JPH0319081A - Information retrieving system - Google Patents

Abstract

PURPOSE:To suppress the number of times of I/O of a file required for the retrieval to a prescribed number of times or below irrespective of whether the retrieval range is large or small by generating plural kinds of range type key data which become a combination element of a range containing numerical data and using them as indexes with respect to numerical data of information to be retrieved. CONSTITUTION:With respect to numerical data (m) of information to be retrieved, plural kinds of range type key data which become a combination element of a range containing (m) are generated in advance and stored in an index part. In this state, plural kinds of range type key data contained in a retrieval range given at the time of retrieving the information are combined and read out, an OR processing of the range type key data is executed, and retrieval data is obtained. Accordingly, a range retrieval can be executed by only read-out which combines the range type key data irrespective of whether a retrieval range is large or small. In such a way, the number of times of I/O of a file for reading out the range type key data of an index part required for the retrieval can be suppressed to a prescribed number or below.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention relates to an information retrieval system, and in particular, the present invention relates to an information retrieval system, and in particular, to numerical data of search target information, numerical data having a range, such as the melting point or boiling point of a chemical substance, is searched. The present invention relates to an information retrieval system that can effectively search numerical data such as the growth temperature of microorganisms.

[Conventional technology]

An information retrieval system is a system that stores information and retrieves it according to the purpose. Information searches include item searches and document searches. Item search involves storing information or data as is in a file and retrieving the relevant items by giving search conditions.The data element given as the search condition is called a search key. This is a search that specifies the content of recorded documents and searches for corresponding documents. For this reason, only important words (keywords) are extracted from among several words included in the literature information, and secondary information unique to the literature is accumulated.

Literature searches can be performed using a string of keywords.

Conventionally, in item searches, numerical data or numerical data with a range, such as the melting point or boiling point of a chemical substance, the growth temperature of microorganisms, etc., are given as the information to be searched, and these numerical data of the information to be searched are For,
When performing a range search in an item search, if a numerical range is given as a search key, the condition of the numerical range of the search key is compared and matched with a lot of information or data of the search item, which is the information to be searched. Therefore, it takes a lot of time to process the item search. Note that 1. Normally, numerical data is recorded as it is as key data in the index as a search key for searching for items.

For this reason, when performing a range search for numerical data in an information retrieval system, it is performed by searching an index for all numerical data that exists within the range, and calculating the logical OR of the obtained data set. .

In addition, as a publicly known document related to an information retrieval method that performs a range search for image information, for example, Japanese Patent Laid-Open No. 62-146
Publication No. 31 is mentioned. this is.

This is a microfilm search method in which a search range is set, a search key is operated, and if the search is within the search range, the search process is repeated; if the search is outside the search range, a warning is issued.

[Problem to be solved by the invention]

By the way, when constructing an information retrieval system that performs a range search using the above-mentioned technology, the following problems occur.

When constructing an information retrieval system, especially an information retrieval system using secondary media such as CD-ROM, the most important thing is the processing capacity represented by the turnaround time, but the range presented as the required specification is the most important. If the search is implemented using conventional methods, there is a concern that the processing time will be slow and the system will become impractical. That is,
As the range of set numerical values to be searched becomes wider, the number of numerical data existing within the range increases, and the number of times the index must be searched increases proportionally. Therefore, the number of times the file is searched increases in accordance with the number of times the index is searched, and the search time becomes longer.

The present invention has been made to solve the above problems.

An object of the present invention is to provide an information retrieval system that can effectively search for numerical data having a range for numerical data of information to be searched.

Another object of the present invention is to suppress the number of file I/Os necessary for a search to a certain number or less in a range search, regardless of the size of the range, to shorten the search time and improve processing capacity. It is in.

The above and other objects and novel features of the present invention will become apparent from the description of this specification and the accompanying drawings.

