JPH0934774A - Information retrieving device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To exactly select data to be referred to according to the retrieval intension of an operator by providing relational intensity by routes to reference source data. SOLUTION: For example, the box of an X company 34 can be reached from any respective boxes of a printer company 31 and a copy company 32. Even when the number of times to refer to the reference data (such as the facsimile) of the X company 34 is almost equal, difference is sometimes generated in the number of times of read when the X company 34 is reached from the printer company 31 and when the X company 34 is reached from the copy company 32. Namely, frequency for the operator, who selects the X company 34 from the box of the printer company 31 to select the box of a printer in the data to be referred to of the X company 34 is high and possibility for a person, who reaches the box of the X company 34 from the box of the copy company 32, to refer to the box of a copy machine is high. Then, since the relational intensity is provided by the routes to reach the reference source data, the data to be referred to according to the intension of the operator can be exactly selected.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an information retrieving apparatus for retrieving desired data from a plurality of stored data, and in particular, based on the strength of relation indicating the strength of relation between data. The present invention relates to an information search device that advances a search.

[0002]

2. Description of the Related Art In information processing apparatuses such as computers and workstations, in order to manage a large amount of data and information efficiently, these are often stored in a database. The database has a hierarchical structure of data,
That is, there is a tree structure. Each node of the tree structure is configured with a group of data items that are strongly related to each other as one unit. Also, there is a parent-child relationship between a node and its subordinate nodes. One parent node can have multiple child nodes, but conversely a child node cannot have multiple parent nodes. In such a tree-structured database, it is difficult to appropriately classify the data and hierarchically structure it, and to add or change nodes. Usually, in the database, the keyword is registered in advance and the data is searched based on this. When searching, a combination of a plurality of keywords is set according to the AND condition or the OR condition, and the target data is searched. However, if an appropriate keyword cannot be set for the data to be searched, the search cannot proceed efficiently.

Therefore, in recent years, a database called a hypertext, which searches for target data based on the relation between the data, has been drawing attention. Hypertext is a so-called mesh type database. Like a tree structure, a net-type database is generally hierarchized by parent-child relationships, but one child node can have multiple parent nodes. . Therefore, the links between the nodes can be configured in a mesh. In hypertext, data is subdivided into small units that a user can understand. The individual pieces of data that have been subdivided are called boxes. In hypertext, the link status between boxes is managed individually. It also manages the strength of the association between boxes.

In hypertext, one box is used as a starting point, and boxes linked to the box are sequentially referred to, so that desired information can be retrieved without knowing an appropriate keyword. . Hereinafter, the box that becomes the reference source will be referred to as reference source data, and the box that is referred to by the reference source data will be referred to as referenced data. Regarding such an information retrieval apparatus, Japanese Patent Laid-Open No. 5-31
It is disclosed in Japanese Patent No. 3986.

FIG. 8 shows an example of a linked state of data managed by a hypertext which has been conventionally used. In this example, information for company guidance is managed as a database. The information is divided into units that can be identified by the user. Here, the information for company guidance is classified by industry, company, and product for each company. Each rectangle in the figure represents a box after dividing the information.

[0006] Boxes of a business type include a box 101 of a printer company and a box 102 of a copy company.
When the box 101 of the printer company is used as the reference source data, the reference data linked to this is company A, X
There are boxes 103 to 105 of company C and company C. Further, when the box 102 of the copy company is used as the reference source data, the referenced data linked to this is X company, C company, D
There are boxes 104-106 of the company. Further, when the company X is the reference source data, there are boxes 107 to 109 of the facsimile, printer and copy machine as the referenced data linked to this. In the mesh-linked database, the upper box can thus refer to the plurality of lower boxes, and the lower box can also be referred to from the plurality of upper boxes.

FIG. 9 shows a data relation table for registering information for managing the relation between boxes. This figure shows the data relation table when the box of company X is the reference source. The data relationship table is individually prepared for each one-to-one combination of reference source data and referenced data. As shown in FIG. 8, there are three boxes referred to by Company X: a facsimile machine, a printer, and a copy machine. Therefore, three data relationship tables are created for each product that is the referenced data when the company X is the reference source data. In the data relation tables 111 to 113, a reference source data identifier 114 for identifying which box is a reference source and a referenced data identifier 115 for identifying which box is a reference destination are registered. There is. Also, the strength of the relationship 116 indicating the frequency of referring to the referenced data from the reference source data, that is, the strength of the relationship between these boxes.
Then, the related read number 117 indicating the number of times the referenced data is selected and read is registered.

