JPH0785093A - Method for automatically setting index - Google Patents

Abstract

PURPOSE:To automatically prepare index to be frequently used by recognizing the access state of index to be used by a user. CONSTITUTION:The frequency to be used of the index which is registered in an index registration part 21 which is used for retrieving terminal equipments A and B is calculated. A changeover discrimination part 24 selects the index to be more frequently used than the frequency to be used which is set in advance from among the calculated frequency to be used of each index. A changeover part 26 updates and registers the selected index instead of the index registered in the index registration part 21.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an index automatic setting method for automatically updating an index for extracting arbitrary data from relational data at high speed.

[0002]

2. Description of the Related Art Conventionally, in a production line such as an automobile, data required for each line has been extracted from a database. This extraction is performed by relying on an index attached to the data. ing.

In the case of relational data, since each data is indexed, there is an advantage that the data required by the user can be freely extracted by designating the index. On the other hand, if the amount of data is huge and the number of indexes is very large, the specified index must be searched from all the indexes, which causes a problem that it takes a very long time to extract the data. I will end up.

[0006] Since this would impair the merit of constructing relational data, recently, some indexes that are often used for searching have been selected from the indexes attached to the data, and this index is used. I try to extract the data.

[0005]

However, in such a conventional method, not all of the selected indexes are used effectively, and the conventional method is very good because the data structure is changed. It may happen that the used index becomes unnecessary. In such a case, it is necessary to newly register an index that is often used when extracting data, or delete an unnecessary index. If this work is neglected, the significance of providing an index for high-speed search will be lost.

Since an automobile production factory has relational data for each work type, the number of relational data and indexes for the entire factory is enormous, and which index is currently most frequently used? Although it is theoretically possible to work by grasping each relational database and each production line and always selecting the optimum index,
It is actually impossible considering the economics.

The present invention has been made in view of the above-mentioned conventional problems, and makes it possible to automatically prepare a frequently used index by grasping the access status of the index used by the user. The purpose is to provide an automatic index setting method.

[0008]

According to the present invention for achieving the above object, any one of a plurality of items attached to relational data is registered as an index to be used for a search, and the index is used for the search. An index automatic setting method used in a search method that enables a high-speed search by using, calculating a frequency of use of each index used in a terminal device for search, and calculating the frequency of use of each index. Among them, an index having a higher frequency of use than the preset frequency of use is selected, and the selected index is replaced with the registered index for renewal registration.

[0009]

According to the present invention thus constructed, the use frequency of each index used in the terminal device for retrieval is calculated, and the use frequency is preset among the calculated use frequencies of the respective indexes. Since an index with a frequency higher than the selected usage frequency is selected and the selected index is replaced with the registered index for renewal registration, it is possible to always register the optimum index without any human intervention. become able to.

Further, if the index to be registered first is selected for the time being, it is likely that it will be used, and if this is registered, only the optimum index will be gradually registered by the subsequent automatic setting of the index. Therefore, it is not necessary to select the index because the initial setting becomes nervous.

[0011]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram showing a schematic configuration of an apparatus that realizes an index automatic setting method according to the present invention.

Terminals A and B in the figure are for searching and extracting relational data stored in the database 10, and are arranged at appropriate positions in the factory. The database 10 and each terminal are connected by a server 20. Server 2
0 has a role of extracting necessary data from the database 10 in response to a request from each terminal and sending the data to the requesting terminal.

The server 20 is provided with the functions necessary for transmitting and receiving data, and FIG. 1 shows only the functions for implementing the method of the present invention. The index registration unit 21 in the figure is a part for registering, as an index, an item that is most likely to be used for a search among the items of the relational data stored in the database 10. Therefore, when searching from the terminal, in principle, one of the registered indexes is designated.
The index counter 22 includes an index registration unit 21.
Among the items registered as the index and the items not registered but designated as the index from the terminal,
It is a part that counts how many times each terminal is specified.

The usage frequency file 23 is a file for storing the usage frequency of each index or item. The switching determination unit 24 causes the usage frequency file 23 to follow the switching command output from the timer 25 at regular intervals.
This is a part of looking at the frequency of use stored in, and selecting an item having a frequency of use of not less than a predetermined percentage as a new index. In addition, the switching determination unit 24 also has a function of automatically inquiring of the user whether the index should be deleted when there is an index to be deleted from the index registration unit 21. Switching unit 2
Reference numeral 6 allows the index registration unit 21 to register the item selected by the switching determination unit 24.

The above is a general description of the apparatus for carrying out the method of the present invention. The structure of data stored in the database 10 will be described in more detail. Although the relational data is stored in the database 10, the relational data has a structure in which the data is organized according to each item. Therefore, for example, in the engine No. item, engine numbers are arranged in order from No. 1 to N, and in the machining date item, the machining date of each engine is arranged in order. Further, in the item of shipping destination, shipping destinations of the respective engines are arranged in order. Usually, dozens or hundreds of such items are provided, and the relational data has a structure like a table in which items are columns and data are rows. Therefore, if the engine number and the processing date are specified as the index, only the table including the engine number and the processing date is extracted from the tables forming the relational data.

The usage frequency file 23 includes a search item usage frequency file and a search item usage frequency cumulative file. As shown in FIG. 2, the search item usage frequency file is a collection of the frequencies of the items used in the search for each line for a certain period of time, and the search item usage frequency cumulative file is shown in FIG. As shown in, the cumulative total of the search item usage frequency files is aggregated.

The switching judgment section 24 is preset with conditions for automatically selecting an index and conditions for inquiring the user whether the index should be deleted when it becomes necessary to delete the index. There is.

This condition is defined as "index automatic conversion condition", for example, as follows.

Number of indexes: 6 Index usage frequency condition: 40% or more Index conversion error output condition: 40% or more Function of each part of the apparatus that realizes the method of the present invention, data structure stored in the database, and file structure The structure is as described above.

Next, the index automatic setting method of the present invention will be described in detail with reference to the flowcharts of FIGS. It is assumed that the following flowchart is processed after the index registration unit 21 index is registered.

First, when the process of the flowchart is started, both files stored in the usage frequency file 23 as shown in FIGS. 2 and 3, that is, the search item usage frequency file and the search item usage frequency cumulative file. All the usage frequencies of are set to 0 (S1). Then, the index counter 22 counts the search items used at the time of a search request from the terminal A or the terminal B, and sequentially records the count result in the search item use frequency file and the search item use frequency cumulative file (S2). The above counting is continuously performed until the timer 25 times out and a switching command is output (S3).

Upon receiving the switching command output from the timer 25, the switching judgment unit 24 indexes each item of the search item usage frequency file shown in FIG. It is calculated whether or not it is used as, and a table as shown in FIG. 6 is prepared. For example, when the search item use frequency file shown in FIG. 2 is obtained, the total number of times the database 10 is searched by the terminal A installed on the A line is 105 times (the timer 25 Within the set time), among these searches, "part number" is used as an index 15 times, "engine NO" is used 85 times, and "shipping destination" is used 5 times. The switching determination unit 24 will calculate the frequency of use of each index by the above numbers. in this case,
Use frequency of "part number" is about 15%, "engine NO"
Is used about 85%, and "ship-to" is used about 5%
Is calculated. This usage frequency is calculated for every line. (S4). When this calculation is completed, the switching determination unit 24 next refers to the above-mentioned "index automatic conversion condition" and selects an item having a usage frequency of not less than%.

Specifically, the "index automatic conversion condition", the number of indexes: 6, the index usage frequency condition: 40% or more, the index conversion error output condition: 40% or more, is applied to the above calculation result.

When the use frequency as shown in FIG. 6 is obtained, there are three A, D, and E that satisfy 40% or more of the index use frequency conditions, so these three are selected as indexes. It will be. Further, in the case of FIG. 2, the calculation result of the usage frequency shows that the usage frequency of "part number" is about 15%, and the usage frequency of "engine NO" is about 8%.
Since the usage frequency of 5% and "shipping destination" was about 5%, only the item "engine NO" satisfies 40% or more of the index usage frequency condition (S5).