[Means to solve the problem]

In order to achieve the above object, the information retrieval system of the present invention is an information retrieval system that performs a search by giving a search range of an upper limit value and a lower limit value to the numerical data of the information to be searched. For numerical data m of search target information, multiple types of range type key data are created and stored in the index section as combination elements of a range including m, and when searching for information, multiple types of range type key data that are included in a given search range are created. The present invention is characterized in that the range type key data are read out in combination, and the range type key data are logically summed to obtain search data.

[Effect]

According to the above means, when performing a range search by giving a search range of an upper limit value and a lower limit value to the numerical data of the search target information, first, the numerical data m of the search target information is set in advance.
, a plurality of types of range type key data are created and stored in the index section as combination elements of the range including m. Then, when searching for information, multiple types of range-type key data included in a given search range are combined and read out.

Performs logical sum processing on range type key data to obtain search data. As a result, regardless of the size of the search range, it is possible to perform a range search just by reading a combination of range type key data, and the number of times a file's I10 number of times the range type key data of the index part required for the search is read is constant. number or less.

In this way, in a range search, regardless of the size of the range, it is possible to suppress the I10 number of times of files required for the search to a certain number or less, shorten the search time, and improve processing performance.

〔Example〕

Hereinafter, one embodiment of the present invention will be specifically described using the drawings.

Before explaining the embodiments in detail, first, the principles of the method for creating key data (range type key data) and the search method for information retrieval in the information retrieval system of the present invention will be described.

(1) Creation of range type key data Three types of range type key data are created for numerical data m, as described below. When a numerical value m (positive integer) is given, ①First range type key data; [m], Create key data whose range is the numerical data m as it is.

■Second range type key data; [(m-2)~m]t [(m-1) ~ (m + 1)].

[m~(m + 2)], Create key data for two ranges before and after the number m.

■Third range type key data; [10a-(10a+9)].

[10a~(10(a+1)+9)].

[10a~(10(a+2)+9)].

[10a~(10(a+3)+9)].

[10(a-1)~ [10(a-1)~ [10(a-1)~ [10(a-1)~ (10a+9)] = (10(a+1)+9)].

(10(a+2)+9)].

(10(a+3)+9)].

[10(a-2) to (10a+9).

[10(a-2) to (10(a+1)+9)].

[10(a-2) to (10(a+2)+9)].

[10(a-2) to (10(a+3)+9)].

However, a is an integer of the quotient obtained by dividing the numerical value m by 10.

When the quotient of dividing the number m by 10 is a (!1 number),
The upper limit value is set by sequentially increasing the value of the positive integer i until it does not exceed the upper limit value of the search tolerance range (1 = O * 1
m 2 t 3 m...), 10 (a+i)+9, and the value of the positive integer j is sequentially increased until the lower limit is not less than the lower limit of the search allowable range. (
j==o, l, 2, 3. ...).

Create multiple types of range type key data with a range of values that are sequentially combined with the values of 10 (a-j) (range = 9.19, 29, 39
．． ...).

With such a key data creation method, the first . Second. Third. Create three types of multiple types of range type key data.

And 1 for the created range type key data.

A range search is performed using the search method described below.

(II) Range search method When performing a range search for a numerical range [x = y], (a
) First, divide the lower limit value X and upper limit value y (x<y) by 10, set the quotients as a and c, and take the remainder as
Then, x=10a+b (a, b are positive integers) y=10c+d
(ct d is a positive integer).

(b) If the numbers in the 10th place of the lower limit value X and the upper limit value y are the same (a=c) (i) If the difference in the 1st place (d-b) is 9, the numerical clause 1! I
[Search the third range type key data of x-yl as is 0 For example, when performing a range search of the numerical range [20 to 29], the range type key data of [20 to 29] is searched as is.

(...) If the difference in 1st place (d-b) is 3 to 8, the difference in 01st place (d-b) divided by 3 equals the number of quotient i (integer),
The range [(X + 3k) to (x + (3k + 2)), respectively, where =o, 1, 2. ”, the second range type key data of (i-1) is searched.