FIG. 10 shows an outline of the configuration of an information retrieval device that manages data by hypertext. This information retrieval device is provided with a main control unit 121 that performs a central function of various controls. Main controller 12
1 is a CPU (central processing unit) and a CPU (not shown)
It also has a ROM (read only memory) (not shown) that stores a program that describes the processing procedure to be performed and various fixed data. Further, it has a RAM (random access memory) (not shown) for storing various data temporarily required for executing the program. The main storage unit 122 is a memory that is temporarily stored when performing a search for various data to be searched.

The data storage section 123 is a section for storing various data to be searched for storage. Here, a magnetic disk device is used as the data storage unit 123. The data relationship storage unit 124 is a memory that stores the data relationship table shown in FIG. The relationship strength update unit 125 is a circuit portion that updates the relationship strength and the relationship read count registered in the data relationship table. The search processing unit 126, based on the data relationship table stored in the data relationship storage unit 124, refers to the referenced data having a relationship strength greater than a predetermined value, that is, the referenced data that is highly related to the reference source data. This is a part for searching reference data. The input / output unit 127 has a function of inputting various instructions from an operator and a function of displaying search results and various information, and includes a keyboard, a mouse, and a display (not shown).

FIG. 11 shows a flow of processing performed when a conventionally used information retrieval apparatus retrieves data. Here, it is assumed that the information for company guidance shown in FIG. 8 is searched. Further, it is assumed that the current reference source data is the box of company X, and the referenced data that is associated with company X is searched for a highly related one. Prior to the search, the operator uses the input / output unit 12 to input a threshold value to be compared with the related strength as a criterion for selecting the referenced data.
Input from 7. Referenced data whose relationship strength registered in the data relationship table is larger than this threshold value is selected as data to be referenced. The identifier representing the current reference source data is stored in the RAM of the main control unit 121.

The retrieval processing unit 126 retrieves and reads one data relation table from the data relation storage unit 124 whose reference source data identifier is equal to the identifier of the current reference source data (step S101). The relationship strength registered in the read data relationship table is compared with the threshold value specified by the operator (step S102). When the relationship strength is equal to or greater than the threshold value (step S102; Y), the number of times of relationship reading registered in the data relationship table is increased by "1" (step S103). The box indicated by the referenced data identifier in the data relation table is highly relevant to the company X, which is the reference source, and will be displayed later. Therefore, the referenced data identifier of the data relationship table is stored in the main storage unit 122.
Is stored in a predetermined area (step S104). When the relationship strength read from the data relationship table is smaller than the threshold value (step S102; N), the data has low relevance, and the process proceeds to search for another data relationship table.

When the reading of all the data relation tables stored in the data relation storage unit 124 whose reference source data identifier matches the box identifier of the company X has not been completed (step S105; N). , Step S
Returning to 101, the next data relation table is read. When the reading of all the data relation tables in which the reference source data identifier matches the box identifier of the company X has been completed (step S105; Y), the main storage unit 122
The referenced data identifier stored in the predetermined area is read (step S106). Then, the data of the boxes corresponding to these are displayed on the display.

The relation strength updating unit 125 reads out the relation reading number registered in the data relation table at regular time intervals, and then initializes it to "0". Since the number of times of reading the relation represents the number of times that the referenced data is read within a certain period of time, if this is a predetermined value or more, the relation strength is strengthened, that is, increased. If the number of times of relational reading is less than a predetermined value, the relational strength is reduced and weakened. As a result, the more the search is performed, the more the value of the relationship strength is corrected to a value according to the frequency with which the referenced data is read.

[0014]

As described above, if the relationship strength is periodically updated based on the number of times of reading from the reference source data, the strength of the relationship becomes large according to the frequency of reading the referenced data. Therefore, it becomes possible to accurately search for the desired data. However, when the number of times of reading a plurality of referenced data linked to the reference source data is dispersed on average, the difference in the relationship strength between the referenced data becomes small. Therefore, even if the magnitude of the relationship strength and the threshold value are compared, it is impossible to select appropriate data, and there is a problem that the satisfaction of the search result is reduced.

Therefore, an object of the present invention is to provide an information retrieving apparatus capable of precisely retrieving target data even when the number of times of referring data is read is evenly distributed. It is in.

