JP2000222418A - Method and device for retrieving data base - Google Patents

Abstract

PROBLEM TO BE SOLVED: To execute retrieval considering the priority of respective data bases in the case of retrieving plural data bases and to display and report a list of retrieved results to a user so that the features of retrieved results can easily be understood. SOLUTION: The data base retrieving device is provided with two or more data bases 1, a retrieving object data base input device 2 for specifying a data base to be retrieved, a priority information input device 3 for setting up priority at the time of retrieving each specified data based to be retrieved, and a retrieving request input device 4 for inputting information indicating a retrieving request. The device is also provided with a data base retrieving device 5 for applying retrieval corresponding to priority to the data base to be retrieved and extracting information having high relationship with the inputted retrieving request, a data base restricting device 13 for restricting a data base to be used for retrieval and a data display device 6 for displaying extracted information.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to computer networks,
With the development of the Internet, the spread of groupware, and the like, it has become an important technical issue how to extract necessary information from a large amount of information accumulated on a computer.

[0002] The present invention relates to a method for retrieving information stored in a database on a computer, and a database retrieval apparatus for realizing the method.

[0003]

2. Description of the Related Art Conventionally, as a method of searching for information stored in a computer, a word indicating a characteristic of the information is registered in advance as an index in each information, and a word input as a search request is used as an index. The method of extracting documents and the characteristics of each information are defined by the space vector composed of the type and appearance frequency of the word included in each information, and these vectors are compared with the space vector composed of search requests However, there is a method of extracting a vector having a similar vector direction.

[0004] In addition, it is difficult to specify only a result that matches 100% of a user's search request by the above search method. In many cases, the information group extracted by the above search method is listed as a search result candidate. In many cases, and let the user make the final selection.

[0005]

As the size of a database to be searched increases, it is expected that the number of searches performed on a plurality of databases will increase. At this time, if all the databases are searched using the same criteria, the search results include a lot of noise, and the risk that desired information cannot be obtained increases. Also, when displaying the search results in a list, it is desirable to take a display form in consideration of the fact that the target is a plurality of databases.

An object of the present invention is to provide a method and an apparatus for performing a search in consideration of the priority of each database when performing a search for a plurality of databases, and a program for causing a computer to execute the method. It is to provide a storage medium for recording.

Another object of the present invention is to provide a display method and means capable of transmitting the characteristics of a search result to a user in a more understandable manner when displaying a list of search results.

[0008]

According to one embodiment of the present invention, there is provided a database system comprising two or more databases for recording information to be searched and a search for inputting information indicating a search request. Request input means, and a database search means for comparing the information of the database and the information of the search request to extract information highly relevant to the search request;
A database search device having a data display means for displaying the extracted information, a database search method using the device, or a storage medium storing a program for causing a computer to execute the method; A database input means for specifying a database to be searched, and database priority information input means for setting a search priority for each of the databases specified as the search object, wherein the database In the search processing in the search means, a search according to the priority is performed on the database specified as the search target.

[0009] In the invention of the above embodiment, further, as a result of the search processing, an extracted data list display means for displaying a list of information having a relevance of not less than a specific value to the search request; From the list displayed on the list display means, a display data designating means for selecting information for displaying more detailed contents, and the contents selected by the display data designating means and the contents of the search request as search history data. A configuration including a search history storage unit that records in a search history database, and a first priority correction process that corrects information on the type of the database to be searched and the priority of the database by using the search history data. It may be.

According to another embodiment of the present invention, there is provided a database for recording information to be searched, a search request input unit for inputting information indicating a search request, and A database search device comprising: a database search unit that compares information and information of the search request to extract information highly relevant to the search request; and a data display unit that displays the extracted information, or a database search using the same. Search request priority information input means for setting a search priority for each word constituting the input search request in a storage medium storing a method or a program for causing the computer to execute the method; As a result, an extracted data list display method for displaying a list of information having a relevance equal to or more than a specific value with the search request. And display data specifying means for selecting information for displaying more detailed contents from the list displayed on the extracted data list display means, and the contents selected by the display data specifying means and the contents of the search request. Search history storage means for recording the search request as search history data in a search history database, and using the search history data to correct the information on the search request and the priority of words constituting the search request. Correction means.

Further, in the above invention, means for interactively changing the priority of a word constituting a search target database or a search request may be used.

In the above invention, each database may be classified into several categories in advance, and a means for specifying a database to be searched by selecting the category may be used.

In the above invention, a means may be used in which priority information is defined in advance for databases included in each category, and the priority value of the search target database is determined using the information. .

In the above invention, when designating a database to be searched, first, one database to be the center of the search is selected, and then a database other than the above database is selected by using a related database. Is also good.

In the above invention, a means for dividing the display screen of the extracted data list for each search target database and displaying each extracted data in the area of the search target database to which the data belongs may be used.

