JP2000099514A - Method and device for deciding retrieval range of database, and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To always obtain the number of retrieval results within the designated range at the time of repeating retrieval to a database that cannot previously and completely grasp a pair of information on the retrieval condition of a database being a retrieval object and the number of retrieval results. SOLUTION: A retrieval range deciding device 12 executes previous sampling investigation based on data specifying a retrieval condition and a database 14 being a retrieval object, which are given from a retrieval repeating device 10, obtains data on the retrieval condition and a retrieval range as the retrieval result and decides the retrieval condition and the retrieval range at the time of retrieving the database 14 being the retrieval object by the retrieval repeating device 10 based on the retrieval result.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method and apparatus for determining a search range of a database and a computer-readable recording medium storing a program for causing a computer to execute the method of determining a search range of a database. The range specified by the number of search results when relaying a search to a database where the information of the pair of the search condition and the number of search results for a different database (hereinafter referred to as a search target database) cannot be completely grasped in advance. The search range is determined before executing the search so that the search range is within the range, and the search is executed. Computer-readable recording medium recording a program to be executed About.

[0002]

2. Description of the Related Art In a conventional database information search, in order to make the number of search results for a search condition fall within a specified range, a search is executed once and the number of search results is not an appropriate number. Even if the search conditions are changed and the search is executed again, or the search range is determined by estimating the number of search results before executing the search, the search conditions and the number of search results in the search target database are all known in advance. The search range of the search target database is determined by a method based on the premise that there is a search condition or a method using a simple rule of thumb of the number of search results, such as inferring from the relationship between the number of search results for a plurality of search conditions.

[0003]

However, in the above-described conventional database information search, the search is relayed to a database in which the information of the pair of the search condition and the number of search results of the search target database cannot be completely grasped in advance. In doing so, there is a problem that it is not possible to estimate the number of search results so that the number of search results falls within a specified range and determine the search range in advance before executing the search. In addition, with a method that uses a simple rule of thumb for the number of search results, such as inferring from the relationship between the number of search results for multiple search conditions, the information on pairs of search conditions and search results in the search target database is completely grasped in advance. When a search is relayed to a database that cannot be used, there is a problem that it is not possible to estimate the number of search results and determine a search range for search conditions having no rule of thumb.

Further, when a search is relayed to a database in which the information of the pair of the search condition and the number of search results of the search target database cannot be completely grasped in advance, the type of information contained in the search target database is not changed. There is a demand for estimating the number of search results and determining the search range for different databases.

The present invention has been made in view of such circumstances, and relays a search to a database in which information on a pair of a search condition and the number of search results of a search target database cannot be completely grasped in advance. It is an object of the present invention to provide a database search range determination method, apparatus, and recording medium that can perform search relay so that the number of search results within a designated range can always be obtained.

[0006]

In order to achieve the above object, the invention according to claim 1 is directed to a database in which information on a pair of a search condition and the number of search results of a search target database cannot be completely grasped in advance. When relaying a search for
A search condition is sampled in advance from a basic database different from the search target database, a search is performed in advance on the search target database only for the sample search condition, and the basic database and the search are performed on those sample search conditions. The number of search results in the target database was obtained, the ratio of the number of search results in the search target database to the number of search results in the basic database was calculated, and the number of search results in the basic database was calculated in advance for any search condition. The present invention is characterized in that the number of search results in the search target database of the specified search condition is estimated by multiplying the ratio of the number of search results in the search target database and the base database.

According to the second aspect of the present invention, when the search is relayed to a database in which the information of the pair of the search condition and the number of search results of the search target database cannot be completely grasped in advance, the number of search results is reduced. In order to determine the search range in advance so that it is within the specified range, the search condition when the search range of the specified search condition is expanded using the basic database different from the search target database and when the search condition is narrowed Are determined, the number of search results described in claim 1 is estimated with respect to the search conditions, and a search range is determined from the estimated result so that the number of search results falls within a desired range. It is characterized by the following.

According to a third aspect of the present invention, in the method for determining a search range of a database according to any one of the first and second aspects, a plurality of basic databases having different types of contained information are prepared, The statistical correlation between the number of search results in the search target database and the number of search results in the basic database for each type of information is calculated, and the type name of the highly correlated information is determined by the type of information contained in the search target database. Judgment, a statistical calculation of the ratio of the number of search results is performed for the search target database and the basic database for each type of the highly correlated information, and the search result of the search target database is calculated based on the calculated ratio of the number of search results. The number is estimated, and a search range is determined from the estimation result.