Namely.

[X~(x+2)ko.

[(x + (3X l)) ~ (x + (3xl) + 2)]
.

[(x + (3X 2)) ~ (x + (3x 2) +
2) Ko.

[Search for second range type key data from (x+3(i-1)) to (x+3(i-1)+2).

(2) Search for second range type key data in the range [(y −2) to y].

■Next, find the logical sum of the data sets obtained from the search in ■ and the search in ■.

Specifically, 1. For example, when performing a range search for the numerical range 20 to 26, since i=2 from d-b=6, the numerical range [20 to 22] and the numerical range [23 to 25] are Search range type key data. Also, from y=26, range type key data in the numerical range [24 to 26] is searched. Then, the logical sum of the data sets obtained by searching these range type key data is calculated.

(m) If the difference in first place (d-b) is 2, the numerical range [x=
The second range type key data of yl is searched to find the corresponding data set.

Specifically, for example, when performing a numerical value range 25-27 range search, range-type key data in the numerical value range [25-27] is searched.

(Short) If the difference in first place (d-b) is 1, [x] and [
The first range type key data of yl is searched, a corresponding data set is obtained, and a logical sum is obtained.

Specifically, when performing a range search in the numerical range 25 to 26, the key data of [25] and [26] are searched and the logical sum thereof is calculated.

(Q) If there is a difference in the numbers in the 10th place of the lower limit value X and the upper limit value y, and the difference (ca) is 1, (i) the difference in the 1st place (d
If b) is 9.

The third range-type key data of the numerical range [x''y] is searched to obtain a corresponding data set.

Specifically, for example, when performing a range search for a numerical value range of 30 to 49, range type key data of [30 to 49] is searched.

(n) otherwise range [x~lOa+9] and range [10c~y] 2
Performing the processing in item (b) above in two ranges,
However, in the case of x=10a+9 or y=loc, the first range type key data of the key itself [x] and [yl is searched.

Next, the logical sum of the two data sets obtained as a result of processing the two data sets is determined.

For example, when performing a range search for the numerical range 25 to 34, first, search for the first numerical range 25 to 29 and the numerical range 3 for voice 2.
Divide into 0-34. Regarding the first numerical range 25 to 29, range type key data in the range [25 to 27] and the range [27 to 29] is searched, and the logical sum thereof is determined. In addition, for the second numerical range [30 to 34, the range [3
0 to 32] and range type key data in the range [32 to 34], and calculate the logical sum thereof. The logical sum of the two data sets obtained through these processes is calculated.

(d) When there is a difference between the numbers in the 10th place of the lower limit value X and the upper limit value y, and the difference (ca) is 2 or more.

(i) If the difference in first place (d-b) is 9, the numerical range [x
= y] to find the corresponding data set.

For example, when performing a range search for the numerical range 20 to 69, 2
Search for key data from 0 to 69.

(it) otherwise range [x~10a+9], range [10(a+1)~10
(c-1)+9], and the range [10C to y].

■First range 1 [x~10a+9] and third range [1
0c=y], the same processing as in the item (it) of (Q) described above is performed.

■Second range [10(a+1) to 10(c-1)+
9], the range data [10(a
+1) to 10 (c-1) Search for the third range type key data of +91.

■Calculate the logical OR of the data sets obtained through these processes (■ and ■).

For example, when performing a range search for the numerical range 25-57, the first numerical range 25-29. Second numerical range 30-49.
and a third numerical range of 50 to 57. For the first numerical range 25 to 29, the range [25 to
27] and the range type key data in the range [27 to 29], and calculate the logical sum. For the second numerical range 50-57, range [50-52], range [53-55],
and the range type key data in the range [55 to 57], and calculate the logical sum.

Also, for the numerical range [30 to 49], the range [30
~49] is searched for range type key data. The logical sum of the data sets obtained through these processes is calculated.