Then, the number of set indexes is selected from the higher frequencies of use of the selected items. This processing is also performed by the switching determination unit 24. In the example based on FIG. 6, there are three A, D, and E that satisfy 40% or more of the index usage frequency conditions, so all three are selected as indexes. This is shown in FIG. That is, six items A, B, C, D, G, and H were selected as indexes in the previous index selection, but three items A, D, and E were indexed in this index selection. Will be selected as

In the example of FIG. 2, since only the item of "engine NO" satisfies 40% or more of the index use frequency condition, only this one is selected as the index, For example, if there are 7 items that satisfy the above conditions, the number of indexes is set to 5 in the "index automatic conversion condition", so 6 of these 7 items that are used most frequently are 6 items. Will be selected as an index (S6).

Next, the switching judgment unit 24 performs a process of integrating all the search counts of the search item use frequency file and the numbers of the search items into the search item use frequency cumulative file shown in FIG. For example, since the number of searches on line A up to now is 1012 as shown in FIG. 3, 105 times in the search item usage frequency file are added to this, and the result is 1117. Such integration is performed for all items for all lines. In this way, a new search item use frequency cumulative file is created (S7).

Upon completion of this processing, this time, it is selected whether or not the index that was the previous index is removed from the index this time. This processing is also performed by the switching determination unit 24. In the example of FIG. 7, the items of B, C, G, and H correspond (S8).

For these items selected in S8, the switching determination section 24 refers to the search item use frequency cumulative file in FIG. 3 to calculate the cumulative use frequency for each selected item. The calculated cumulative usage frequency is set as the "index automatic conversion condition",
It is checked for each item whether it is 40% or more of the index conversion error output condition (S9). As a result of this check, the index conversion error output condition 4
If there is 0% or more, it is displayed on the display device so that the administrator can confirm whether or not it can be deleted from the index (S10, S11). As a result of this confirmation, if an instruction to delete may be output, the old index is rewritten to the new index based on the instruction, and the rewritten index is registered in the index registration unit 21. This processing is performed by the switching unit 26. If the item to be deleted does not have the usage frequency equal to or higher than the index conversion error output condition, rewriting to the new index is performed without the administrator checking (S12). When the above process is completed, the switching determination unit 24 sets all the items in the search item usage frequency file to 0 and accumulates the count of the index counter 22 for each item until the next switching command is output. (S13).

As described above, the index registered in the index registration section 21 is updated to the optimum index at every time interval set in the timer 25. This update will be performed for each line in the example of FIG. Therefore, the search work from each line can be made very efficient. In addition, the index to be first registered in the index registration unit 21 may be selected relatively roughly, and the labor of this work can be saved.

[0031]

As described above, according to the present invention, the frequency of use of each index used in the terminal device for retrieval is calculated, and the frequency of use is calculated from among the calculated frequency of use of each index. Since an index with a frequency higher than the preset frequency of use is selected and the selected index is replaced with the registered index to be updated and registered, the index change operation is not required, and an optimum index is always available. You will be able to register the index.

[Brief description of drawings]

FIG. 1 is a block diagram showing a schematic configuration of an apparatus that realizes an index automatic setting method according to the present invention.

FIG. 2 is a diagram showing an example of a search item use frequency file stored in the use frequency file shown in FIG.

FIG. 3 is a diagram showing an example of a search item use frequency integration file stored in the use frequency file shown in FIG.

FIG. 4 is a flowchart showing the procedure of the method of the present invention.

FIG. 5 is a flow chart showing the procedure of the method of the present invention.

FIG. 6 is a diagram showing an example of usage frequencies calculated by a switching determination unit shown in FIG. 1.

FIG. 7 is a diagram for explaining selection of an index selected by the switching determination unit shown in FIG.

[Explanation of symbols]

10 ... Database 20 ... Server 21 ... Index registration unit 22 ... Index counter 23 ... Usage frequency file 24 ... Switching determination unit 25 ... Timer 26 ... Switching unit

Claims (1)

[Claims] 1. A search method for enabling a high-speed search by registering an arbitrary item as an index used at the time of search among a plurality of items attached to relational data and using the index at the time of search. An index automatic setting method, which calculates a usage frequency of each index used in a terminal device for a search, and the calculated usage frequency of each index is higher than a preset usage frequency. An automatic index setting method, characterized in that an index having a high frequency is selected, the selected index is replaced with a registered index, and updated and registered.