According to the first to third aspects of the present invention, when a search is relayed to a database in which information on a pair of a search condition and the number of search results of a search target database cannot be completely grasped in advance, Since the search range is determined in advance so that the number of search results falls within the specified range, it is possible to perform search relay so that the number of search results within the specified range is always obtained.

According to a fourth aspect of the present invention, when the search is relayed to a database in which the information of the pair of the search condition and the number of search results of the search target database cannot be completely grasped in advance, the search target In a database search range determining apparatus for determining a search range of a database, a search condition extracting unit that samples a search condition in advance from a basic database different from the search target database, and a search condition extracted by the search condition extracting unit. Sample search execution means for executing a sample search on a search target database in advance only for a sample search condition, and a coefficient storing coefficient data for predicting the number of search results for an arbitrary search condition for each search target database Database, and basic database and database to be searched for the sample search condition The number of search results in the base database is calculated, a coefficient which is the ratio of the number of search results in the search target database and the number of search results in the base database is calculated, and coefficient determination means to be stored in the coefficient database. The number of results is multiplied by a coefficient read from the coefficient database, which is the ratio of the number of search results of the search target database and the base database calculated in advance for the arbitrary search condition, and the search target of the specified search condition is Means for estimating the number of search results in the database.

According to a fifth aspect of the present invention, in the apparatus for determining a search range of a database according to the fourth aspect, use basic database determining means for determining a basic database to be used which is different from the search target database; From the basic database to be used, a plurality of search conditions when the search range of the specified search condition is expanded and a plurality of search conditions when the search range is narrowed are determined, and the determined search condition is searched in the basic database to be used. Acquiring the number of results, and acquiring the coefficients relating to the basic database and the search target database to be used from the coefficient database, and multiplying the coefficient by the number of search results in the basic database to be used and the number of search results of the search target database Means for estimating the number of search results for estimating Results Results number and having a search range determining means for the determination of the search range so as to fall within a desired range.

According to a sixth aspect of the present invention, there is provided the apparatus for determining a search range of a database according to any one of the fourth and fifth aspects, further comprising: a plurality of information-type basic databases having different types of contained information; An information type database in which an information type name contained in the search target database is stored for each database; and a type name of the highly correlated information is determined as the type of information contained in the search target database, and the information type name is determined as the information. Information type determining means for storing in a type database, wherein the coefficient determining means statistically calculates the number of search results in the search target database for the sample search condition and the number of search results in the basic database for each type of information. And the search result estimating means includes a search target database and information having the high correlation. Statistical calculation of the ratio of the number of search results is performed on the basic database for each type of information to be performed, and the number of search results of the search target database is estimated based on the calculated ratio of the number of search results. A search range is determined from the estimation result.

According to the present invention, when a search is relayed to a database in which the information of the pair of the search condition and the number of search results of the search target database cannot be completely grasped in advance, Since the search range is determined in advance so that the number of search results is within the specified range, the search range of the database can be used to perform search relay so that the number of search results within the specified range is always obtained. A decision device can be realized.

According to a seventh aspect of the present invention, when a search is relayed to a database in which information on a pair of a search condition and the number of search results of the search target database cannot be completely grasped in advance, the search target database A search condition is sampled in advance from another basic database, and a search is performed in advance on the search target database only for the sample search condition, and the sample search conditions are compared in the basic database and the search target database. Find the number of search results, calculate the ratio of the number of search results between the search target database and the base database, and calculate the number of search results in the base database and the search target database calculated in advance for any search conditions. Multiplied by the ratio of the number of search results in the basic database to the search target database of the specified search condition. It is summarized as a computer-readable recording medium having recorded thereon a program for causing a computer to execute the search range determining method of the database, characterized in that to search results number of estimation in.

According to the invention described in claim 8, when the search is relayed to a database in which the information of the pair of the search condition and the number of search results of the search target database cannot be completely grasped in advance, the number of search results is reduced. In order to determine the search range in advance so that it is within the specified range, the search condition when the search range of the specified search condition is expanded using a basic database different from the search target database, In this case, a plurality of search conditions are determined, and the number of search results according to claim 1 is estimated with respect to the search conditions, and a search range is determined from the estimated results so that the number of search results falls within a desired range. A gist of the present invention is a computer-readable recording medium that stores a program for causing a computer to execute a method of determining a search range of a database.

Further, according to the ninth aspect of the present invention, a plurality of basic databases having different types of contained information are prepared,
Calculates the statistical correlation between the number of search results in the search target database for the sample search condition and the number of search results in the basic database for each type of information, and the information that the search target database contains the type name of the highly correlated information And performing a statistical operation on the ratio of the number of search results for the search target database and the basic database for each type of the highly correlated information, and based on the calculated ratio of the number of search results, the search target database. Estimate the number of search results for,
A computer-readable recording medium storing a program for causing a computer to execute the database search range determination method according to claim 1, wherein a search range is determined from the estimation result. Make a summary.

