JP2011197809A - Search apparatus, search method and search program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To perform efficient full-text search irrespective of characteristics of document data.SOLUTION: A search apparatus includes a storage part 110 for storing a pattern file 110a, a first index 110b and a second index 110c, a determination part 120, and a search part 130. The determination part 120 receives a search keyword, and determines whether to use the first index 110b or with the second index 110c for search, on the basis of the search keyword and the pattern file 110a. The search part 130 searches document data using the first index 110b or second index 110c, according to the result of determination by the determination part 120.

Description

  The present invention relates to a search device and the like.

  A full-text search for searching for a specific character string from a plurality of document data is known. In this full-text search, an inverted index is used. The transposed index corresponds to an index for storing position information of words included in character data. Methods for creating an inverted index are roughly divided into a character delimiter method and a word delimiter method.

  In the character delimiter method, an inverted index is created for each character without considering the meaning of the word. An inverted index created by the character delimiter method is referred to as a character index. If a character index is used, a complete partial match search is possible. However, it is necessary to compare the search keyword and the character index one character at a time, and there is a disadvantage that a long search time is required.

  In the word segmentation method, a transposed index is created for each meaningful word. The transposed index created by the word break method is expressed as a word index. If the word index is used, the search keyword is compared for each word, so that the search time can be shortened compared to the case where the character index is used. However, search omission may occur depending on how words are separated.

  As described above, since the character delimiter method and the word delimiter method have advantages and disadvantages, it is important how to properly use the character delimiter method and the word delimiter method. For example, a technique of automatically selecting a character index and a word index according to the length of a search keyword is disclosed as a conventional technique using a character partitioning system and a word partitioning system.

Japanese Patent Laid-Open No. 10-307835 JP 2001-34623 A JP 2008-77673 A

  However, depending on the document data for which full text search is performed, even when the length of the search keyword is the same, there are cases where efficient full text search can be performed by using different character indexes and word indexes.

  For example, document data stored in a biodatabase includes an ID (Identification) for linking to another database in addition to the document. In general, a word index is effective for document data having no symbol such as an ID, and a character index is effective for document data having a symbol.

  Here, a case where a search expression “1.1.1.1 ANDsuppressor” is given will be described as an example. When a full-text search is attempted using a character index for such a search expression, it is preferable to search only document data that includes the ID “1.1.1.1” and the word “suppressor”. However, if a full-text search is attempted using the character index for the above search expression, document data including IDs such as “1.1.1.11” and “1.1.1.12” will actually be hit.

  On the other hand, if a full-text search is attempted using the word index for the above search expression, only document data including the ID “1.1.1.1” can be searched. However, since “suppressors” and “suppressor” do not completely match, document data including “suppressors” cannot be searched.

  The disclosed technology has been made in view of the above, and an object thereof is to provide a search device, a search method, and a search program capable of efficiently performing a full-text search regardless of the characteristics of document data. .

  In one aspect, the search device disclosed in the present application is partitioned based on the first partitioning method, partitioned based on the first index associated with the document data, and based on the second partitioning method, and is stored in the document data. A second index that is associated, a storage unit that stores a pattern file that defines features of a predetermined character, a search keyword is received, and the first index is determined based on the search keyword and the pattern file. A determination unit that determines whether to use the second index to search for document data, and the first index or the first index based on the determination result of the determination unit. And a search unit for searching for document data using the index of No. 2.

  According to one aspect of the search device disclosed in the present application, there is an effect that a full-text search can be executed efficiently regardless of the characteristics of document data.

FIG. 1 is a diagram illustrating the configuration of the search device according to the first embodiment. FIG. 2 is a diagram illustrating the system according to the second embodiment. FIG. 3 is a diagram illustrating the configuration of the search device according to the second embodiment. FIG. 4 is a diagram showing the data structure of the pattern file. FIG. 5 is a flowchart of the process procedure of the search device according to the second embodiment. FIG. 6 is a diagram illustrating a hardware configuration of a computer that configures the search device according to the embodiment.