In the above invention, when displaying a list of extracted data, a display form of each extracted data, for example, display color, transparency of display, display, etc., according to the degree of relevance between the search request and each extracted data. Alternatively, a means of changing the size may be used.

In the above invention, when displaying a list of extracted data, the degree of relevance between the extracted data and each search target database is determined for each extracted data. Means may be used in which a radar chart centered on the axis or a bar graph color-coded for each search target database is displayed.

According to the present invention, for example, the following effects can be obtained.

(A) Since a database to be searched can be limited according to a search request, a search can be performed in consideration of the priority of each database.

(B) Since a search target database that reflects the user's request can be limited, a search can be performed in consideration of the priority of each database.

(C) Since a search request reflecting the user's request can be input, a search can be performed in consideration of the priority of each database.

(D) Since it is possible to set the priority directly reflecting the user's intention, it is possible to perform a search in consideration of the priority of each database.

(E) Since a search target database is simply set according to the purpose of the user, a search can be performed in consideration of the priority of each database.

(F) Since it is possible to more appropriately determine the intention of the user when selecting the database to be searched, it is possible to perform a search in consideration of the priority of each database.

(G) Since it is possible to recognize at a glance which database contains the extracted information, the features of the search result can be conveyed to the user more easily.

(H) Since the degree of relevance between the search request and the extracted information can be more accurately recognized, the features of the search result can be more easily conveyed to the user.

(I) Since the relationship between the extracted information and the search target database can be more accurately recognized, the characteristics of the search result can be more easily conveyed to the user.

[0028]

1 to 3 show an embodiment of the present invention.
6 will be described.

FIG. 1 shows the basic configuration of an embodiment of the database search system according to the present invention. This system basically includes a plurality of databases (1... 1 ′), a search target database input device (2), a priority information input device (3), a search request input device (4), a database search device (5), It consists of six devices, a data display device (6) and a database limiting device (13).

The present invention relates to a search for information recorded on a computer. Information to be searched is recorded in a database (1... 1 ′). It is assumed that a plurality of databases (1... 1 ′) exist according to the type of information to be recorded.

When performing a search, the user first selects one or more databases to be searched using the search target database input device (2). Next, using the priority information input device (3), information on the priority at the time of performing a search for each selected database is set. Further, information of the database (1) selected as a search target is extracted using the database limiting device (13). Next, using the search request input device (4), words or sentences related to the information to be searched are input as a search request. Using the above information, the system uses the database search device (5) to extract information highly relevant to the search request input by the user from the specified database and display it on the data display device (6).

FIG. 2 shows an example of a configuration when the database search method according to the present invention is systemized on a computer.

Database to be searched (1 ... 1 ')
Is a hard disk (21) or network (2
It is recorded on the storage device connected in 7). Further, a program for implementing the search method according to the present invention on a computer is also stored on the hard disk (21). And
The program is temporarily stored on the memory (22).

Next, the processing of the above program is executed by the CPU (2
Execute in 3). At this time, the display of the processing contents to the user is performed on the display (24), and the input of the command from the user is performed using the keyboard (25) or the mouse (26). When executing the search process, part or all of the database (1... 1 ′) to be searched is temporarily stored in the memory (22), and the search result is stored in the display (24). Is displayed.

Hereinafter, an embodiment of the present invention will be described in detail using a system having the configuration shown in FIG. 3 as a specific example of the present invention.

The information to be searched is a database (1)
Will be recorded. Then, in order to increase the efficiency of the search, the information recorded in the above-mentioned database (1) is converted into a more easily searchable format using the search database creator (7), and these are converted into the search database ( Record in 8). At this time, there are a plurality of search database creation devices (7... 7 ') according to the type of information recorded in the databases (1... 1'), and the search database creation devices (7... 7 ') according to the number of databases (1). There are also a plurality of databases (8... 8 ').

When performing a search, the user first uses the search request input device (4) to input a word or a sentence related to the information to be searched as a search request. Next, priority information for each search request is set using the search request priority information input device (9). Then, one or more databases to be searched are selected by using the search target database input device (2), and then, when the selected databases are searched by using the priority information input device (3). Set the priority information. In addition, a database limiting device (13)
Is used to extract information of the database (1) to be searched.

Using the above information, the system uses the database search device (5) to extract, from the specified database, some information highly relevant to the search request entered by the user, and to extract the extracted information and the information. Information on the relevance to the search request is displayed in a list on the extracted data list display device (10). Then, the user sees the list display and specifies necessary information with the display data specifying device (11). The system displays the specified information on the data display device (6) based on the specified result.