According to the seventh to ninth aspects of the present invention, when a search is relayed to a database in which the information of the pair of the search condition and the number of search results of the search target database cannot be completely grasped in advance, Since the search range is determined in advance so that the number of search results is within the specified range, a database search that can perform search relay so that the number of search results within the specified range is always obtained A range determination method can be implemented.

[0018]

Embodiments of the present invention will be described below in detail with reference to the drawings. FIG. 1 shows an outline of a configuration of a search relay system to which the present invention is applied. In FIG. 1, a search relay system performs an information search on a search target database in response to a request from a searcher 1 and relays the search result to inform the searcher 1 of a search result. A pre-sampling survey is performed on the search target database 14, and the search conditions, the search target database identifier, and the desired number of search results given by the search relay device are referred to by referring to the search results obtained by the pre-sampling survey. And a search range determining device 12 for determining a search condition and a search range at the time of execution of the search.

It is assumed that the search target database 14 cannot completely grasp in advance the information of the pair of the search condition and the number of search results.

When the searcher 1 makes a search request to the search relay device 10 in the above configuration, the search relay device 10 responds to the search request of the searcher 1 by specifying search conditions and data (for example, data specifying the search target database 14). , The search target database identifier) and the desired number of search results are sent to the search range determination device 12.

On the other hand, the search range determination device 12 performs a pre-sampling survey on the search target database 14 before executing the search. That is, a search condition is sampled in advance from a basic database different from the search target database 14, a search is performed in advance on the search target database 14 only for the sample search condition, and the given search condition and Data as a basis for estimating the number of search results for the search range is obtained as search results. This pre-sampling survey is performed only once in the entire search relay system, not every time the search relay is performed, and is performed according to the search condition given by the search relay device 10 at the time of the search relay. Not something.
The search range determination device 12 refers to the search result of the pre-sampling survey, and searches for the search condition provided by the search relay device 10, data for specifying the search target database 14 (for example, a search target database identifier), and the number of desired search results. The search condition and the search range at the time of executing the search to the search target database 14 by the search relay device 10 are determined based on each data of
Send to 0.

The search relay device 10 is a search range determining device 1
2, the data of the search condition and the search range are received and sent to the search target database 14. The search target database 14 searches for information based on the given search conditions and search range, and outputs the search results to the searcher 1 via the search relay device 10.

Next, FIG. 2 shows the configuration of the database search range determining apparatus according to the embodiment of the present invention. In the figure, the apparatus for determining the search range of the database includes
00 and a pre-operation part 120.

The constantly operating portion 100 is a portion that operates when a process of determining a search range at the time of executing a search of the search target database 14 is performed. The search relay device 10 in FIG. Data 300 indicating the desired number of search results is received, and data 301 indicating the determined search condition and search range is output to search relay device 10.

The pre-operation portion 120 is a portion which is executed in advance of the process of determining the search range when executing the search of the search target database 14, and is used for determining the search range when the search of the search target database 14 is executed. , An information type database 130 and a coefficient database 132 are constructed. Here, the information type database 130 is a database in which data indicating the type of information contained in the search target database is stored for each of the plurality of search target databases 14, and the coefficient database 132 is a search result of the search target database 14. This is a database that stores data indicating the coefficient of the number of search results obtained by statistically processing the ratio between the number and the number of search results in the basic database 140 or the basic database 142 described below.

The basic database 140 is a database holding a general number of information, and the basic database 142 is a basic database 140 for each type of information.
This is a database consisting of a plurality of basic databases holding the same data.

The pre-operation section 120 includes a search condition extracting section 122 for sampling a search condition and a search condition extracting section 1.
A sample search execution unit 124 that executes a preliminary sample investigation (sample search) based on the search conditions sampled by the sample search unit 22, the search target database 14 and the basic database 1 received from the sample search execution unit 124
A coefficient determination unit 126 that performs statistical processing on the number of search results of the search target databases 40 and 142 to determine a coefficient for estimating the number of four search results of the search target database 14, a search target database 14 received from the sample search execution unit 124, An information type determination unit 128 that performs statistical processing on the number of search results of the basic databases 140 and 142 and determines the type of information for each search target. Here, the search condition extraction unit 122 is a search condition extraction unit of the present invention, the sample search execution unit 124 is a sample search execution unit of the present invention, the coefficient determination unit 126 is a coefficient determination unit of the present invention, and the information type determination unit 128 Respectively correspond to information type determination means of the present invention.