In a range search, as mentioned above, a range search is performed by combining range-type key data from the numerical data of the lower limit value and upper limit value for a given search range. It is no longer necessary to search all numerical data sequentially, and the number of searches within the index can be kept below a certain number of times.

FIG. 1 is a block diagram showing the configuration of essential parts of an information retrieval system according to an embodiment of the present invention.

In FIG. 1, 1 is a CD-ROM device for a database in which information to be searched is stored, 2 is a data processing device having a search processing mechanism, 3 is a printer as an output device for outputting search results, 4 is a display device, Further, 5 is a keyboard of an input device.

Here, a case will be described as an example in which an information retrieval system is constructed that can perform item retrieval by range retrieval for database data of microorganism data stored in CD-ROM.

The numerical data for performing the item search is the numerical data of each item in the microorganism data. For each numerical data,
Based on the above-mentioned principle, range type key data is created for each numerical value 24, for example, and stored as key data in the index section, as shown in FIG. 2a. As shown in Fig. 2b, for each key data of the range type key data of this index section, in order to associate the storage position of the data main body of the corresponding microorganism data, a pointer indicating the storage position of each data is used. are stored in correspondence with each key data to form an index section.

In the information retrieval system configured in this way, a range search is performed by the user inputting a search expression (attributes = key data) from the keyboard. In this case, the input search expression is a desired numerical range ( By inputting a search expression that is a combination of upper and lower limit values, CD
- Search the microorganism attribute data (vIt organism name, storage institution number, type, etc.) on the ROM and output the results.

Among the microorganism data to be searched, growth temperature and guanine/cytosine content consist only of numerical data, so in order to obtain the desired data, a range search is performed.CD-ROM
The data structure inside consists of an index section with key data and pointers, and a data section containing characteristic data of each microorganism. Therefore, for example, when searching for microorganisms whose growth temperature is in the range of 21 to 67°C, the search condition TM=21 to 67 is given as the search formula.

FIG. 3 is a flowchart illustrating search processing according to an embodiment of the present invention. Further, FIG. 4 is a diagram illustrating an example of combinations of range type data read out by the search process of FIG. 3.

This will be explained with reference to FIGS. 3 and 4.

■First, in step 31, refer to the range type key data stored in the growth temperature index section, divide the lower limit value 21 and upper limit value 67 of the numerical range by 10, and calculate the quotient and removal. .

2l-10X2+1, @-a=2. b=1.67=
10X6+7 to c=6. d=7 From the obtained quotient and remainder, the difference in the 10th place number (c-a) is 4, and the difference in the 1st place number (d-b) is determined to be 6, so change the numerical range from 21 to
It is classified into three ranges: 29.30-59° and 60-67, and a range search is performed for each numerical range.

(2) In the next step 32, a range search is performed for the numerical ranges 21-29.

First, the quotient i obtained by dividing the difference 8 between the first-place numbers by 3 is 2, so the 0th order range [27-26] is searched for range-type key data in the ranges [21-23] and [24-26]. 29] is searched for range type key data.

The logical sum of the data sets obtained through these search processes is calculated. As a result, range type key data a.

The pointers of b and c are read and combined.

(2) Next, in step 33, a range search is performed for the numerical range 30-59.

This range search is completed by directly searching the range type key data in the range [30 to 59]. As a result, the pointers of the range type key data d are read and combined.

■In the next step 34, the numerical value range 60 to 67
Perform a range search for.

In this range search process, the quotient i obtained by dividing the 1's digit difference 7 by 3 is 2, so the range [60-62], the range [
63 to 65] Search for range type key data Next, search for range type key data in the range [65 to 67] and calculate the logical sum of the data sets obtained by these search processes. As a result, the pointers of the range type key data elt fs g are read and combined.

(2) Subsequently, in step 35, the logical OR of the data sets obtained through these series of search processes is calculated, and the search for the given numerical range 21-67 is completed.