  Embodiments of a search device, a search method, and a search program disclosed in the present application will be described below in detail with reference to the drawings. Note that the present invention is not limited to the embodiments.

  FIG. 1 is a diagram illustrating the configuration of the search device 100 according to the first embodiment. As illustrated in FIG. 1, the search device 100 includes a storage unit 110, a determination unit 120, and a search unit 130.

  The storage unit 110 stores a pattern file 110a, a first index 110b, and a second index 110c. The pattern file 110a is data defining features of a predetermined character. The first index 110b is data that is partitioned based on the first partitioning method and associated with the document data. The second index 110c is data that is partitioned based on the second partitioning method and associated with the document data.

  The determination unit 120 receives a search keyword, and determines whether to perform a search using the first index 110b or the second index 110c based on the search keyword and the pattern file 110a. .

  Based on the determination result of the determination unit 120, the search unit 130 searches for document data using the first index 110b or the second index 110c.

  The search device 100 uses the pattern file 110a to determine whether to perform a search using the first index 110b or a search using the second index 110c. For this reason, since an optimal index can be selected according to the characteristics of the search keyword, a full-text search can be efficiently executed regardless of the characteristics of the document data.

  Next, an example of a system according to the second embodiment will be described. FIG. 2 is a diagram illustrating the system according to the second embodiment. As shown in FIG. 2, this system includes a user terminal 60 and a search device 200. The user terminal 60 and the search device 200 are connected via the network 50.

  The user terminal 60 is a device that transmits a search keyword to the search device 200 and receives a search result for the search keyword from the search device 200.

  The search device 200 is a device that performs full-text search of document data. FIG. 3 is a diagram illustrating the configuration of the search device 200 according to the second embodiment. As illustrated in FIG. 3, the search device 200 includes a storage unit 210, an indexing processing unit 220, an input receiving unit 230, a search expression analysis processing unit 240, a scoring processing unit 250, and a search result output unit 260.

  The storage unit 210 stores a pattern file 210a, a document data group 210b, a word index 210c, and a character index 210d.

  The pattern file 210a is data defining features of a predetermined character. FIG. 4 shows the data structure of the pattern file 210a. As shown in FIG. 4, this pattern file has No and a pattern. No identifies each pattern. The pattern represents the characteristics of a predetermined character with a regular expression. Here, it is assumed that the characters include numbers and symbols in addition to general characters.

  Here, an example of a pattern description method will be described. [] In the pattern means a letter, number, or symbol that matches any one of the characters written in [and]. For example, [0-9] means a single digit. {N, m} in the pattern means that the immediately preceding character is repeated from n to m times. For example, [0-9] {1, 3} means a 1-digit, 2-digit, 3-digit number.

Moreover, + in the pattern means that the immediately preceding character is repeated one or more times. For example, [0-9] + means a character string consisting of numbers. * In the pattern means that the immediately preceding character is repeated zero or more times. For example, [0-9] ^* means a character string consisting of an empty character or a number.

  Returning to the description of FIG. The document data group 210b includes a plurality of document data. Each document data is assigned a unique ID and includes various character strings.

  The word index 210c is a transposed index that associates the word of each document data included in the document data group 210b with the ID of the document data in which the word exists. The character index 210d is a transposed index in which characters of each document data included in the document data group 210b are associated with IDs of document data in which the characters exist.

  The indexing processing unit 220 is a processing unit that generates a word index 210c and a character index 210d from the document data group 210b. The indexing processing unit 220 generates a word index 210c from the document data group 210b by a word segmentation method. Further, the indexing processing unit 220 generates a character index 210d from the document data group 210b by a character delimiter method. The generation of the word index 210c by the word segmentation method is the same as the well-known word segmentation method. The generation of the character index 210d by the character delimiter method is the same as the well-known character delimiter method.

  The input receiving unit 230 receives a search keyword from the user terminal 60 and outputs the search keyword to the search formula analysis processing unit 240. Note that the input reception unit 230 may acquire a search keyword from an input device connected to the search device 200. The input device corresponds to, for example, a mouse or a keyboard.