Next, the specific processing contents (pre-processing) of the pre-operation section 120 will be described with reference to FIG. First, the search condition extracting unit 122 receives the search condition list from the basic database 140, and also receives the search condition list from the basic database 142 that holds similar data for each type of contained information, and samples some search conditions ( Step 200). Next, the sample search execution unit 12
4 executes a sample search in the search target database 14 based on the search condition received from the search condition extraction unit 122. That is, the search condition received from the search condition extraction unit 122 and the search range determined by the sample search execution unit 124 itself are sent to the search target database 14, and the search result is received from the search target database 14 (step 202).

Next, the coefficient determination unit 126 receives information on pairs of search conditions and the number of search results for each of the plurality of search target databases 14 from the sample search execution unit 124, and retrieves the search conditions from the basic databases 140 and 142, respectively. And a pair of information of the number of search results and the ratio of the number of search results in the search target database to the number of search results in the base database 140 or the base database 142 are statistically calculated for each search target database 14. (Step 204) A coefficient of the number of search results for estimating the number of search results of the search target database 14 is determined (Step 206). Next, the coefficient determining unit 126 converts the data of the set of the search target database identifier, the basic database identifier, and the number of search result coefficients into the coefficient database 132.
(Step 208). The number of search results in the search target database 14 is estimated by multiplying the coefficient of the number of search results by the number of search results in the basic database in the execution process of the constant operation part described later.

On the other hand, the information type determination unit 128 sends a plurality of search target databases 1 from the sample search execution unit 124.
The information of the pair of the search condition and the number of search results is received for each of the four, and the information of the pair of the search condition and the number of search results is received from the basic databases 140 and 142, respectively, and these data are statistically operated. It is checked whether the data of the pair of the search condition and the number of search results obtained from the target database 14 has a higher correlation with the basic database 140 or the basic database 142 provided for each information type (step 210). As a result of examining the correlation of the number of search results in step 210, it is determined that the type of information in the basic database 140 or the type-specific basic database 142 having a high correlation is the type of information for each search target database 14 (step 212). , Search target database identifier,
Basic database 142 by type of information with high correlation
The data of the set of identifiers (or the basic database 140) is stored in the information type database 130 (step 214). The data of a set of the search target database identifier and the identifier of the basic database 142 (or the basic database 140) for each type of information having a high correlation is referred to in a later-described runtime process performed in a constantly operating part, and is referred to by the search target database 14. In order to estimate the number of search results, the basic database 142 (or the basic database 140) for each type of information having a high correlation is used.

On the other hand, the constantly operating part 100 includes a used basic database determining unit 102 for determining a basic database to be used, a basic database determined by the used basic database determining unit 102, and a coefficient database 13
2, a search result number estimating unit 104 for estimating the number of search results of the search target database 14 with reference to the content of the search target database 14, and a coefficient determining unit 106 for determining a search range based on the estimation result by the search result number estimating unit 104. Have. Here, the used basic database determining unit 102 is used as the used basic database determining unit of the present invention, the search result number estimating unit 104 is used as the search result number estimating unit of the present invention, and the search range determining unit 106 is used as the search range determining unit of the present invention. , Respectively.

Next, the specific processing content (execution processing) of the constantly operating part 100 will be described with reference to FIG. First, the use basic database determination unit 102 receives the search condition, the search target database identifier, and the data 300 indicating the desired number of searches from the search relay device 10 shown in FIG. Basic database 1 for each type of information indicating the type of information and the identifier of the search target database and the information having a high correlation
The data group of the identifier set 42 (or the basic database 140) is read out, a match is made with the search target database identifier, and the information type of the search target database 14 is determined. That is, the identifier of the basic database to be used for determining the search range of the search target database 14 is obtained (step 400).

Next. The search result number estimating unit 104 receives the identifier of the basic database (140 or 142) to be used from the use basic database determining unit 102, and refers to the basic database specified by the basic database identifier to specify the search condition. A search condition when the search range is expanded or a search condition when the search range of the search condition is narrowed is determined (step 402).

Next, the search result number estimating unit 104 acquires the number of search results in the basic database from the basic database to be used for the search condition determined in step 402 (step 404), and obtains the coefficient database 1
32, a coefficient obtained by statistically calculating the ratio of the number of search results between the search target database and the basic database to be used is obtained (step 406), and the number of search results in the basic database obtained in step 404 is obtained. Is multiplied by the coefficient obtained in step 406 to estimate the number of search results in the search target database 14 (step 408).

The search range determining unit 106 receives the number of search results in the search target database 14 estimated by the search result number estimating unit 104, and determines search conditions so as to fall within a desired search range (step 410).

The pre-operation part 120 and the always-operation part 10
0 (FIGS. 3 and 4) corresponds to the database search range determination method according to the embodiment of the present invention.

A computer-readable recording medium that records a program for causing a computer to execute the database search range determining method according to the embodiment of the present invention also belongs to the technical scope of the present invention.

Next, the database search range determining device 12
Will be described with reference to FIGS. 5 to 9. FIG. In this embodiment, when an "address" of a certain place is given by a searcher as a search condition, the search range determination device 12 of the database sets the number of search results in a desired range according to the search target database. The search range of the “address” as the search condition received in the above is adjusted, changed, and output.

First, the processing contents of the pre-operation section 120 in FIG. 2 will be described. The basic database 140 and the information-specific basic database 142 are, for example, as shown in FIG.
It is as shown in. The information type-specific basic database 142 has a plurality of basic databases, and includes a basic database 142-1 relating to hotel information. Here, it is assumed that the search target database 14 has two types of search target databases A and B, and the following processing is performed once.

First, the retrieval condition extraction unit 122 retrieves three addresses of "Chuo-ku, Tokyo", "Iruma-gun, Saitama" and "Kita-ku, Osaka-shi" from the basic databases 140 and 142 shown in FIG. Sampling is performed as a condition. The sample search execution unit 124 executes a sample search for each of the search target databases A and B for these three addresses, and obtains an execution result as shown in FIG.

The coefficient determining unit 126 determines “Chuo-ku, Tokyo”,
For the search conditions of “Iruma-gun, Saitama” and “Kita-ku, Osaka, Osaka”, the result of the sample search by the sample search execution unit 124 is compared with the contents of the basic databases 140 and 142, and statistical calculation processing is performed. As shown in FIG. 7, the basic databases 140, 1
42 and is stored in the coefficient database 132. The coefficients stored in the coefficient database 132 shown in FIG. 7 are the results of statistical calculation of the ratio between the number of search results in the search target databases A and B and the number of search results in the basic databases 140 and 142. As one of the statistical processes, for example, taking a simple average is conceivable, and FIG. 7 shows an example in which a simple average is taken. That is, as shown in FIG. 5, the number of information items for the search conditions of “Chuo-ku, Tokyo”, “Iruma-gun, Saitama”, and “Kita-ku, Osaka-shi” in the basic database 140 are 74,000 and 7,000 respectively. 000
And 50,000, and the results of the sample search of the search target database A are as follows: The number of information items for the search conditions of "Chuo-ku, Tokyo", "Iruma-gun, Saitama", and "Kita-ku, Osaka-shi" 7,000,500,3,40 respectively
Since there are no records, the coefficient between the search target database A and the basic database 140 is (7000/7400 + 500/7000 + 3400 /
50,000) /3@0.078 Also, the number of information for the search conditions of “Chuo-ku, Tokyo”, “Iruma-gun, Saitama”, and “Kita-ku, Osaka-shi” in the basic database (hotel) 142-1 is as follows. 1000 cases, 1
Since there are 0 and 800 cases, the coefficient between the search target database A and the basic database 142-1 is (7000/1000 + 500/10 + 3400/80)
0) /3≒20.42.

The sample search execution results of the search target database B are “Chuo-ku, Tokyo”, “Iruma-gun, Saitama”,
The number of information items for the search condition of “Kita Ward, Osaka City, Osaka Prefecture” is 900, 9, and 720, respectively. Therefore, the coefficient between the search target database B and the basic database 140 is (900/74000 + 9). / 7000 + 720/500
00) /3≒0.0092, and the coefficient between the search target database B and the basic database (hotel) 142-1 is (900/1000 + 9/10 + 720/800) / 3
≒ 0.90. Therefore, the basic database 1 of the search target databases A and B stored in the coefficient database 132
The coefficients for 40 and 142-1 are as shown in FIG. On the other hand, the information type determination unit 128 compares the sample search result by the sample search execution unit 12 with the basic database 14.
0, 142-1 are compared, statistical calculation is performed, the correlation between each of the search target databases A, B and each of the basic databases 140, 142-1 is checked, and the information type is determined. Then, the processing result shown in FIG. 8 is obtained.

In this embodiment, the information type is determined by the information type determining unit 128 based on the level of the correlation coefficient r between the databases. Generally, the correlation coefficient r is obtained by the following calculation. The variables of the search target databases A and B are x, the variables of the basic databases 140 and 142-1 are y,
Assuming that the average values of the variables x and y are x and y, the average values x and y
Is

(Equation 1) Becomes If the square of the standard deviation of the variables x and y is Sx and Sy,

(Equation 2)

(Equation 3) Becomes

Also, Sxy is

(Equation 4) Then, the correlation coefficient r from Sx, Sy, and Sxy is

(Equation 5) Becomes The correlation coefficient r between the databases is specifically calculated by the equations (1) to (5) below. ｛X｝ = ｛7
000,500,3400}, {y} = {74000,
If it is set to 7000, 50000, then, according to the equations (1) to (4), x becomes 3633, y is 43667 Sx is 21206667, Sy is 2304666667 Sxy is 215333333. Therefore, from the equation (5), the correlation coefficient r between the search target database A and the basic database 140 is obtained.
1 becomes r1 = 0.97.

Similarly, the correlation coefficient r2 of the search target database A with the basic database (hotel) 142-1 is obtained. {X} = {7000,500,3}
Assuming that 400｝, {y} = {1000, 10, 800}, x {3633, y} 603 Sx {21206667, Sy ≒ 548067 Sxy ≒ 3148667 from equations (1) to (4). Therefore, from Equation (5), the correlation coefficient r2 of the search target database A with the basic database 142-1 is r2 = 0.92.

Further, the correlation coefficient r3 between the search target database B and the basic database 140 is {x} = {9
00,9,720}, {y} = {74000,700
If it is 0.50000, x {543, y} 43667 Sx {443934, Sy} 230466667 Sxy @ 31530000 is obtained from the equations (1) to (4). Therefore, from the equation (5), the correlation coefficient r between the search target database B and the basic database 140 is obtained.
3 becomes r3 = 0.98. The correlation coefficient r4 between the search target database B and the basic database (hotel) 142-1 is
{X} = {900,9,720}, {y} = {100
If 0, 10, 800}, x {543, y} 603 Sx {443934, Sy} 548067 Sxy @ 493260 from equations (1) to (4). Therefore, from equation (5), the correlation coefficient r4 between the search target database B and the basic database 142-1 is r4 = 1.00.

The correlation coefficient between the databases shown in FIG. 8 is obtained by the above-described calculation. Therefore, as is clear from FIG. 8, the information type determination unit 128 determines that the search target database A has a high correlation with the basic database 140 containing general information, and that it is appropriate to use the basic database 140. I do. For the search target database B, a basic database (by information type) 142
Of which information type is the basic database 142-1 of the hotel
Is determined to be highly correlated, and the type of information contained is a hotel, and it is determined that it is appropriate to use the basic database of the hotel among the basic databases (by information type) 142. The information type determination unit 128 determines the result of the determination marked with a circle in FIG. 8, that is, the data of the set of the identifier of the search target database A and the identifier of the basic database 140 for the search target database A, and the search result for the search target database B. The data of the set of the identifier of the search target database B and the identifier of the basic database 142-1 of the hotel among the basic databases (by information type) 142 is stored in the database 130 by information type.

Next, the processing content of the constantly operating portion 100 in FIG. 2 will be described. It is assumed that “1” near Nishi-Shinjuku, Shinjuku-ku, Tokyo is specified as a search condition and “B” is specified as a search target database identifier by the search relay device 10. Use basic database determination unit 102
Matches the search target database identifier B with reference to the information type database 130, and finds the basic database (by information type) 142 for the search target database B.
It is determined that the basic database 142-1 of the hotel is used. Next, the number-of-retrieval-result estimating unit 104 includes the basic database (hotel) 142-1 and the coefficient database 1
32, the coefficient of the number of search results for the basic database (hotel) 142-1 is determined to be “0.9” with respect to the search target database B, and “Tokyo” in the basic database (hotel) 142-1 is determined. Tokyo "," Shinjuku-ku, Tokyo "," Nishi-Shinjuku, Shinjuku-ku, Tokyo ",
The number of information for each search condition of "Nishi-Shinjuku 1-chome, Shinjuku-ku, Tokyo" is 30,000, 3,000, respectively.
Since there are 200 and 20 search results, the number of search results in the search target database B for these search conditions is calculated by multiplying the above information count by the coefficient “0.9” as shown in FIG. 27,000, 2,700, 180 for each search condition of "Shinjuku-ku, Tokyo", "Nishi-Shinjuku, Shinjuku-ku, Tokyo" and "1-Chome Nishi-Shinjuku", Shinjuku-ku, Tokyo
And 18 cases.

Next, assuming that the range of the desired number of information is set to, for example, 100 to 200, the search range determining unit 106 determines that the number of search results belongs to the above range from the estimation result shown in FIG. Search condition is "Nishi-Shinjuku, Shinjuku-ku, Tokyo"
Is determined, and data 3 indicating a search condition and a search range
01 and search for “Nishi-Shinjuku, Shinjuku-ku, Tokyo”
Output to According to the embodiment of the present invention, when a search is relayed to a database in which the information of the pair of the search condition and the number of search results cannot be completely grasped in advance, the number of search results falls within a specified range. As described above, since the search result of the search target database is estimated and the search range is determined in advance for all the search executions, it is possible to carry out the search relay so that the number of search results within the designated range is always obtained.

Further, according to the present embodiment, the search target database is not based on the empirical rule of the number of search results as in the prior art, but based on the information of the pair of the search condition and the number of search results in the basic database. Since the number of search results is estimated in, the number of search results can be estimated and the search range can be determined even for search conditions having no rule of thumb. Furthermore, according to the present embodiment, the number of search results is estimated and the search range is determined in consideration of the type of information contained in the search target database. The number of search results can be estimated and the search range can be determined for the database.

[0051]

As described above, according to the first to third aspects of the present invention, a database in which the information of the pair of the search condition and the number of search results of the search target database cannot be completely grasped in advance. When the search is relayed, the search range is determined in advance so that the number of search results is within the specified range. It can be carried out. According to the invention as set forth in claims 4 to 6, when the search is relayed to a database in which the information of the pair of the search condition and the number of search results of the search target database cannot be completely grasped in advance, the number of search results is reduced. The search range is determined in advance so that the search range is within the specified range. Therefore, a database search range determination device that can perform a search relay so that the number of search results within the specified range is always obtained is provided. Can be realized. According to the invention described in claims 7 to 9,
When relaying a search to a database in which the information of the pair of the search condition and the number of search results of the search target database cannot be completely grasped in advance, set the search range in advance so that the number of search results falls within the specified range. Since the determination is made in advance, it is possible to implement a database search range determination method capable of performing a search relay so that the number of search results within a designated range is always obtained.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a database search and relay system to which the present invention is applied.

FIG. 2 is a block diagram showing a configuration of a database search range determining device according to the embodiment of the present invention.

FIG. 3 is a flowchart showing processing contents executed by a preliminary operation part in the database search range determination device shown in FIG. 2;

FIG. 4 is a flowchart showing processing contents executed by a constantly operating part in the database search range determining device shown in FIG. 2;

FIG. 5 is an explanatory diagram showing the contents of data stored in each of a basic database and a basic database (by information type).

6 is an explanatory diagram showing an execution result of a sample search on a search target database by a sample search execution unit in the database search range determination device in FIG. 2;

FIG. 7 is an explanatory diagram showing the contents of data stored in a coefficient database in the database search range determining device in FIG. 2;

FIG. 8 is an explanatory diagram showing a processing result of an information type determination unit in the database search range determination device in FIG. 2;

9 is an explanatory diagram showing an estimation result of the number of search results for a search target database by a search result number estimation unit in the database search range determination device in FIG. 2;

[Explanation of symbols]

 REFERENCE SIGNS LIST 10 search relay device 12 database search range determination device 14 search target database 100 constantly operating portion 102 basic database used determining portion 104 search result number estimating portion 106 search range determining portion 120 pre-operation portion 122 search condition extracting portion 124 sample search executing portion 126 coefficient determination unit 128 information type determination unit 130 information type database 132 coefficient database 140 basic database 142 basic database (by information type)

 ──────────────────────────────────────────────────続 き Continued on the front page (72) Inventor Seiji Yokoji 3-19-2 Nishi-Shinjuku, Shinjuku-ku, Tokyo Japan Telegraph and Telephone Corporation (72) Inventor Kenichi Shima 3-9-1-2, Nishi-Shinjuku, Shinjuku-ku, Tokyo No. F-term in Nippon Telegraph and Telephone Corporation (reference) 5B075 NK10 NK31 NK37 NR03 NR05 NR12 PP23 PQ05 PR03 PR10 QP05 QP10 UU40

Claims (9)

