associated with a keyword such as “document” closely related to “document” can be retrieved and obtained.

In this way, keywords can be expanded into related keyword groups without having a knowledge base storing keywords related to the keywords, and the user's desired data can be obtained by searching using the keyword groups. The frequency of detection increases, and the search can be performed efficiently.
In addition, even when data is retrieved from a search device that uses controlled words, the user can perform an appropriate search without being aware of the controlled words, and the user can select the controlled words in the search device to be searched. You can save time and effort.

  In the third embodiment described above, when the search formula designated by the user is a logical product of keywords, the search result can be further examined by creating the search formula by the following procedure. That is, when the keywords specified by the user are k1, k2,..., Kn, first, the keyword group expanded from the keyword ki (0 ≦ i ≦ n) is defined as Ki (not including ki), and each of Ki An expression Oi in which keywords are combined by logical sum is created, and a partial search expression of logical product of the expression Oi and a logical product expression Ai obtained by removing the keyword ki from the search expression input by the user is created. Next, the search formula input by the user and the n partial search formulas for each keyword are combined with a logical sum to obtain a search formula. As a result, the keyword group expanded from each keyword input by the user is combined with a logical sum for each input keyword, and the combined expression is combined into a search expression that is combined with the logical product. It is possible to reduce the situation where undesired data is retrieved.

Next, an integrated search device according to a fourth embodiment of the present invention will be described with reference to FIG. In addition, the same number is attached | subjected to the part similar to the above-mentioned Example, and the overlapping description is abbreviate | omitted.
This integrated search device is an integration that performs an operation that combines the expansion of keywords by preliminary search in the integrated search device of the third embodiment and the operation in data registration in the integrated search device of the first embodiment. The search device includes search devices G and H, a registration mediation unit 1, a search mediation unit 4, a data registration unit 10, and a profile storage unit 11.

  The search devices G and H have a database that stores data and keywords related to the contents of the data. When registering data, the search devices G and H search the data while registering the data and keywords related to the data in association with each other. In doing so, data associated with the specified keyword is searched and detected. Note that the search devices G and H do not have a keyword expansion function.

The data registration unit 10 includes an input unit 2a, a search expression generation unit 5e, a keyword expansion unit 8b, and a data registration control unit 2d.
The keyword expansion means 8b has preliminary search control means 8c and keyword selection means 8d.
The profile storage means 11 relates to the profile information of the profile storage means 3 in the first embodiment, information on whether or not a keyword group related to data can be acquired, information on the format of the search expression, and keyword information on the search result. Information.

  In this integrated search device, a keyword expansion operation using the preliminary search shown in the third embodiment is performed on a user keyword input by the input means 2a, and a keyword group obtained by the expansion operation is obtained. By performing the data registration operation shown in the first embodiment, the data is registered in the search target storage device in association with the data.

  Therefore, keywords can be expanded into related keyword groups without having a knowledge base storing keywords related to the keywords, and keywords based on the keyword groups can be registered corresponding to the data. When searching, data required by the user can be appropriately searched without newly expanding keywords.

  In addition, when the search device to be registered uses a controlled word, even if the user inputs a keyword different from the controlled word, the input keyword is expanded so that the related controlled word is registered in the search device. Therefore, the user does not need to input a keyword in consideration of the controlled word to be associated with the data, and can save the trouble of checking the controlled word.

In the first to fourth embodiments described above, the integrated search device includes two search devices. However, the integrated search device is not limited to this, and may include two or more search devices.
In the present embodiment, each functional unit described above is configured by a processor executing a control program (stored in advance in a ROM or the like). However, in the present invention, these functional units are configured as independent hardware circuits. May be.

It is a block diagram of the integrated search device which concerns on 1st Example of this invention. It is a figure explaining each search device of the integrated search device concerning the 1st example of the present invention. It is the figure which showed the profile information which concerns on 1st Example of this invention. It is a figure explaining operation | movement of the integrated search apparatus based on 1st Example of this invention. It is a block diagram of the integrated search device which concerns on 2nd Example of this invention. It is a figure explaining each search device of the integrated search device concerning the 2nd example of the present invention. It is the figure which showed the profile information which concerns on 2nd Example of this invention. It is a figure explaining operation | movement of the integrated search apparatus which concerns on 2nd Example of this invention. It is a figure explaining the search formula creation process of the integrated search device which concerns on 2nd Example of this invention. It is a block diagram of the integrated search device which concerns on 3rd Example of this invention. It is a figure explaining each search device of the integrated search device concerning the 3rd example of the present invention. It is the figure which showed the profile information which concerns on 3rd Example of this invention. It is a figure explaining operation | movement of the integrated search apparatus based on 3rd Example of this invention. It is a block diagram of the integrated search device which concerns on 4th Example of this invention. It is a figure explaining integration of the search device concerning a conventional example.

Explanation of symbols

A, B, C, D, E, F, G, H ...
2a, 5a, input means,
2c, 5d, keyword expansion means,
2d .. data registration control means,
5f .. data search control means,
8c .. Preliminary search control means,
8d ・ Keyword selection means

Claims (3)

In an integrated search device that associates data with a keyword and registers them in a database, controls a plurality of search devices, and searches for data based on a specified keyword.
An input means for inputting a keyword related to data to be searched;
Keyword expansion means for expanding the input keywords into related keywords,
An integrated search device, comprising: a data search control means for causing a predetermined search device to search for data based on the developed keyword. The integrated search device according to claim 1,
Preliminary search control means for causing the search device to search for data based on the keyword input from the input means;
Extraction means for extracting a keyword related to the keyword input from the search result according to a preset rule;
The data search control unit causes a predetermined search device to search for data based on the extracted keyword. A method of expanding keywords used for database search into related keywords,
The profile management means manages information about whether or not it has a required keyword expansion function for expanding keywords into related keywords for a plurality of search devices,
The search device selection means expands the keywords into related keywords by the search device having a keyword expansion function based on the management information of the profile management means,
A database search keyword expansion method, wherein the data search control means provides the expanded keyword for database search to another search device that does not have the required keyword expansion function.

Images