  The search expression analysis processing unit 240 compares the search keyword with the pattern file 210a to determine whether to search for document data using the word index 210c or to search for document data using the character index 210d. It is a processing unit. In the following, searching for document data using the word index 210c is referred to as a word segmentation type search. Searching for document data using the character index 210d is referred to as a character delimiter search.

  First, the search expression analysis processing unit 240 performs syntax analysis on the search keyword. For example, the search keyword is “1.1.1.1 ANDsuppressor”. The search expression analysis processing unit 240 performs syntax analysis on the search keyword “1.1.1.1 ANDsuppressor”, so that the conditional sentence “AND” included in the search keyword and the character string “1.1.1.1” sandwiching the conditional sentence are included. ”And“ suppressor ”are extracted.

  The search expression analysis processing unit 240 compares each character string extracted from the search keyword with the pattern in the pattern file 210a, and performs a word delimiter search for each character string. Determine whether to perform.

  Specifically, the search formula analysis processing unit 240 determines that a character delimiter search is performed when a character string matches any of the patterns in the pattern file 210a. For example, the character string “1.1.1.1” matches the pattern “No 2” in the pattern file 210a shown in FIG. For this reason, the search expression analysis processing unit 240 determines to perform a word break search for the character string “1.1.1.1”.

  Further, the character string “suppressor” does not match the pattern of the pattern file 210a shown in FIG. For this reason, the search formula analysis processing unit 240 determines that the character string “suppressor” is to be searched by the character delimiter method.

  The search expression analysis processing unit 240 outputs data in which the character string is associated with the determination result to the scoring processing unit 250. In addition, the search expression analysis processing unit 240 also outputs the conditional sentence included in the search keyword to the scoring processing unit 250.

  The scoring processing unit 250 is a processing unit that acquires the character string, the determination result of the character string, and the conditional sentence of the search expression analysis processing unit 240 and searches the document data corresponding to the search keyword based on the acquired data. . Here, as an example, the determination result corresponding to the character string “1.1.1.1” is “search for the word separation method”, and the determination result corresponding to the character string “suppressor” is “search for the character separation method”. And the conditional statement is “AND”.

  In this case, the scoring processing unit 250 compares the character string “1.1.1.1” with the character index 210d, specifies the document data corresponding to the character string “1.1.1.1”, and specifies the specified document data Is obtained from the document data group 210b. Further, the scoring processing unit 250 compares the character string “suppressor” with the word index 210c, specifies the document data corresponding to the character string “suppressor”, and acquires the specified document data from the document data group 210b. To do.

  Then, since the conditional sentence is “AND”, the scoring processing unit 250 compares the document data corresponding to the character string “1.1.1.1” with the document data corresponding to the character string “suppressor”, and the duplicate document The data is output to the search result output unit 260. When the conditional statement is “OR”, the scoring processing unit 250 corresponds to the document data corresponding to the character string “1.1.1.1” and the character string “suppressor” because the conditional statement is “OR”. The document data to be output is output to the search result output unit 260.

  When the document data is searched, the scoring processing unit 250 may assign a score to the document data according to the frequency of the character string included in the document data.

  The search result output unit 260 notifies the user terminal 60 of the document data received from the scoring processing unit 250. The search result output unit 260 may adjust the order of the document data displayed on the user terminal 60 according to the score of the document data. The search result output unit 260 may output document data to a display device connected to the search device 200. The display device corresponds to, for example, a monitor or a liquid crystal display.

  Next, a processing procedure of the search device 200 according to the second embodiment will be described. FIG. 5 is a flowchart of the process procedure of the search device 200 according to the second embodiment. As shown in FIG. 5, the search device 200 acquires a search keyword (step S101) and executes syntax analysis (step S102).

  The search device 200 acquires an unselected pattern from the pattern file 210a (step S103), and determines whether or not the search keyword matches the pattern (step S104). When the search keyword matches the pattern (Yes at Step S104), the search device 200 determines that the word segmentation type search is performed (Step S105), and proceeds to Step S108.