[Claims] When a search is relayed to a database in which information on a pair of a search condition and the number of search results of a search target database cannot be completely grasped in advance, a search is performed in a basic database different from the search target database. , Sample the search conditions in advance, perform a search in advance on the search target database only for the sample search conditions, obtain the number of search results in the basic database and the search target database for those sample search conditions, The ratio of the number of search results between the database and the basic database is calculated in advance, and the number of search results in the basic database and the number of search results between the search target database and the basic database calculated in advance for arbitrary search conditions are calculated. Estimation of the number of search results in the search target database of the specified search condition multiplied by the ratio And determining the search range of the database. 2. When a search is relayed to a database in which information on a pair of a search condition and a number of search results of a search target database cannot be completely grasped in advance, the number of search results is within a specified range. In order to determine the search range in advance, a plurality of search conditions are determined when the search range of the specified search condition is expanded and narrowed using a basic database different from the search target database. A database search method comprising: estimating the number of search results according to claim 1 with respect to the search conditions; and determining a search range from the estimated result so that the number of search results falls within a desired range. Range determination method. 3. A plurality of basic databases having different types of contained information are prepared, and the number of search results in a database to be searched for sample search conditions and the number of search results in the basic database for each type of information are statistically calculated. Calculate the correlation, determine the type name of the highly correlated information as the type of information contained in the search target database, and calculate the statistics of the ratio of the number of search results between the search target database and the basic database for each type of the highly correlated information. 3. The method according to claim 1, further comprising: performing a logical operation, estimating the number of search results of the search target database based on a ratio of the calculated number of search results, and determining a search range from the estimated result. Method for determining search range of database described in 1. 4. A database for determining a search range of the search target database when a search is relayed to a database in which information on a pair of a search condition and the number of search results of the search target database cannot be completely grasped in advance. A search condition extracting device for sampling a search condition in advance from a basic database different from the search target database; and a search target database for only the sample search conditions extracted by the search condition extracting device. Sample search execution means for executing a sample search in advance, a coefficient database storing coefficient data for estimating the number of search results for an arbitrary search condition for each search target database, and the sample search condition Find the number of search results in the basic database and the search target database A coefficient determining means for calculating a coefficient which is a ratio between the search target database and the number of search results of the basic database, and storing the coefficient in the coefficient database; Estimating the number of search results in the search target database of the designated search condition by multiplying the read coefficient of the number of search results of the search target database and the basic database calculated in advance for the arbitrary search condition Means for estimating the number of search results, and a search range determining apparatus for a database. 5. A used basic database determining means for determining a basic database to be used which is different from the search target database, and a search condition when a search range of a specified search condition is expanded from the basic database to be used. And determining a plurality of search conditions in the case of narrowing, obtaining the number of search results in the basic database to be used for the determined search conditions, and obtaining coefficients relating to the basic database and the search target database to be used from the coefficient database. Means for estimating the number of search results of the search target database by multiplying the number of search results in the basic database to be used by the coefficient, and a search result from the estimation result by the search result number estimator. Search range determinator that determines the search range so that the number falls within the desired range When the search range determination apparatus of the database according to claim 4, characterized in that it comprises a. 6. A plurality of information type basic databases having different types of information to be contained, an information type database storing information type names contained in the search target database for each search target database, and the information having a high correlation. Information type determining means for determining the type name of the information as the type of information contained in the search target database, and storing the information type name in the information type database, wherein the coefficient determining means searches for the sample search condition. A statistical correlation between the number of search results in the target database and the number of search results in the basic database for each type of information is calculated, and the search result estimating means includes the search target database and information having a high correlation. Statistical calculation of the ratio of the number of search results is performed for the basic database for each information type, and the calculated search results are calculated. The number of search results of the search target database is estimated based on the ratio of the search target, and the search range determination unit determines a search range from the estimated result. Search range determination device. 7. When a search is relayed to a database in which the information of the pair of the search condition and the number of search results of the search target database cannot be completely grasped in advance, a search is performed in a basic database different from the search target database. , Sample the search conditions in advance, perform a search in advance on the search target database only for the sample search conditions, obtain the number of search results in the basic database and the search target database for those sample search conditions, The ratio of the number of search results between the database and the basic database is calculated in advance, and the number of search results in the basic database and the number of search results between the search target database and the basic database calculated in advance for arbitrary search conditions are calculated. Estimation of the number of search results in the search target database of the specified search condition multiplied by the ratio A computer-readable recording medium storing a program for causing a computer to execute a method for determining a search range of a database, the method comprising: 8. When a search is relayed to a database in which information on a pair of a search condition and the number of search results of a search target database cannot be completely grasped in advance, the number of search results is within a specified range. In order to determine the search range in advance, a plurality of search conditions are determined when the search range of the specified search condition is expanded and narrowed using a basic database different from the search target database. And a search range is determined so that the number of search results falls within a desired range based on the estimated result. A computer-readable recording medium that records a program for causing a computer to execute the search range determining method. 9. A plurality of basic databases having different types of contained information are prepared, and the number of search results in the search target database for the sample search condition and the number of search results in the basic database for each type of information are statistically calculated. Calculate the correlation, determine the type name of the highly correlated information as the type of information contained in the search target database, and calculate the statistics of the ratio of the number of search results between the search target database and the basic database for each type of the highly correlated information. 3. The method according to claim 1, further comprising: performing a logical operation, estimating the number of search results of the search target database based on a ratio of the calculated number of search results, and determining a search range from the estimated result. Computer-readable recording of a program for causing a computer to execute the method for determining the search range of the database described in 1. Recording medium.