The search history database (12) uses the search request entered by the user, the information of the designated database, the priority information for each of them, and the information about which information the user finally displayed as a search history. To record. Then, by using the information of the search history database, the information recorded in the search database (8) is corrected, the search request input device (4) automatically adds the search request, and the search request priority information input device (9). ), The automatic addition of the search target database in the search target database input device (2), the correction of the priority information set in the database priority information input device (3), and the like.

FIG. 4 shows an example of information recorded in the database (1).

Each information is composed of the following elements. No.
(401) is a sequential number indicating the ID of each information in the database (1). Title (402)
Is a sentence indicating the outline of the content of each information. Creator (40
3) is the name of the creator of each information. Affiliation (404) is the department to which the creator of each information belongs. The creation date (405) is the date when each information was first registered. Update date (40
6) is the newest date when each information was corrected. The text (407) is the content of each information.

FIG. 5 shows a processing flow of the search database creating device (7) for the information recorded in the database (1).

First, data is extracted from the database (1) for each information unit (step 501). Thereafter, words included in the words recorded in the data are extracted by using a method such as morphological analysis (Step 502), and the extracted words are recorded as a set of words in the extracted word database 503a (Step 503). . Then, the above processing is executed for all information included in the database (1) (step 504).

Thereafter, for all the words recorded in the extracted word database in step 503, the number of times each word appears in the information is defined as a space vector of each information as shown in FIG. 6 (step 505). ), And record them in the search database (8). The number of appearances (601) of each word in FIG. 6 indicates the degree of association of each word with each piece of information.

The sum of the space vectors of all the information recorded in the database (1) is stored in the database (1).
It is defined as a space vector indicating its own feature, and is recorded in the search database (8) in a format as shown in FIG. 7 (step 506). The number of appearances of each word in FIG.
01) represents the degree of relevance of each word to the database (1). Also, the user separately defines the classification information (702) of the database (1), and the information is recorded at the same time.

FIG. 8 shows a processing flow of the search request input device (4).

First, a search request is interactively input using a search request input screen (step 801). FIG. 9 shows an example of the search request input screen at that time. In the search request input field (901) in FIG. 9, a search request is input as a request sentence or a word related to information to be searched. Then, "O
By selecting “K” (902), the process proceeds to the next process, and by selecting “Cancel” (903), the process returns to the immediately preceding process.

Next, when the search request is input as a sentence, the words included in the sentence are extracted by using a method such as morphological analysis (step 802).

Next, a keyword related to the search request input by the user is added using the information recorded in the search history database (12) (step 803). FIG. 10 shows the data structure of search history data relating to a search request. In the search history data, the degree of relevance (1001) between all words included in all databases handled in the past by the system is defined. This value is defined as 50 when the system is used for the first time, and is updated based on the flow of processing to be described later in the process of repeating the search. The upper limit of the value of the degree of association is 100, and the lower limit is 0.

FIG. 1 shows a flow of processing for adding a keyword.
It is shown in FIG. With respect to all the words included in the search request, words whose relevance to each word is 80 or more are extracted using the search history data shown in FIG. 10 (step 1101,
1102). The word extracted by the above processing is added to the search request as an additional keyword 1101a.

Returning to the processing of FIG. 8, a vector having a word group constituting the search request defined in the processing of FIG. 11 as a component is defined as a space vector of the search request (step 804).

The processing of the search request priority information input device (9) will be described.

The priority value is defined in the range of 0 to 100. The initial value of the priority value is determined according to the processing flow of FIG. If the keyword is a word input by the user or a word extracted from a search sentence input by the user, the initial value of the priority is uniformly 100 (step 1).
201).

If the keyword is automatically added using the information on the search history, the average of the relevance values used for extracting the keyword is calculated (step 120).
2) The value obtained by further reducing the average value by と is defined as the initial value of the priority (step 1203).

The dialog setting of the priority value is performed using the screen shown in FIG. For each keyword, the value of the priority is a numerical value (1
301) and a bar graph (1302). The length of the bar graph (1302) is proportional to the magnitude of the priority value. A pointer (1303) is attached to the right end of the bar graph. The user clicks the pointer (1303) with the mouse cursor (1304) and slides the pointer left and right to set a priority value. At this time, the movement of the pointer (1303) is matched with the bar graph (1303).
2) The length and the value of the numerical value (1301) change. Also,
The keyword input by the user and the keyword automatically added using the search history are displayed separately. After setting,
By selecting “OK” (1305), the process proceeds to the next process,
By selecting “cancel” (1306), the priority value is returned to the initial value.

FIG. 14 shows a processing flow of the search target database input device.

First, the user specifies a method of selecting a search target database (step 1401). FIG. 15 shows a selection screen at this time. The selection method of the search target database is specified by direct selection (1501) by the user, automatic selection by inputting a search target (1502), or automatic selection by using history data (1503). Specifically, the user moves the mouse cursor (1504) to the block of the display method of the selection method to be specified, and clicks the mouse button. Then, the blocks displaying the specified selection method are displayed in different colors. In that state,
By selecting “OK” (1505), the process proceeds to the next process,
By selecting “cancel” (1506), the current selection state is canceled and the state returns to the initial state. If no selection method is specified, "OK" (1505)
Cannot be selected.

If it is specified by the user that the user directly selects (1501) in the processing of step 1401, the user selects one of the databases (1) connected to the system to be the center of the search. (Step 1402).

FIG. 16 shows a selection screen at this time. On this screen, a list of databases (1) to be selected is displayed (1601), and a check box (1602) is displayed at the head of each database name. The user moves the mouse cursor (1603) to the check box (1602) of the database name to be selected and presses the mouse button to specify the database to be the center of the search. At this time, the designated check box (1602)
Are displayed in different colors. If another check box (1602) is designated in a state where one of the check boxes (1602) is already displayed in a different color, the previously designated check box (1602) is designated.
The designation of 2) is canceled, and the newly designated check box (1602) is displayed in a different color. In that state,
By selecting "OK" (1604), the process proceeds to the next process,
When "cancel" (1605) is selected, the process returns to the immediately preceding process. At this time, if "OK" (1604) is selected while no check box (1602) is specified, an error message "No database has been selected" is displayed, and the database selection process is performed again. Return to

After the database that is the center of the search is selected in the above process, a database to be searched other than the database is selected (step 140).
3). FIG. 17 shows a selection screen of the search target database. The screen displays a list of databases (1) to be selected (1701), and a check box (1702) is displayed at the beginning of each database name. Also, the check box (1702) of the database that is the center of the search is color-changed from the beginning. The user moves the mouse cursor (1703) to a check box (1702) of a database name to be selected other than the database that is the center of the search, and presses a mouse button to specify a database to be searched. At this time, the designated check box (1702) is the check box (170) of the database that is the center of the search.
The color is displayed with a color different from that of 2). It is also possible to select a plurality of databases to be searched. In this state, by selecting “OK” (1704), the process proceeds to the next process. When “Cancel” (1705) is selected, all the designations of the search target database are released. In addition, it is possible not to select any database to be searched other than the database that is the center of the search.

After the above selection is completed, an initial value of the priority value of each database is defined using the database selection history data (step 1404). FIG. 18 shows the structure of the history data used in the above processing. In the history data, the degree of relevance (1801) between all databases handled by the system in the past is defined. When this system is used for the first time, the values of all the degrees of relevance are defined as 50, and thereafter, are updated based on the processing flow described later in the process of repeating the search. The upper limit of the value of the degree of association is 100, and the lower limit is 0. As the priority value of the database, the database that is the center of the search is set to 100, and the value of the relevance to the database that is the center of the search of the history data is used for the database to be searched.

Next, a case where the user designates automatic selection (1503) of the search target database (1) in the processing of step 1401 in FIG. 14 will be described. First, using the space vector of each database recorded in the search database (8) in the structure as shown in FIG. 7 and the space vector of the search request defined in the processing of step 804 in FIG. For (1), an evaluation value is calculated by the following formula.

Evaluation value = Σk1 × k2 (1) where k1 is the weight value of one component of the space vector of the search request, and k2 is the space vector of the database.
The value of the weight of the component of the same word as the component of the space vector of the search request is shown. If the same word does not exist, the evaluation value for that component is set to 0.

Then, of the databases (1) to be processed by the system, the database having the largest evaluation value is defined as the database to be the center of the search (step 14).
05). Next, the reference value of the database is set to 100, and the value of the relevance of another database to the database is calculated using the history data shown in FIG. 18 (step 1406). Then, the database (1) having the relevance value of 50 or more is defined as a database to be searched (step 1407). The initial value of the priority value of each database is 100 for the database that is the center of the search, and the database to be searched defines the relevance value as the priority value (step 1408).

Next, a case where the user has specified the input of the search target (1502 in FIG. 15) in the processing of step 1401 in FIG. 14 will be described.

First, using the classification information (702) of each database (1) recorded in the search database (8) in the structure as shown in FIG.
02), a list of databases (1) belonging to the classification is defined in advance by a user or a system administrator as shown in FIG. At this time, an initial value of the priority value of the database (1) included in each classification information (702) is also defined (1901).

Then, first, using the screen as shown in FIG. 20, one of the classification information (702) defined in FIG. 19 is selected as the search target of the user. On this screen, a list of classification information (702) to be selected is displayed (200).
1) A check box (2002) is displayed at the top of each classification information (702). The user moves the mouse cursor (2003) to the check box (2002) of the classification information (702) to be selected and presses a mouse button to specify the classification information (702) to be searched. At this time, the designated check box (1602)
Are displayed in different colors. If another check box (2002) is designated while any one of the check boxes (2002) is already displayed in a different color, the previously designated check box (200) is designated.
The designation of 2) is canceled, and the newly designated check box (2002) is displayed in a different color. In that state,
By selecting "OK" (2004), proceed to the next process,
If "Cancel" (2005) is selected, the process returns to the immediately preceding process. At this time, if “OK” (2004) is selected while no check box (2002) is specified, an error message “No search target is selected” is displayed, and the search target is displayed again. Return to the selection process.

The classification information (70
The database (1) included in 2) is read from the definition information shown in FIG. 19, and is set as the search target database (1) (step 1409). At the same time, the initial value of the priority value of each database is read from the information in FIG. At this time, among the databases included in the specified classification information (702), the database having the largest initial value of the priority value is defined as the database to be the center of the search.

The processing of the database priority information input device (3) will be described. As shown in FIG. 21, the priority of each database is determined as shown in FIG. 21 by using the database which is the center of the search defined by the search target database input device (2), the search target database, and the initial value of the priority value of each database. Display the degree setting screen. The value of the priority is displayed as a numerical value (2101) and a bar graph (2102) for each database. The length of the bar graph (2102) is proportional to the magnitude of the priority value. A pointer (2103) is attached to the right end of the bar graph, and the user can move the pointer (2103) with the mouse cursor (2104).
Click, and slide it left and right to set the priority value. At this time, according to the movement of the pointer (2103), the length of the bar graph (2102) and the numerical value (2
101) changes. After completing the setting, click "OK" (13
05), the process proceeds to the next process, and "Cancel"
By selecting (2106), the priority value is returned to the initial value.

The database limiting device (13) extracts information of the database (1) to be searched using the above information.

FIG. 22 shows the flow of processing of the database search device (5).

In this processing, the space vector of the search request is expressed as follows: V = {kv1 · t1, kv2 · t2,..., Kvn · tn} (Equation 2) where tn: word kvn: nth word , And the space vector of each of the databases A and B to be searched is DA = {ka1 · a1, ka2 · a2,..., Kan · an} (Equation 3) DB = ｛kb1 B1, kb2b2,..., Kbnbn... (Equation 4) where an: word kan: weight value for the nth word (defined in FIG. 7) bn: Word kbn: The weight value for the nth word (defined in FIG. 7), and the priority values of the search target database are αA and αB, respectively (step 2201).

In this processing, the vectors VA and VB obtained by projecting the space vector V set as described above onto the space defined by DA and DB are obtained as follows (step 2).
202).

VA = {0, kv1 · ka2 · t1,..., Kvi · kap · ti} (Equation 5) VB = {kv2 · kb1 · t2, 0,..., Kvj · kbq · Tj｝ ... (Equation 6) where t1 = a2, ..., ti = ap, and t2 = b
1,..., Tj = bq.

Then, VA and VB are normalized, and vectors obtained by projecting them on the space of DA∩DB are VA ′ and V
B ′ (step 2203). Next, a final search vector W is defined as in the following equation (step 220).
4).

W = αA · VA ′ + αB · VB ′ (Equation 7) Next, all information vectors (defined in FIG. 6) existing in the space of the DA∩DB, and a search vector The angle formed with W is calculated (step 2205). As shown in FIG. 23, the angle θ to be formed is calculated by dividing two vectors into W and B1.
Then, cos (θ) = (W · B1) / || W |||| B1 || (Equation 8) Here, || W ||: is obtained by the norm of W. Then, an ID of information whose angle is less than or equal to a certain value is extracted (step 2206), and the ID and the information of the angle (step 2207) are sent to the extracted data list display device (10).

The processing of the extracted data list display device (10) will be described.

FIG. 24 shows an example of the extracted data list display. The display screen is divided by the number of databases specified as search targets. Each divided area (2401)
Displays the database name (2402). Then, a part of the title of the extracted information is replaced with a divided area (240
1) Displayed above (2403).

At this time, if the number of characters in the title is large, several characters are extracted from the head of the title and displayed. In addition, the display of each title (2403) is performed by changing the color according to the size of the angle between the title and the search vector. Specifically, the smaller the size of the formed angle, the deeper the display color. further,
When the user positions the mouse cursor (2404) on the display of each title, a detailed information display window (2405) at the bottom of the screen displays the full title of the information (402), creator (403), affiliation (404), creation date (405) and the update date (406) is displayed.

As a method of dividing the area for each database, a method of radially dividing the center of the screen as a base point as shown in FIG. 24, a method of dividing the screen in the horizontal direction as shown in FIG. When the information cannot be displayed, the screen is scrolled using a scroll bar (2601), or the area is divided into a matrix as shown in FIG. The configuration is used. At this time, which division method is used depends on the selection of the user.

As an optional function, the degree of coincidence between the components of the space vector of each piece of information and the components of the space vector of each database to be searched (calculated as coincidence in the case of the same word) is calculated. Defined as a value indicating the degree of relevance between information and each search target vector,
By using the above values, it is also possible to display the relevance of each information and the database specified as a search target in a radar chart (2801) as shown in FIG. At this time, it is assumed that the larger the value of each axis of the radar chart, the higher the relevance to the database indicated on that axis.

Further, using the above-mentioned value of the degree of relevancy, it is possible to represent the relevancy of each information and the database designated as a search target by a bar graph (2901) as shown in FIG. At this time, the length of the bar graph of each database is normalized so that the length of the entire bar graph obtained by adding the lengths of the bar graphs of each database is constant. An area (290) indicating the name of each database
The same color is used for the display color of 2) and the color of the bar graph indicating the degree of relevance to the database in the bar graph.

As means for reflecting the value of the angle between the space vector of the information and the search vector in the display of the title of each information, as shown in FIG. In addition to changing the color, a method of increasing the size of the information display as the value of the angle formed is smaller as shown in FIG. 30 or a method of increasing the transparency of the information display as the value of the angle formed is larger is used. It is also possible.

A list of search results is displayed by the above processing. At this time, as shown in FIG. 31, the priority values of the databases and the search keywords are corrected, the search is performed again, and the state of the change in the list display is changed. It is also possible to narrow down the search results while checking the degree of change and the degree of change.

The processing of the display data specifying device (11) will be described.

FIG. 32 shows an example of a screen for designating display data.
In this example, the display designation window (3203) is opened by moving the mouse cursor (3201) to the title display (3202) and pressing the mouse button. During the display designation window, the title full text (402), creator (403), affiliation (404), creation date (405), and update date (406) are displayed as detailed information of the designated information. . And on the window (3203),
By selecting “OK” (3204), the process proceeds to the information display process. By selecting “Cancel” (3205), the designation is canceled, and the display designation window (3203)
Close.

When the display data is specified, a part of information of the search history database (12) and the search database (8) is updated in the following steps.

FIG. 33 shows a flow of a process of updating the weight value of each information in the search history database (12) and the search database (8) of the degree of association between words.

First, as a search history, words included in the search request, additional keywords, information on the priority of the keywords,
The word and its weight information (defined in FIG. 6) included in the information designated for display are extracted (step 33).
01).

Next, words having a weight value of 5 or more are extracted from words included in the information designated for display (step 3302). Then, in the search history database (12) of the degree of association between words, the value of the degree of association between the word included in the search request and the additional keyword and the word extracted in the processing of (Step 3302) is calculated as follows: + ABS (weight value × (priority / 100)) (Equation 9) (step 3303). At this time, if the word is registered for the first time, the old relevance is set to 0. Further, in the search database (8) for each information, the value of the weight of the same word as the word included in the search request + the additional keyword is calculated as follows: new weight value = old weight value + ABS (5 × (priority / 100)) ) (Equation 10) (Step 3304) At this time, in the case of a word registered for the first time, the value of the old weight is set to 0.

Next, the flow of the process of updating the search history database (12) of the degree of association between the databases (1) is shown in FIG.
It is shown in FIG.

First, the priority values of the database serving as the center of the search and the other databases to be searched are extracted (step 3401). Then, so that the priority value of the database that is the center of the search is 100,
The priority value of the search target database is adjusted (step 3402).

The average value of the priority value of the search target database calculated in the above processing and the relevance value of the search target database with respect to the database serving as the center of the search in the history data is calculated as It is calculated for each search target database (step 3403). And
The average value is recorded in the search history database (12) as a new relevance value (step 3404).

Next, regarding the update of the search history database (12), FIG. 35 shows the flow of processing of the history correction method based on the update interval.

First, the update date of the history data is recorded in each of the word data included in the information, the data indicating the relevance between words, and the data indicating the relevance between the databases (1) (step). 3501). Then, all the above data are checked at regular intervals (step 350).
2) For data for which data has not been updated for a certain period of time, a value of the weight or the degree of association is subtracted by a certain amount (step 3503).

The processing of the data display device (6) will be described. FIG. 36 shows an example of a data display screen. This screen presents the information shown in FIG. 4 recorded in the database (1) with respect to the information designated by the user for display.
When "OK" is selected, the display of the data is terminated, and the process proceeds to the search request input device (4).

[0097]

According to the present invention, when a search is performed on a plurality of databases, the search is performed in consideration of the priority of each database, and when the search results are displayed in a list, the search results are not displayed. The features can be conveyed to the user in a more easily understood manner.

More specifically, by using the present invention, the effect of improving the efficiency and accuracy of information retrieval for a plurality of databases can be obtained.

Further, according to the present invention, it is possible to improve the search efficiency and the search accuracy by utilizing the search history.

Further, according to the present invention, an effect is obtained that a search can be performed in accordance with the intention of the user.