Searching for these range-type key data only requires reading pointer data from each range-type key data, so the search process is completed in a short time. Further, the process of calculating the logical sum of data sets read from each range type key data is also a logical sum process regarding each pointer data, and is completed in a short time. Therefore, the entire search process is completed in a short time.

When performing a range search, if the search range becomes large, normally all the key data within that range must be searched to obtain the result data, but as mentioned above, it is possible to combine each range type key data. Search processing can be performed simply by reading the file. By configuring the range type key data as described above, it is possible to suppress the number of searches to a maximum of seven times by combining the range type key data.

A CD-ROM has a storage capacity of 540 yen on a single medium.
Although it is a recording medium with a large storage capacity of MB, it takes longer to access compared to hard disks and floppy disks. Therefore, even if the data in the index section increases, by reducing the number of accesses, the overall search time can be reduced. It can be shortened.

As explained above, according to this embodiment, the number of searches for key data and the number of logical operations required for one range search in the search processing method are compared with the conventional method as shown in Table 1 below. This will improve search efficiency.

(Hereafter the margin of this page) Table 1 Comparison of the conventional method and this method As shown in Table 1, the number of searches for key data is n
No more than 7 times, regardless of the number of times. Therefore, search time can be shortened and search efficiency can be improved.

The present invention has been specifically explained above based on examples, but
It goes without saying that the present invention is not limited to the embodiments described above, and can be modified in various ways without departing from the spirit thereof.

〔Effect of the invention〕

As explained above, according to the information retrieval system of the present invention, in a range search, regardless of the size of the range, it is possible to suppress the I10 number of times of files required for the search to a certain number or less, thereby shortening the search time. As a result, the processing capacity of the system can be improved.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the configuration of the main parts of an information retrieval system according to an embodiment of the present invention, FIG. 2a is a diagram illustrating a method for creating range type key data, and FIG. 2b is an index section. FIG. 3 is a flowchart illustrating search processing according to an embodiment of the present invention. FIG. 4 is a diagram illustrating an example of combinations of range type data read out by the search process of FIG. 3. In the figure, 1...CD-ROM device, 2...Data processing device, 3...Printer, 4...Display device,
5...Keyboard.

Claims (4)

[Claims] (1) An information retrieval system that performs a search by giving a search range of an upper limit value and a lower limit value to the numerical data of search target information, in which m Multiple types of range type key data are created and stored in the index section as combination elements of a range that includes, and when searching for information, multiple types of range type key data included in a given search range are read out in combination, and the range type An information retrieval system characterized by performing logical sum processing on key data to obtain search data. (2) In the information retrieval system according to claim 1, when creating and having numeric key data with a range in the index to be searched, and performing a range search for the set numeric value,
An information retrieval system characterized in that a search is performed by combining range type key data. (3) In the information retrieval system according to claim 1, the plurality of types of range type key data stored in the index section are ranges that are combination elements of a range including m for numerical data m of the search target information. The type key data is first range type key data whose range is numerical data m as it is, second range type key data whose range is two before and after numerical data m, and a quotient obtained by dividing numerical value m by 10. When is a (integer), the upper limit value is a value that does not exceed the upper limit value of the search allowable range, and the value of an integer i greater than or equal to 0 is sequentially increased, and (i
= 0, 1, 2, ...) 10(a+i)+9, and sequentially increase the value of an integer j greater than or equal to 0 until the lower limit is not less than the lower limit of the search allowable range. and third range type key data having a range of sequentially combined values of (j = 0, 1, 2, ...) 10 (a-j). Information retrieval system. (4) In the information retrieval system according to claim 3, when performing a range search by giving a lower limit value x and an upper limit value y to multiple types of range type key data stored in the index section, the lower limit value x and the upper limit value y are given. Divide the upper limit y by 10 and get 1
For the range of 0th place and the range of 1st place, multiple types of range type key data included in each given search range are combined and read, and the range type key data is ORed to obtain search data. An information retrieval system characterized by: