JP2001282811A - Knowledge data search device, knowledge data search method, and computer-readable recording medium storing program for searching knowledge data - Google Patents

Abstract

(57) [Summary] (With correction) [Problem] In a device for searching for an association between knowledge data desired by a user from a knowledge database created in association with knowledge data such as ideas and ideas,
A user can search and output a desired node and a link between the nodes. An unnecessary condition input means for inputting unnecessary knowledge data and an unnecessary database creating means are provided, and a search result output means for deleting knowledge data stored in the unnecessary database from a search result database.
Is further provided, even when the search result obtained by the search condition input means 6 includes knowledge data other than that desired by the user, such knowledge data can be reliably deleted. Only the desired knowledge data can be stored and output in the search result database 10 as the search result.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a knowledge data retrieval apparatus for retrieving knowledge data desired by a user and an association between the knowledge data from a knowledge database created in association with knowledge data such as ideas and ideas.
It relates to a knowledge data search method.

[0002]

2. Description of the Related Art Hitherto, as a method of associating ideas and ideas or words and sentences obtained through the Internet, a graph in a so-called graph theory in which character strings such as words and the like are connected by a line is known. In the graph, a display unit (hereinafter, referred to as a node) of knowledge data such as a phrase or a sentence is associated with another node (hereinafter, referred to as a link), and a database using such a graph has been acquired. By linking and extending nodes and links to existing databases,
It aims to accumulate knowledge.

[0003] A user can obtain knowledge by searching the database as an accumulation of knowledge for another node associated with a desired node.

[0004]

However, when the user obtains desired knowledge from the database, the search method as described above uses the node (and the homologous word, etc.) originally searched for by the user in the data after the search. There is a problem that a node (and a link with the node) other than the link with the node is searched and displayed.

The present invention has been made to solve such a problem, and an object of the present invention is to enable a user to search for and output a desired node and a link between the nodes.

[0006]

According to the present invention, there is provided a knowledge data searching apparatus for searching for desired knowledge data from a knowledge database comprising knowledge data and association between the knowledge data. A knowledge database 7 composed of knowledge data and an association between the knowledge data, and a search condition input means 2 for inputting knowledge data desired by a user.
Search database creating means 3 for creating a search database 8 from the knowledge data desired by the user acquired by the search condition input means 2, unnecessary condition input means 4 for inputting unnecessary knowledge data, and unnecessary condition input means Unnecessary database creating means 5 for creating an unnecessary database 9 from the unnecessary knowledge data obtained in 4;
7, a search result output means 6 for outputting search results from the search database 8 and the unnecessary database 9, and a search result database 10. The search result output means 6 includes knowledge data in the search database 8,
After extracting the knowledge data associated with the knowledge data and the association thereof from the knowledge database 7, the extracted knowledge data and the association are stored in the search result database 10, and the knowledge data in the unnecessary database 9 are stored in the search result database 10. The knowledge data associated with the knowledge data in the unnecessary database 9 and the association are deleted from the search result database 10 to update the search result database 10, and the updated search result database is output by the output device. It is characterized by having been done. Thus, unnecessary condition input means 4 for inputting unnecessary knowledge data and unnecessary database creating means 5 are provided, and search result output means 6 for deleting knowledge data stored in unnecessary database 9 from search result database 10. Is further provided, even if the search result obtained by the search condition input means 6 includes knowledge data other than that desired by the user, such knowledge data can be reliably deleted, and therefore, Only the desired knowledge data can be stored and output in the search result database 10 as the search result.

Further, according to the present invention, a search database 8 is created from knowledge data desired by a user, an unnecessary database 9 is created from unnecessary knowledge data, and the knowledge data in the search database 8 and the knowledge database in the search database 8 are stored. The knowledge data associated with the knowledge data and the association thereof are extracted from the knowledge database 7 composed of the knowledge data and the association between the knowledge data, and the extracted knowledge data and the association are extracted from the search result database 10.
The knowledge data in the unnecessary database 9, the knowledge data associated with the knowledge data in the unnecessary database 9, and the association are stored in the search result database.
10 and update the search result database 10 by deleting
A knowledge data search method characterized by outputting the updated search result database 10 can also be used.

The present invention also provides a procedure for storing knowledge data desired by a user in a search database 8, a procedure for storing unnecessary knowledge data in an unnecessary database 9, a procedure for storing knowledge data in the search database 8, and a procedure for storing the knowledge data in the search database 8. 8
The knowledge data associated with the knowledge data in the database and its association are extracted from the knowledge database 7 composed of the knowledge data and the association between the knowledge data, and the extracted knowledge data and the association are extracted from the search result database.
10 to update the search result database 10 by deleting from the search result database 10 the procedure for storing in the unnecessary database 9, the knowledge data in the unnecessary database 9, the knowledge data associated with the knowledge data in the unnecessary database 9 and the association thereof. The procedure for outputting the updated search result database 10 on an output device may be executed by a computer, and may be configured by a computer-readable recording medium that records a program for searching for knowledge data.

[0009]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of a knowledge data search device according to the present invention will be described below with reference to the drawings.
FIG. 1 is a block diagram showing a hardware configuration of a knowledge data search device according to the present invention. In FIG. 1, reference numeral 1 denotes a control unit composed of a general-purpose microcomputer, a dedicated LSI, or the like. The control unit 1 has a keyboard capable of inputting commands and the like,
An input device 11 such as a mouse; a display device 12 such as a CRT display or a liquid crystal display for displaying output results on a display screen; a printing device 13; and an external storage device 14 such as a hard disk.
Are connected via a cable. In addition, the display device 12,
The printing device 13 functions as an output device that outputs an output result and the like. Although not shown, the control unit 1 is provided with lower-level control units that control input / output and search, and these lower-level control units are configured by firmware or software such as a ROM. .

In the knowledge data search apparatus shown in FIG. 1, a search condition input means for inputting knowledge data desired by the user is used as means for searching for knowledge data desired by the user and knowledge data associated with the knowledge data. 2, search database creation means 3, unnecessary condition input means 4 for inputting unnecessary knowledge data 4, unnecessary database creation means 5,
A search result output means 6 for outputting search results and a knowledge database 7 are provided. Each of the units is configured by a program stored in the external storage device 14 or the like, and is appropriately called by a temporary storage unit such as a RAM under the control of the control unit 1 to realize the function of each unit.

In FIG. 1, reference numeral 7 denotes a knowledge database created by associating knowledge data composed of words and phrases such as ideas and ideas or words and sentences obtained from the Internet or the like. Reference numeral 8 denotes a search database created by the search database creation means 3 using the user's desired knowledge data acquired by the search condition input means 2. Reference numeral 9 denotes an unnecessary database created by the unnecessary condition database creating means 5 using the knowledge data that is unnecessary by the user and acquired by the unnecessary condition input means 4. Reference numeral 10 denotes a search result database created by the search result output means 6. In addition, the knowledge database
7, the search database 8, the unnecessary database 9, and the search result database 10 are configured by an external storage device 14 such as a hard disk. Also, in the present embodiment, a graph is described as an example of a structure for associating knowledge data, a knowledge data is called a node, and an association between knowledge data is called a link, and the knowledge database 7 is created in advance. .

FIG. 2 is a flowchart showing a search method using the knowledge data search device of the present invention. First, the search condition input means 2 is controlled by the control unit based on the operation of the input device 11 by the user.
Under the control of 1, the user inputs a node to be left (step S1 in FIG. 2). For example, as shown in FIG. 3, the search condition input means 2 first supports the user by displaying a question A "What is the node to search?" On the display screen of the display device. When a user inputs a desired node using the input device 11, the search condition input means
2 acquires the input contents and displays them on the display screen. The nodes input in this way (nodes that the user intends to keep) are stored in the search database 8 by the search database creating means 3 under the control of the control unit 1 (step S2 in FIG. 2).

Next, the unnecessary condition input means 4 inputs a node unnecessary for the user under the control of the control unit 1 based on the operation of the input device 11 by the user (step S3 in FIG. 2).
For example, as shown in FIG. 3, the unnecessary condition input means 4 first asks the user a question B "What is an unnecessary node?"
Is displayed on the display screen of the display device to provide input support. When a user inputs an unnecessary node using the input device 11, the unnecessary condition input means 4 acquires the input content and displays it on a display screen. The unnecessary nodes input in this manner are stored in the unnecessary database 9 by the unnecessary database creating means 5 under the control of the control unit 1 (see FIG. 2).
Step S4).

When the search database 8 and the unnecessary database 9 are created, the nodes stored in the search database 8 are compared with the nodes stored in the knowledge database 7 (step S5 in FIG. 2). Such a comparison will be described in detail with reference to FIG. In the present embodiment, a case where a node is a word will be described as an example.

FIG. 4A shows an example in which the knowledge database 7 is displayed on the display screen. In FIG. 4, "multimedia", "database", "object-oriented", "still image", "moving image" ”,“ Voice ”,“ AV ”,
"Adult", "network", "broadband", etc. indicate the nodes 21a, 21b, 21c,..., 21l, respectively, and the bars in the figure indicate links 22a, 22b, 22c,.・ ・,
Indicates 22j. FIG. 4B shows the contents stored in the search database 8. According to FIG. 4B, the search database 8 contains “multimedia”, “moving image”, “A”
V "is stored.

The search result output means 6 includes a search database
The nodes stored in 8 are extracted from the nodes stored in the knowledge database 7. Here, the search result output means 6 compares the knowledge database 7 with the search database 8 and extracts not only the nodes stored in the search database 8 but also the nodes associated with the nodes and their associations (links). Search results database 10
To be stored. In the present embodiment, a node indirectly associated with a certain node, for example, a node associated via two links (in FIG. 4A, the node 21i is , Node 21a
, Etc.) are all extracted. That is, in the present embodiment, of the nodes stored in the knowledge database 7, all nodes and links except “broadband” and “network” are extracted in the search result database 10 and stored in the search result database 10. (Step S6 in FIG. 2, FIG. 4C). It should be noted that an ID for identifying each node is given to each node, and a node associated with a certain node is designated by an ID, indicating that a link exists in the node. Therefore, whether or not the extracted node has a link can be detected based on whether or not the node specifies an ID.

Further, the search result output means 6 calls up the unnecessary database 9 under the control of the control unit 1 and compares it with the search result database 10 (step S7 in FIG. 2). FIG.
(A) shows the contents stored in the unnecessary database 9. In FIG. 5A, for example, the unnecessary database 9
Shows a case where “AV” and “adult” are stored. The search result output means 6 deletes the nodes stored in the unnecessary database 10 from the search result database 10. Here, since "AV" is also stored in the search database 8, it is determined that the node is a necessary node and is left in the search result database 8, and as a result, "adult" is deleted. At this time, the search result database 10 is updated by deleting not only the node stored in the unnecessary database 10 but also the node associated with the node and the link as the association (step S8 in FIG. 2, FIG. 5).
(B)). Also at the time of this deletion, whether or not there is a node associated with the node stored in the unnecessary database can be detected based on whether or not the node specifies an ID. Finally, the search result output means 6 outputs the content stored in the search result database 10 by displaying it on the display screen of the display device 12 or printing it by the printing device 13 (step S9 in FIG. 2). .

As described above, by using the knowledge data search device according to the present invention, when a user searches for a desired node, unnecessary nodes and links such as homonyms are excluded (deleted), and the user can be more reliably searched. Only the desired nodes and links can be extracted into the search result database.

Incidentally, a floppy (registered trademark) disk, C
It is also possible to record a procedure (program) for causing the computer to execute the steps S1 to S8 on a computer-readable recording medium such as a D-ROM, a CD-R, and an MO.

In the above embodiment, the graph is used for the knowledge database, the nodes are used for the knowledge data, and the links are used for associating the nodes. However, the knowledge data in the present invention is not limited to the nodes. Appropriate knowledge data such as knowledge used for the semantic network and keys and values used for the associative structure can be used. In the link, not only the bar, but also an appropriate association such as a bar in which the weight for the link is indicated by a numeral or an arrow or the like may be used. Further, as the knowledge data, not only words but also arbitrary data such as sentences and images can be used.

The knowledge database may be created in advance as in the above-described embodiment, or may be created by the user each time, or by adding knowledge data and links separately created from data taken from the Internet or the like. It may be updated. Further, the knowledge database created in advance may have a configuration in which knowledge data can be added after it is created in advance, or a configuration in which knowledge data cannot be added. When inputting knowledge data and links to the knowledge database, not only data taken from the Internet etc. but also nodes were automatically extracted from data read by a scanner etc. to create knowledge data, or input by voice An appropriate method such as extracting knowledge data from data can be used.

In the above-described embodiment, the knowledge database and the search database were compared after the search database and the unnecessary database were created. However, the knowledge database and the search database were compared after the search database was created, and then the unnecessary database was created. May be created. Furthermore, a search database may be created after creating an unnecessary database. In short, as long as the knowledge database is compared with both the search database and the unnecessary database, the order in which the unnecessary databases are created or compared does not matter.

The unnecessary database and the search database are not necessarily limited to one, and a plurality of unnecessary databases and search databases may be created. For example, two kinds of unnecessary databases A and B
When two types of search databases C and D are created, first, the unnecessary database A is compared with the knowledge database to create a search result database E, and then the search database
The search result database E is updated by comparing C with the search result database E. Similarly, after comparing the unnecessary database B with the knowledge database to create a search result database F, the search database D is changed to the search result database.
Update the search result database F by comparing it with F. The updated search result databases E and F are displayed as search results.

Also, by combining the unnecessary database A with the search database D and combining the unnecessary database B with the search database C, both search result databases can be displayed as search results in the same manner as described above. In this way, by creating a plurality of unnecessary databases and a search database, comparing them appropriately and comparing them with a knowledge database, and creating a search result database,
The search result most desired by the operator can be obtained more reliably.

Further, in the above-described embodiment, the search result database is updated once, but steps S1 to S8 in FIG.
May be repeated to repeatedly update the search result database.

In the above-described embodiment, the unnecessary database is created based on unnecessary knowledge data manually input in advance. However, knowledge data that appears more than a predetermined number of times is automatically determined by the user as being known. An unnecessary database may be created based on the number of appearances so as to delete the unnecessary database. In such a case, each time the search result database is updated, a program that configures the search result output means is provided so as to have a function of counting the number of appearances of the knowledge data remaining in the search result database. Then, the knowledge data exceeding the preset number of appearances is input from the search result output unit to the unnecessary condition input unit. The knowledge data input to the unnecessary condition input means is accumulated by the unnecessary database creating means as in the above embodiment, and an unnecessary database is created.

For the search database, in addition to the method of inputting the knowledge data desired by the user into the search condition input means, for example, when a word desired by the user is input, all the knowledge data including the word is input by the search condition input means. The search condition input means may be configured to create a search database from the extracted knowledge data extracted by the search database creation means.

In the above-described embodiment, the CRT display and the printer connected to the control unit by a cable are used as output devices. However, the output device does not have to be directly connected to the control unit by a cable or the like. Alternatively, a display, a printer, or the like connected indirectly to the control unit via a network, the Internet, or the like may be used.

Further, with respect to the output screen of the search result database, all data stored in the search result database may be output, or only the nodes within a specific number of links from a desired node may be output. good.

Further, with regard to the unnecessary database, according to the above embodiment, all the nodes connected (associated) to the unnecessary nodes by links are deleted.
The search condition output means may be configured to delete a specific number of links connected to unnecessary nodes.

[0031]

According to the knowledge data search apparatus of the present invention, unnecessary condition input means and unnecessary database creation means are provided, and a search result output for deleting knowledge data stored in the unnecessary database from the search result database is provided. Since the means is further provided, the user can reliably retrieve and output only desired knowledge data.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an embodiment of a knowledge data search device according to the present invention.

FIG. 2 is a flowchart showing a knowledge data search method according to the present invention.

FIG. 3 is a display screen of a display device when a search condition input unit and an unnecessary condition input unit urge a user to input;

4A is a display screen showing the contents of the knowledge database, FIG. 4B is a character display of the contents stored in the search database, and FIG. 4C is a search result database after comparing the search database with the knowledge database Display screen showing the contents stored in.

5A is a character display of contents stored in an unnecessary database, and FIG. 5B is a display screen showing contents stored in a search result database updated after a comparison between the unnecessary database and the search result database.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 Control part 2 Search condition input means 3 Search database creation means 4 Unnecessary condition input means 5 Unnecessary database creation means 6 Search result output means 7 Knowledge database 8 Search database 9 Unnecessary database 10 Search result database

 ────────────────────────────────────────────────── ─── Continuing from the front page (72) Inventor Masahiro Nishida 4-16-20 Kikawahigashi, Yodogawa-ku, Osaka Computer Consultant Co., Ltd. F-term (reference) 5B069 AA01 CA18 KA06 LA03 5B075 NK25 NK33 NR02 PP02 PP03 PQ02 QP01 5E501 AA01 AC17 AC23 AC33 BA03 CA03 CA04 CB02 CB09 CB13 EA12 EB17 FA13

Claims (3)

[Claims] 1. A knowledge data retrieval apparatus for retrieving desired knowledge data from a knowledge database comprising knowledge data and an association between the knowledge data, comprising a knowledge data and an association between the knowledge data. A knowledge database, search condition input means for inputting knowledge data desired by the user, search database creation means for creating a search database from the knowledge data desired by the user acquired by the search condition input means, and unnecessary knowledge data. Unnecessary condition input means for inputting, unnecessary database creating means for creating an unnecessary database from the unnecessary knowledge data obtained by the unnecessary condition input means, and a search result for outputting a search result from the knowledge database, search database and unnecessary database Output means, and a search result database, The search result output means extracts, from the knowledge database, the knowledge data in the search database, the knowledge data associated with the knowledge data in the search database, and the association thereof, and outputs the extracted knowledge data and the association thereof to the search result. And updating the search result database by deleting the knowledge data in the unnecessary database, the knowledge data associated with the knowledge data in the unnecessary database, and the association from the search result database. A knowledge data search device configured to output a result database on an output device. 2. A knowledge data retrieval method for retrieving desired knowledge data from a knowledge database composed of knowledge data and an association between the knowledge data, comprising: creating a retrieval database from knowledge data desired by a user; An unnecessary database is created from unnecessary knowledge data, and the knowledge data in the search database, the knowledge data associated with the knowledge data in the search database, and the association between the knowledge data, the knowledge comprising knowledge data and the association between the knowledge data. Extracting from the database, creating the search result database with the extracted knowledge data and its association, the knowledge data in the unnecessary database, the knowledge data associated with the knowledge data in the unnecessary database and the association thereof Delete from the search result database and search Knowledge data search method characterized by updating the results database, and outputs the updated search result database. 3. A procedure for storing knowledge data desired by a user in a search database, a procedure for storing unnecessary knowledge data in an unnecessary database, a knowledge data in the search database, and a knowledge data in the search database. Extracting the knowledge data and its association from the knowledge database composed of the knowledge data and the association between the knowledge data, and storing the extracted knowledge data and the association in a search result database; Updating the search result database by deleting the knowledge data, the knowledge data associated with the knowledge data in the unnecessary database, and the association from the search result database, and outputting the updated search result database on the output device. Run on computer and search for knowledge data Computer-readable recording medium recording a program to be executed.