Further, according to the present invention, it is possible to easily select a search target database and set information on the priority of the database.

Further, according to the present invention, it is possible to obtain the effect that the accuracy of the search history data can be improved by more appropriately recognizing the intention of the user.

Further, according to the present invention, an effect is obtained that the extracted data can be presented to the user in an easily understandable manner.

Further, according to the present invention, an effect is obtained that the relationship between the extracted information and each database can be grasped more appropriately.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a basic configuration in an embodiment of a database search method.

FIG. 2 is a block diagram showing a specific example of a hardware configuration of a database search method.

FIG. 3 is a block diagram showing a configuration in another embodiment of a database search method.

FIG. 4 is an explanatory diagram showing an example of information recorded in a database.

FIG. 5 is a flowchart showing the flow of processing for creating a search database.

FIG. 6 is an explanatory diagram showing an example of space vector information of each information in a search database.

FIG. 7 is an explanatory diagram showing an example of space vector information of a search database.

FIG. 8 is a flowchart showing the flow of a search request input process.

FIG. 9 is an explanatory diagram showing an example of a screen for inputting a search request.

FIG. 10 is an explanatory diagram showing an example of search history data indicating the degree of association between words.

FIG. 11 is a flowchart showing the flow of a keyword addition process based on search history information.

FIG. 12 is a flowchart showing the flow of processing for defining an initial value of the priority of a keyword.

FIG. 13 is a flowchart illustrating an example of a screen for setting dialogue for priority of a keyword.

FIG. 14 is a flowchart showing the flow of processing for selecting a search target database.

FIG. 15 is an explanatory diagram showing an example of a screen for designating a selection method of a search target database.

FIG. 16 is an explanatory diagram showing an example of a screen for selecting a database serving as a search center.

FIG. 17 is an explanatory diagram showing an example of a screen for selecting a database to be searched;

FIG. 18 is an explanatory diagram showing an example of search history data indicating the degree of association between databases.

FIG. 19 is an explanatory diagram showing an example of a database according to classification information and definitions of its priorities.

FIG. 20 is an explanatory diagram showing an example of a screen for inputting a search target.

FIG. 21 is an explanatory diagram showing an example of a screen for setting a priority of a database;

FIG. 22 is a flowchart showing the flow of a database search process.

FIG. 23 is an explanatory diagram of an angle between vectors.

FIG. 24 is an explanatory view showing a screen example of a list display of extraction information.

FIG. 25 is an explanatory view showing a screen example of a list display of extraction information.

FIG. 26 is an explanatory view showing a screen example of a list display of extraction information.

FIG. 27 is an explanatory view showing a screen example of a list display of extraction information.

FIG. 28 is an explanatory view showing a display screen example of the relevance of each information and the search target database.

FIG. 29 is an explanatory diagram showing an example of a display screen showing the relevance of each piece of information and the search target database.

FIG. 30 is an explanatory view showing a screen example of a list display of extraction information.

FIG. 31 is an explanatory diagram of updating a search result based on a priority correction.

FIG. 32 is an explanatory diagram showing an example of a screen for specifying display data.

FIG. 33 is a flowchart showing the flow of processing for updating a search history database.

FIG. 34 is a flowchart showing the flow of processing for updating a search history database.

FIG. 35 is a flowchart showing the flow of processing for updating the search history database.

FIG. 36 is an explanatory diagram showing an example of a display screen of a search result.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... database, 2 ... search target database input device, 3 ... database priority information input device, 4 ... search request input device, 5 ... database search device, 6 ... data display device, 7 ... search database creation device, 8 ... Search database, 9 ... Search request priority information input device, 10
... Extracted data list display device, 11 ... Display data designation device, 12 ... Search history database, 13 ... Database limited device.

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Haruki Ogawa 882-Chair, Oaza-shi, Hitachinaka-city, Ibaraki Pref. Hitachi, Ltd. Instrumentation Division (72) Inventor Toshihiko Nakano 5-2-1 Omika-cho, Hitachi City, Ibaraki Prefecture No. F-term in Hitachi Ltd.'s Omika Plant (reference) 5B075 KK03 KK07 ND03 NK02 NK32 PP03 PP13 PQ02 PQ14 PQ23 PQ36 PQ46 PR08

Claims (17)