[0016]

According to a first aspect of the present invention, a reference is made to a plurality of data linked in a net-like manner in which one data can be referred to by two or more other data. Strength of the relationship between the reference relationship storage unit that stores the association between the original data and one or more referenced data to be referenced, and the referencing source data and the referenced data to be referenced For each one-to-one combination of the reference source data and the referenced data stored in the reference relationship storage means in association with each other, the reference reference data is positioned as the referenced data. Before the relation strength storage means stored for each reference source data that is the original data and the reference source data when referencing the referenced data position this as the referenced data Pre-reference source information storage means for storing the pre-reference source identification information indicating from which one of the reference source data is referred to, and the referenced data associated with this from any reference source data is referred to. Of the relationship strength between the reference source data stored in the relationship strength storage means and the referenced data associated therewith, the front reference source identification information stored in the front reference source identification information storage means The referenced data selecting means for selecting the referenced data based on the relationship strength corresponding to the preceding referencing data, and storing the referencing source identification information when the referenced data is selected by the referenced data selecting means. The information retrieving apparatus is provided with pre-reference source identification information updating means for updating the value of the pre-reference source identification information stored in the means to a value representing the reference source data when the referenced data is selected. There.

That is, according to the first aspect of the invention, the strength of relationship between the reference source data and the referenced data is prepared for each higher-order reference source data that positions the reference source data as the referenced data. Then, the strength of the relationship used when selecting the referenced data is changed depending on which of the higher-level reference source data the current reference source data has arrived. In a net-type linked database, there may be two or more routes that reach the reference source data.

For example, the company X box shown in FIG. 8 can be reached from either the printer company box or the copy company box. Even when the number of times the referenced data of the X company, that is, the facsimile, the printer, and the copy machine are referred to each other, is almost the same when the printer company box reaches the X company and the copy company reaches the X company. If classified according to time, the number of times of reading may differ. That is, an operator who selects company X from the printer company box has a high frequency of selecting the printer box from the reference data of company X, and a person who reaches the company X box from the company copy box is the copy machine. More likely to refer to the box. Therefore,
By providing the relationship strength for each route that has reached the reference source data, it is possible to more accurately select the referenced data according to the operator's search intention.

According to the second aspect of the present invention, one piece of data can be referred to by two or more other pieces of data, and a plurality of pieces of data to be searched linked in a mesh form can be referred to as reference source data and will be referred to. A reference relationship storage unit that stores a correspondence between one or two or more referenced data, and a relationship strength that represents the strength of the relationship between the reference source data and the referenced data to be referenced from now on. For each one-to-one combination of the reference source data and the referenced data whose correspondence is stored in the reference relationship storage means, the reference source data is a higher reference source data that positions the reference source data as the referenced data. Correlation between the relation strength storage means stored for each original data and the relation read number indicating the number of times the referenced data is referred to from the reference source data is associated with the reference relation storage means. For each one-to-one combination of the stored reference source data and referenced data, the relational read count storage means stores the reference source data as the referenced data for each preceding reference source data, and the referenced data. Pre-reference source information storage means for storing the pre-reference source identification information indicating from which of the pre-reference source data the reference source data when referencing is referred to as the referenced data, When referencing the referenced data associated with it from any reference source data, the relationship strength between the referencing data stored in the relationship strength storage means and the referenced data associated with it Of the referenced data based on the relationship strength corresponding to the previous reference source data indicated by the previous reference source identification information stored in the previous reference source identification information storage means Referenced data selection means to be performed, and when the referenced data is selected by the referenced data selection means, the previous reference of the relation read times assigned to the combination of the reference source data and the selected referenced data The referenced data is selected by the referenced data selection means and the related read number increasing means for increasing the related read number corresponding to the previous reference source data indicated by the previous reference source identification information stored in the original identification information storage means. When the value of the front reference source identification information stored in the front reference source identification information storage means is updated to a value representing the reference source data when the referenced data is selected, It is associated with the same pre-reference source data, reference source data, and referenced data as the associated pre-reference source data, reference source data, and referenced data. The information retrieval apparatus is provided with relationship strength updating means for strengthening or weakening each relationship strength based on the value of the number of times of relationship reading.

That is, according to the second aspect of the present invention, the number of times the referenced data is read is registered for each route that reaches the reference source data. Then, based on this number of times, the corresponding relationship strength is strengthened or weakened. Therefore, it is possible to appropriately update the value of the relationship strength provided for each route that reaches the reference source data.