  On the other hand, when the search keyword does not match the pattern (No at Step S104), the search device 200 determines whether there is an unselected pattern (Step S106). When there is an unselected pattern (step S106, Yes), the search device 200 proceeds to step S103.

  On the other hand, when there is no unselected pattern (No in step S106), the search device 200 determines that the character delimiter search is performed (step S107) and executes the search (step S108).

  As described above, the search device 200 according to the second embodiment uses the pattern file 210a to determine whether to perform a word delimiter search or a character delimiter search. For this reason, since an optimal index can be selected according to the characteristics of the search keyword, a full-text search can be efficiently executed regardless of the characteristics of the document data.

  In the second embodiment, when the search keyword is a search expression, the search expression is divided into a plurality of partial keywords, and a word-delimited search or a character-delimited search is performed for each partial keyword. Is judged. For this reason, a full-text search can be executed using a search formula that follows the existing technology as it is.

  Further, the pattern file 210a of the second embodiment can be easily customized according to the preference of each user.

  Incidentally, each component of the search device 200 shown in FIG. 3 is functionally conceptual and does not necessarily need to be physically configured as illustrated. In other words, the specific form of distribution / integration of the search device 200 is not limited to the illustrated one, and all or a part thereof is functionally or physically distributed in arbitrary units according to various loads or usage conditions. -Can be integrated and configured. For example, the storage unit 210 may be mounted on a removable external device or portable terminal, and the external device or portable terminal may be connected to the search device 200 by wire or wirelessly.

  The search device 200 can also be realized by mounting each function of the search device 200 on an information processing device such as a known personal computer, workstation, mobile phone, PHS terminal, mobile communication terminal, or PDA. .

  FIG. 6 is a diagram illustrating a hardware configuration of a computer that configures the search device according to the embodiment. As shown in FIG. 6, the computer 300 includes a CPU (Central Processing Unit) 301 that executes various arithmetic processes, an input device 302 that receives input of data from a user, and a monitor 303. The computer 300 includes a medium reading device 304 that reads a program and the like from a storage medium, and a network interface device 305 that exchanges data with other computers via a network. The computer 300 also includes a RAM (Random Access Memory) 306 that temporarily stores various information and a hard disk device 307. Each device 301 to 307 is connected to a bus 308.

  The hard disk device 307 stores a search program 307a having the same functions as the search expression analysis processing unit 240, the scoring processing unit 250, and the indexing processing unit 220 shown in FIG. Further, the hard disk device 307 stores various data 307b corresponding to the various data 210a to 210d shown in FIG.

  When the CPU 301 reads the search program 307a from the hard disk device 307 and develops it in the RAM 306, the search program 307a functions as the search process 306a. In addition, the CPU 301 reads various data 307 b into the RAM 306. The search process 306a performs a full text search using various data 306b.

  Note that the search program 307a is not necessarily stored in the hard disk device 307, and the computer 300 may read and execute a program stored in a storage medium such as a CD-ROM. Alternatively, the program may be stored in a public line, the Internet, a LAN (Local Area Network), a WAN (Wide Area Network), or the like, and the computer 300 may read and execute the program therefrom.

  The following supplementary notes are further disclosed with respect to the embodiments including the above examples.

(Supplementary Note 1) A first index that is partitioned based on the first partitioning scheme and associated with the document data, and a second index that is partitioned based on the second partitioning scheme and associated with the document data; A storage unit for storing a pattern file defining characteristics of a predetermined character;
Whether to search for document data using the first index or to search for document data using the second index based on the search keyword and the pattern file. A determination unit for determining;
A search device, comprising: a search unit that searches for document data using the first index or the second index based on a determination result of the determination unit.

(Additional remark 2) It further has the keyword division part which divides | segments the said search keyword into a some partial keyword, The said determination part searches document data using the said 1st index for every partial keyword, The search apparatus according to appendix 1, wherein it is determined whether to search for document data using the second index.

(Additional remark 3) The said 1st division | segmentation system is a word division | segmentation system which divides | segments a character string for every meaningful word, The said determination part is when the characteristic defined in the said pattern file corresponds with the said search keyword. The search apparatus according to appendix 1 or 2, wherein it is determined that the document data is searched using the first index.

(Supplementary Note 4) A first index that is partitioned based on the first partitioning scheme and associated with the document data, and a second index that is partitioned based on the second partitioning scheme and associated with the document data; A search device having a storage device for storing a pattern file defining features of a predetermined character,
Whether to search for document data using the first index or to search for document data using the second index based on the search keyword and the pattern file. A determination step for determining;
And a search step of searching for document data using the first index or the second index based on the determination result of the determination step.

(Additional remark 5) It further includes the keyword division | segmentation step which divides | segments the said search keyword into a some partial keyword, In the said determination step, search of document data is performed using said 1st index for every partial keyword. 5. The search method according to appendix 4, wherein it is determined whether or not the document data is searched using the index of 2.

(Additional remark 6) The said 1st division | segmentation system is a word division | segmentation system which divides | segments a character string for every meaningful word, and when the characteristic defined in the said pattern file corresponds with the said search keyword in the said determination step. The search method according to appendix 4 or 5, wherein it is determined that the document data is searched using the first index.

(Supplementary Note 7) A first index that is partitioned based on the first partitioning scheme and associated with the document data, and a second index that is partitioned based on the second partitioning scheme and associated with the document data; A computer having a storage device for storing a pattern file defining the characteristics of a predetermined character,
Whether to search for document data using the first index or to search for document data using the second index based on the search keyword and the pattern file. A determination procedure for determining;
And a search procedure for executing a search for document data using the first index or the second index based on a determination result of the determination step.

(Supplementary Note 8) Whether or not the computer further executes a keyword dividing procedure for dividing the search keyword into a plurality of partial keywords, and the determination procedure searches the document data using the first index for each partial keyword. The search program according to appendix 7, wherein it is determined whether or not to search for document data using the second index.

(Additional remark 9) The said 1st division | segmentation system is a word division | segmentation system which divides | segments a character string for every word with a meaning, and the said determination procedure is a case where the characteristic defined in the said pattern file corresponds with the said search keyword. The search program according to appendix 7 or 8, wherein it is determined that the document data is searched using the first index.

DESCRIPTION OF SYMBOLS 100 Search apparatus 110a Pattern file 110b 1st index 110c 2nd index 120 Determination part 130 Search part

Claims (5)

A first index associated with the document data, and a second index associated with the document data and associated with the document data; and a predetermined character A storage unit for storing a pattern file defining features;
Whether to search for document data using the first index or to search for document data using the second index based on the search keyword and the pattern file. A determination unit for determining;
A search device, comprising: a search unit that searches for document data using the first index or the second index based on a determination result of the determination unit.   It further includes a keyword dividing unit that divides the search keyword into a plurality of partial keywords, and the determination unit searches the document data using the first index for each partial keyword, or the second index. The search apparatus according to claim 1, wherein it is determined whether or not document data is to be searched.   The first delimiter is a word delimiter that delimits a character string for each meaningful word, and the determination unit determines that the feature defined in the pattern file matches the search keyword. The search apparatus according to claim 1, wherein it is determined that the document data is searched using the first index. A first index associated with the document data, and a second index associated with the document data and associated with the document data; and a predetermined character A search device having a storage device for storing a pattern file defining features,
Whether to search for document data using the first index or to search for document data using the second index based on the search keyword and the pattern file. A determination step for determining;
And a search step of searching for document data using the first index or the second index based on the determination result of the determination step. A first index associated with the document data, and a second index associated with the document data and associated with the document data; and a predetermined character A computer having a storage device for storing a pattern file defining features;
Whether to search for document data using the first index or to search for document data using the second index based on the search keyword and the pattern file. A determination procedure for determining;
And a search procedure for executing a search for document data using the first index or the second index based on a determination result of the determination step.

Images