[Claims] A search request input unit for inputting information indicating a search request; comparing the information in the database with the search request information to perform the search; In a database search method using a database search device that includes a database search unit that extracts information highly relevant to a request and a data display unit that displays the extracted information, a search object is selected from the two or more databases. A search target database input process for specifying a database to be set; and a database priority information input process for setting a search priority for each of the databases specified as the search target. In the processing, the priority is given to the database specified as the search target. Database search method and performing search Flip was. 2. The database search method according to claim 1, wherein, as a result of the search processing, an extracted data list display processing for displaying a list of information having a relevance equal to or greater than a specific value with the search request; A display data designation process for selecting information for displaying more detailed contents from the list displayed on the extracted data list display means; and a search history of the selection content in the display data designation process and the content of the search request. A search history storage process for recording data in a search history database; and a first priority correction process for correcting information on the type of the database to be searched and the priority of the database by using the search history data. A database search method, further comprising: 3. A database for recording information to be searched, search request input means for inputting information indicating a search request, and comparing the information in the database with the information on the search request to determine the relevance of the search request. In a database search method using a database search device comprising: a database search unit for extracting information having a high degree of search; and a data display unit for displaying the extracted information, a search priority is set for each word constituting the input search request. A search request priority information input process, and an extracted data list display process for displaying a list of information having a relevance equal to or more than a certain value with the search request as a result of the search process. A display data designation process for selecting information for displaying more detailed contents from a list displayed on the means; and the display data designation process. Search history storage processing for recording the selected content and the content of the search request in a search history database as search history data, and by using the search history data, priority of the search request and words constituting the search request Second to correct degree information
And a priority correction process. 4. The database search method according to claim 1, wherein an input of an instruction from a user is received, and a priority of said search target database input in said database priority information input processing is changed. A database search method further comprising a change process. 5. The database search method according to claim 3, wherein an input of an instruction from a user is received, and a priority of a word constituting the search request input in the search request priority input processing is changed. A database search method further comprising a search request priority change process. 6. The database search method according to claim 1, wherein the two or more databases are classified into a plurality of categories in advance, and when the database to be searched is specified, A database search method characterized by designating a database to be searched by selecting one of them. 7. The database search method according to claim 6, wherein priority information is defined in advance for databases included in each of the plurality of categories, and the priority of the database specified as the search target is determined. When setting, a database search method characterized in that a priority value of the search target database is determined by using the previously defined information. 8. The database search method according to claim 1, wherein when the database to be searched is specified, one database which is the center of the search is selected, and the relevancy is determined by a database other than the center database. A database search method characterized by allowing a user to select a database that has a database. 9. The database search method according to claim 2 or 3, wherein when displaying the list of the extracted data, the display screen is divided for each search target database, and each extracted data is displayed in the database. A database search method characterized in that the search is performed within the area of the search target database to which the user belongs. 10. The database search method according to claim 2, wherein when the list of the extracted data is displayed, each extracted data is displayed according to a degree of relevance between the search request and each extracted data. A database search method characterized by changing the display form of a database. 11. The database search method according to claim 10, wherein when changing the display mode of the extracted data, at least one of a display color, transparency, and size of each extracted data is changed. Characteristic database search method. 12. The database search method according to claim 2, wherein, when displaying the list of the extracted data, the degree of relevance between each extracted data and each search target database. A database search method characterized by displaying the data in a radar chart centering on the degree of association with each search target database. 13. The database search method according to claim 2 or 3, wherein when displaying the list of the extracted data, the degree of relevance between the extracted data and the respective search target database for each extracted data. , The length of a bar graph color-coded for each search target database. 14. A storage medium storing a program for causing a computer to execute the database search method according to any one of claims 1 to 13. 15. A search request inputting means for inputting information indicating a search request, two or more databases for recording information to be searched, and comparing the information in the database with the information in the search request. In a database search device including a database search unit for extracting information highly relevant to a request and a data display unit for displaying the extracted information, a database to be searched is specified from the two or more databases. Search target database input means, and database priority information input means for setting a search priority for each of the databases specified as the search target, wherein the database search means is specified as the search target. Characterized by performing a search according to the priority to the database Search equipment. 16. The database search device according to claim 15, wherein as a result of the search processing, an extracted data list display means for displaying a list of information having a relevance equal to or greater than a specific value with the search request; Display data specifying means for selecting information to be displayed in more detail from the list displayed on the extracted data list display means; and searching for the selected content in the display data specifying process and the content of the search request. A search history database that records as history data; and a first priority correction unit that corrects information on a type of a search target database and a priority of the database by using the search history data. Characteristic database search device. 17. A database for recording information to be searched, search request input means for inputting information indicating a search request, and comparing the information in the database with the information in the search request to determine whether the search request is relevant. A database search device comprising: a database search unit that extracts information having a high level of search; and a data display unit that displays the extracted information. The search request priority setting a search priority for each word constituting the input search request Degree information input means; extracted data list display means for displaying a list of information having a relevance equal to or more than a specific value with the search request as a result of the search processing; and a list displayed on the extracted data list display means Display data specifying means for selecting information for displaying more detailed contents from among the above, and the selection contents and the search request in the display data specifying process. A search history database that records the content of the search request as search history data, and using the search history data to correct the information on the search request and the priority of words constituting the search request.
And a priority correction means.