According to the third aspect of the present invention, one piece of data can be referred to by two or more other pieces of data, and a plurality of pieces of data to be searched linked in a mesh form are referred to as reference source data and will be referred to. A reference relationship storage unit that stores a correspondence between one or two or more referenced data, and a relationship strength that represents the strength of the relationship between the reference source data and the referenced data to be referenced from now on. For each one-to-one combination of the reference source data and the referenced data whose correspondence is stored in the reference relationship storage means, the reference source data is a higher reference source data that positions the reference source data as the referenced data. The relationship strength storage means for storing each original data, and the reference source data for referencing the referenced data, the referencing source data positioning this as the referenced data. A reference source information storage unit for storing the reference source identification information indicating from which one of the reference source data is referred to, and the reference when referring to the referenced data associated with this from any reference source data. As the strength of the relationship between the original data and the referenced data associated therewith, the strength of the relationship is the strength of the relationship corresponding to the previous reference source data indicated by the previous reference source identification information stored in the previous reference source information storage unit. Initial relationship strength registration means for registering a predetermined initial value as the relationship strength when the initial value is registered by the initial relationship strength registration means when the initial value is not stored in the storage means, and is referred to based on that value. When data is selected and the referenced data associated with it is referenced from any reference source data, between the reference source data and the referenced data associated with it When the relationship strength corresponding to the previous reference source data indicated by the previous reference source identification information stored in the previous reference source information storage means as the relationship strength is stored in the relationship strength storage means, the relationship strength is calculated based on the relationship strength. Referenced data selection means for selecting reference data, and referred to the value of the previous reference source identification information stored in the previous reference source identification information storage means when the referred data is selected by this referenced data selection means The information retrieving apparatus is provided with pre-reference source identification information updating means for updating the data to a value representing the reference source data when the data is selected.

That is, according to the third aspect of the present invention, when selecting the referenced source data from the referencing source data, when the relationship strength corresponding to the route reaching the referencing source data is not registered, a predetermined initial value is set. Is registered in the relationship strength storage means. When setting a new link between the referencing source data and the referenced source data, it is not necessary to register the relationship strength for the referenced data subordinate to the referenced source data, so it is easy to change the link setting. Can be done.

[0023]

BEST MODE FOR CARRYING OUT THE INVENTION The present invention will be described in detail below with reference to Examples.

FIG. 1 shows the outline of the configuration of an information retrieval apparatus according to an embodiment of the present invention. The information retrieval device includes a CPU (central processing unit) 11 that performs a central function of various data processing. Various circuit devices are connected to the CPU 11 through a bus 12 such as a data bus. Of these, ROM (Read Only Memory)
Reference numeral 13 is a read-only memory for storing a program representing a data processing procedure and various fixed data.
A RAM (Random Access Memory) 14 is a memory that temporarily stores data necessary for executing a program. The magnetic disk device 15 is a storage device for storing data to be searched. The display unit 16 displays various data in addition to the searched data. Here, a CRT display is used as the display unit 16. The keyboard 17 and the mouse 18 are input devices for inputting operator's instructions.

FIG. 2 functionally shows the configuration of the information retrieval apparatus shown in FIG. The information search device includes a main control unit 21 that performs a central function of various controls.
The main storage unit 22 is a memory for temporarily storing various data to be searched after being read from the magnetic disk device. The data storage unit 23 is a unit that stores various data to be searched for storage. The data relationship storage unit 24 is a memory that stores a data relationship table that indicates whether or not there is a relationship between data and its strength. The relationship strength update unit 25 is a circuit unit that updates the relationship strength and the relationship read count registered in the data relationship table. The search processing unit 26 is a unit that searches for data that is highly related to the reference source data, based on the data relationship table stored in the data relationship storage unit 24. The input / output unit 27 includes a keyboard and a mouse as input devices and a display as an output device.

The initial relationship strength giving section 28 is a circuit section for adding a new item to the data relationship table. When the reference history holding unit 29 selects one referenced data and advances the search as this new reference source data, which reference source data in which this new reference source data is arranged above it is selected. It is a part that holds information indicating whether or not it has been referred to from.

FIG. 3 shows an example of a link state of data to be searched by the information search device. This is information for company guidance, and is classified by industry, company, and product for each company. There are printer companies and copy companies in the industries, and boxes 31 and 3 are provided for each.
2 are provided. As a printer company, Company A 3
3, X company 34, and C company 35 exist, and copy companies include X company 34, C company 35, and D company 36. Company X 34 and company C 35 are both printer companies and copy companies. In the figure, the product-specific boxes are shown only for company X, and the description of the product-specific boxes for companies A, C, and D is omitted. Company X's products include fax machines, printers, and copy machines.
7 to 39 are provided.

These boxes have relevance in the direction of the arrow shown by the solid line, and accordingly, the search can be advanced from the top to the bottom in the figure. The box selected on the basis of the relevance is referred to as referenced data, and the box from which selection is made is referred to as reference source data.
For example, when a company is selected based on the printer company box 31, the printer company box 31 becomes the reference source data. And Company A 33, Company X 34, Company C 3
The box 5 is the referenced data. These boxes are given an identifier for identifying each other. The information for company guidance is divided into boxes and stored in the data storage unit 23. The information retrieving apparatus stores information indicating a relation between data called a data relation table for each arrow in order to store the connection between the data represented by the solid lines in FIG. Based on the data relation table, the related data are sequentially searched to find the target data.

FIG. 4 shows an example of registered contents of the data relation table. This represents a data relation table when the box of company X shown in FIG. 3 is used as the reference source data. One data relationship table represents a one-to-one relationship between the reference source data and the referenced data. Therefore, when the company X is used as the reference source, the data relationship tables 41, 42, and 43 are created for each of the facsimile, printer, and copy machine. Since the data structures of the data relation tables 41 to 43 are the same, the data relation table 41 will be described here.
Descriptions of other elements are omitted as appropriate.

In the data relation table 41, a reference source data identifier 44 for indicating the reference source data and a referenced data identifier 45 for indicating the referenced data are registered. The identifiers assigned to the corresponding boxes are registered in this area. Although an identification number is actually used as an identifier, the contents of each box are shown as an identifier in the figure for convenience of explanation. For example, the data relationship table 41 represents the relationship corresponding to the arrow 61 in FIG. 3, and company X is registered as the reference source data identifier, and facsimile is registered as the referenced data identifier.

In the data relation table, the relation strength indicating the relation with the reference source data and the relation reading number indicating the number of times the referenced data are referred to are registered. These are prepared for each box in which the box indicated by the reference source data identifier is positioned as the referenced data. A box in which the box indicated by the reference source data identifier is positioned as the referenced data is referred to as previous reference source data. For example, in the data relation table 41, the company X shown in FIG. 3 is the reference source data, so the box 31 of the printer company and the box 32 of the copy company are the respective reference source data.

The pre-reference source data identifier 46 for identifying the pre-reference source data is registered in the data relation table. In addition, a relationship strength 47 indicating the strength of the relationship between the previous reference source data and the referenced data, and a relationship read number 48 indicating the number of times the referenced data is read are registered. In the data relation table 41, for one piece of front reference source data, the front reference source data identifier 46 and the relation strength 47
Also, the item of the relation read count 48 is registered as a set. This will be called a related item by reference source. In the data relationship table 41, the relationship item 51 by the previous reference source when the previous reference source data is the printer company and the relationship item 52 by the previous reference source when the previous reference source data is the copy company are registered.

Relationship strength in relationship item 51 by reference source
47₁Reached Company X from Box 31 of the printer company
Strength of the relationship between Company X and the facsimile
Is represented. Also, in the related item by reference source 51
Relationship read count 48₁Is the printer company box 31?
Refer to the box 37 of the facsimile after arriving at Company X
It represents the number of times that it was done. Similarly, related terms by reference source
Relationship strength 47 of eye 52 _TwoAnd related read count 48_TwoIs
When you reach company X from the copy company box
The relationship strength and the number of times of reading are shown. Ma
In addition, when adding a new item by reference source,
The initial relationship strength 53, which is the initial value to be set as the degree, is registered.
Have been.

Now, the flow of processing performed when the information search apparatus searches for the target data using the data relation table shown in FIG. 4 will be described.

FIG. 5 shows the flow of a search process performed by the information search device. Here, a case of searching the information for company guidance shown in FIG. 3 will be described as an example. Further, the current reference source data is the box 34 of company X, and it is assumed that the data to be referenced is selected from the referenced data of the facsimile, printer, and copy machine. The data relation table when the company X is the reference source is the same as that shown in FIG. Further, the previous reference source data when it arrives at the box of company X is assumed to be the printer company.
In this case, the reference history holding unit 29 shown in FIG. 2 stores that the previous reference source data is the printer company. It is assumed that the threshold value serving as a reference when selecting the data to be referred to according to the strength of the relation of the data is input by the operator with the keyboard or the mouse.

When an instruction to select the data to be referred to next with the box of company X as the reference source data is input from the keyboard or mouse, the CPU 11 selects from the data relation table stored in the data relation storage unit 24. 1 when company X is registered as a reference source data identifier
One is retrieved and read (step S201). Next, CP
The U11 checks whether or not the previous reference source data identifier same as the previous reference data identifier held by the reference history holding unit is registered in the read data relation table (step S202). When the matching previous reference source data identifier is registered (step S202; Y), the relationship strength for the corresponding previous reference source data is read (step S203). The case where the matching previous reference source data identifier is not registered will be described later.

For example, if the data relation table 41 shown in FIG. 4 is read in step S201, "10" is read as the relation strength. The CPU 11 compares the read relationship strength with a previously input threshold value. If the relationship strength is equal to the threshold value or is higher than the threshold value (step S204; Y), the value of the referenced data identifier registered in this data relationship table is stored in a predetermined area of the RAM 14 ( Step S205). For example, assuming that the threshold value input by the operator is "40", the referenced data identifier "facsimile" in the data relation table 41 is
Not stored in RAM 14.

Next, the CPU 11 checks whether or not there is another data relation table in which company X is registered as a reference source data identifier (step S206). As shown in FIG. 4, there are three data relationship tables whose company X is the reference source data. When reading is not performed for all of these (step S206; N), step S
Returning to step 201, processing for checking the relationship strength of each data relationship table is performed. Since the previous reference source data is the printer company, in the case of the data relation table 42 in which the referenced data is the printer, the relation strength is “8”.
0 "is compared with a threshold value. In the case of the data relation table 43 in which the referenced data is a copy machine," 20 "is compared with the threshold value as the relation strength.
In a predetermined area of 14, the referenced data identifier “printer” of the data relationship table 42 having the relationship strength “80” is registered.

When the processing has been completed for all the data relation tables whose reference source data is company X (step S206; Y), the data corresponding to the referenced data identifier stored in the RAM 14 is set. CR of display 16
It is displayed on the T display (step S207). Based on this display, the operator specifies his or her desired data with the mouse (step S208). As a result, the data desired by the operator is selected from the three referenced data that can be referred to by company X. Therefore, the CPU 11 increases the value of the number of times of relational reading associated with the corresponding previous reference data in the data relational table corresponding to the selected referenced data by 1 (step S209). Here, since the referenced data is the printer and the previous reference source data identifier is the printer company, the number of relational readings registered in the column 54 of the data relational table 42 is changed from "15" to "16".

When referring to data in the next stage, the box 38 of the printer becomes the reference source data. And
Box 34 of company X becomes the previous reference source data. Therefore, the CPU 11 changes the value of the previous reference source data identifier held by the reference history holding unit 29 to the identifier of company X (step S210). Finally, the CPU 11 displays information about the printer, which is the selected box, on the display (step S211), and ends the process. Although not shown in FIG. 3, when an instruction to further refer to the data related to the printer box 38 is input, the printer box is used as the reference source data and the company X is used as the previous reference source data. The search will be performed according to the flow shown in.

Although not shown in the flow chart of FIG. 5, the CPU
Reference numeral 11 checks the number of times of relational reading at regular time intervals, and if this is greater than or equal to a constant value, increases the value of the corresponding relational strength and strengthens it. On the contrary, when the number of times of relational reading is equal to or less than a certain value, the value of the corresponding relational strength is reduced and weakened. Then, after checking the number of times of relational reading, this is initialized to "0". In this way, by changing the relationship strength according to the number of times of reading for each fixed time, the strength of the relationship between the reference source data and the referenced data becomes an appropriate value according to the frequency of reading.

An operator who selects company X next to the printer company and proceeds with data retrieval is highly likely to be interested in information about the printer among products of company X. However, the operator who selects company X next to the copy company is often interested in the copy machine among the products of company X. Therefore, there is a high probability that the search will proceed according to the route shown by the dotted line in FIG. That is, the degree of association between company X and its products depends on the route to company X.

As shown in FIG. 4, when the referenced data is the data relation table 42 of the printer, the relation strength when the pre-reference source data is the printer company is as large as "80", and the pre-reference source is the copy company. In case of, the relationship strength is "30"
And relatively small. On the contrary, in the data relationship table 43 in which the referenced data is the copy machine, the relationship strength is as small as “20” when the previous reference source data is the printer company, and the relationship strength is “20” when the previous reference source data is the copy company. It is as large as 90 ". In this way, since the relationship strength is managed separately for each route reaching the company X, it is possible to more accurately retrieve data that matches the operator's purpose. Furthermore, the relationship strength prepared for each previous reference source data is updated at regular time intervals based on the number of times of relationship reading, so as the number of searches increases, an appropriate relationship strength according to the route reaching company X can be obtained. Obtainable.

Next, a case where the item of the previous reference source data is not registered in the data relation table in step S202 of FIG. 5 will be described.

FIG. 6 shows a data link state in the company guide information when a facsimile company is added to the industry. The box 71 of the facsimile company added to the industry hierarchy has a link with company X as indicated by the dotted line 72. Now, it is assumed that company X is referred to for the first time by the facsimile company based on this link. At this time, company X becomes the reference source data, and facsimile company 71 becomes the previous reference source data. The reference history holding unit holds the identifier assigned to the box 71 of the facsimile company. Further, it is assumed that the data relationship table having company X as the reference source data is in the state shown in FIG.

In step S201 of FIG. 5, the data relation table 41 shown in FIG. 4 is read, and the CPU 11 checks whether or not the facsimile company is registered as the previous reference source data identifier (step S202). This is the first time that a facsimile company has reached company X, and since the previous reference source data does not yet have an item for the facsimile company in the data relation table (step S202; N), the CPU
The process 11 adds the item to the data relation table (step S212).

FIG. 7 shows a data relation table in which a facsimile company is added as the previous reference source data. The same parts as those in FIG. 4 are designated by the same reference numerals. This is a part of the data relation table 41 to which the relation item 81 for each reference source is added. A facsimile company is copied in the previous reference source data identifier 82, and a value of the initial relationship strength 53 is copied and stored in the relationship strength 83. The relation read count 84 is set to "0" as an initial value. Also in the data relationship tables 42 and 43, relationship items by front reference source 85 and 86 in which the front reference source data identifier is a facsimile company are added, respectively.

Returning to FIG. 5, the description will be continued. After adding a new item to the data relationship table, the initial relationship strength is read (step S213), and this value is compared with the threshold value in step S204. Subsequent processing is similar to the case where the related item by reference source is not added. By performing the above processing also on the data relation tables 42 and 43 shown in FIG. 4, the tables are updated to the state shown in FIG.

As described above, when the newly added facsimile company refers to company X, the data relation table with reference to company X is updated. It doesn't get complicated. That is, when additionally registering the box of the facsimile company in the industry, it is only necessary to create the data relation table having the facsimile company as the reference source data.

In the embodiment described above, the relationship strength is updated, but the relationship strength does not have to be updated as long as an appropriate value can be set from the initial setting time. Further, although the reference source data and the referenced data are in one-to-one correspondence and each has an independent data relationship table, these do not have to exist independently. For example, 1
All the referenced data referenced from one referencing data may be put together to form one data relation table.

[0051]

As described above, according to the first aspect of the present invention, since the relationship strength is provided for each route reaching the reference source data, it is possible to select the referenced data according to the operator's search intention.

According to the second aspect of the present invention, the number of times the referenced data is read is registered for each route that has reached the reference source data, and based on this, the corresponding strength of the relationship is strengthened or weakened. Since this is done, the value of each relationship strength can be updated appropriately. As a result, as the number of searches is increased, it is possible to accurately search the referenced data in accordance with the operator's search intention.

Further, according to the invention described in claim 3, when the referred source data is selected from the referred source data, the case where the relationship strength corresponding to the route reaching the referred source data is not registered is dealt with. A predetermined initial value is registered in the relationship strength storage means as the relationship strength. Therefore, it is possible to construct or change the mesh link structure without increasing the burden on the operator.

[Brief description of drawings]

FIG. 1 is a block diagram showing an outline of a configuration of an information search device according to an embodiment of the present invention.

FIG. 2 is an explanatory diagram functionally showing the configuration of the information search device shown in FIG.

FIG. 3 is an explanatory diagram showing an example of a link state of data to be searched by the information search device.

FIG. 4 is an explanatory diagram showing an example of registered contents of a data relation table.

FIG. 5 is a flowchart showing a flow of search processing performed by the information search device.

FIG. 6 is an explanatory diagram showing a linked state of data in company guide information when a facsimile company is added to an industry.

FIG. 7 is an explanatory diagram showing an example of registered contents of a data relation table when a facsimile company is added.

FIG. 8 is an explanatory diagram showing an example of a linked state of data managed by hypertext that has been conventionally used.

FIG. 9 is an explanatory diagram showing an example of registered contents of a data relation table in a conventionally used information search device.

FIG. 10 is a block diagram showing an outline of a configuration of an information search device that has been conventionally used.

FIG. 11 is an explanatory diagram showing a flow of processing performed when a conventionally used information search device searches for data.

[Explanation of Codes] 11 ... CPU, 12 ... Bus, 13 ... ROM, 14 ... RA
M, 15 ... Magnetic disk device, 16 ... Display unit, 17 ... Keyboard, 18 ... Mouse, 21 ... Main control unit, 22 ... Main storage unit, 23 ... Data storage unit, 24 ... Data relation storage unit,
25 ... Relationship strength updating unit, 26 ... Search processing unit, 27 ... Input / output unit, 28 ... Initial relationship strength giving unit, 29 ... Reference history holding unit, 41-43 ... Data relationship table, 44 ... Reference source data identifier, 45 ... Referenced data identifier, 46 ... Previous reference source data identifier, 47 ... Relationship strength, 48 ... Relationship read number, 53 ... Initial relationship strength

Claims (3)

[Claims] 1. A reference source data which is a reference source for a plurality of data to be searched linked in a mesh form in which one data can be referred to by other two or more data, and one or more reference data In the reference relationship storage means, a reference relationship storage means that stores the association with the referenced data, and a relationship strength that represents the strength of the relationship between the reference source data and the referenced data that is to be referred to in the reference relationship storage means. For each one-to-one combination of the reference source data and the referenced data in which the association is stored, the reference source data is stored for each previous reference source data that is the higher-level reference source data that is positioned as the referenced data. The relation strength storage means and the reference source data used when referring to the referenced data are referred to from any of the previous reference source data that positions this as the referenced data. Front reference source information storage means for storing front reference source identification information indicating whether or not the reference reference data associated with the reference reference data is stored in the relationship strength storage means. Corresponding to the pre-reference source data indicated by the pre-reference source identification information stored in the pre-reference source identification information storage means in the relationship strength between the reference source data and the referenced data associated therewith. The referenced data selecting means for selecting the referenced data based on the relationship strength, and when the referenced data is selected by the referenced data selecting means, it is stored in the previous reference source identification information storage means. An information retrieving apparatus, comprising: pre-reference source identification information updating means for updating the value of the reference source identification information to a value representing the reference source data when the referenced data is selected. 2. One reference data, which is a reference source for a plurality of data to be searched, which are linked in a mesh form, in which one data can be referred to by two or more other data, and one or more reference data In the reference relationship storage means, a reference relationship storage means that stores the association with the referenced data, and a relationship strength that represents the strength of the relationship between the reference source data and the referenced data that is to be referred to in the reference relationship storage means. For each one-to-one combination of the reference source data and the referenced data in which the association is stored, the reference source data is stored for each previous reference source data that is the higher-level reference source data that is positioned as the referenced data. The relation strength storage means and the relation read number representing the number of times the referenced data are referred to from the reference source data are stored in the reference relation storage means in association with each other. For each one-to-one combination of the reference source data and the referenced data, the related read number storage means storing the reference source data as the referenced data for each preceding reference source data, and referring to the referenced data. Pre-reference source information storage means for storing the pre-reference source identification information indicating from which of the pre-reference source data that the reference source data of the above is positioned as the referenced data, and any reference source When referring to the referenced data associated with it from the data, the relationship strength between the reference source data stored in the relationship strength storage means and the referenced data associated with the referenced data is The referenced data is selected based on the relationship strength corresponding to the previous reference source data indicated by the previous reference source identification information stored in the previous reference source identification information storage means. The referenced data selecting means, and when the referenced data is selected by the referenced data selecting means, the above-mentioned pre-reference among the relational read times assigned to the combination of the reference source data and the selected referenced data Relationship reading number increasing means for increasing the relationship reading number corresponding to the pre-reference source data indicated by the pre-reference source identification information stored in the source identification information storage means, and the referenced data is selected by the referenced data selecting means. And the pre-reference source identification information updating means for updating the value of the pre-reference source identification information stored in the pre-reference source identification information storage means to a value representing the reference source data when the referenced data is selected. Pre-reference source data, reference source data, and referenced data that have the same strength are paired with the pre-reference source data, referenced source data, and referenced data. Information retrieval apparatus characterized by comprising a relationship intensity updating means strengthening or weakening the relationship strength by the value of the relationship read times that are attached to the group. 3. One reference data, which is a reference source for a plurality of data to be searched linked in a mesh form, in which one data can be referred to by two or more other data, and one or two or more data to be referred to. In the reference relationship storage means, a reference relationship storage means that stores the association with the referenced data, and a relationship strength that represents the strength of the relationship between the reference source data and the referenced data that is to be referred to in the reference relationship storage means. For each one-to-one combination of the reference source data and the referenced data in which the association is stored, the reference source data is stored for each previous reference source data that is the higher-level reference source data that is positioned as the referenced data. Relationship strength storage means for referencing the referenced data, and the referencing data used when referencing the referenced data from the previous referencing data that positions this as the referenced data. Pre-reference source information storage means for storing the pre-reference source identification information indicating whether the reference source data is associated with the referenced data associated with it from any reference source data. The relationship strength corresponding to the previous reference source data indicated by the previous reference source identification information stored in the previous reference source information storage means is stored in the relationship strength storage means as the relation strength with the attached referenced data. If not, the initial relationship strength registration means for registering a predetermined initial value as the relationship strength, and when the initial value is registered by this initial relationship strength registration means, the referenced data is selected based on the value. When referring to the referenced data associated with it from any reference source data, the relationship strength between the referencing data and the referenced data associated with it When the relationship strength corresponding to the previous reference source data indicated by the previous reference source identification information stored in the previous reference source information storage means is stored in the relationship strength storage means, the relationship strength is calculated based on the relationship strength. Referenced data selection means for selecting reference data, and a value of the previous reference source identification information stored in the previous reference source identification information storage means when the referenced data is selected by the referred data selection means. An information retrieving apparatus comprising: pre-reference source identification information updating means for updating the reference source data when the reference data is selected.