JP2006012000A - Method, device for retrieving document, computer program and computer readable recording medium - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To make character strings to be a keyword easy to be visually observed in the case of displaying a result by performing vicinity retrieval. <P>SOLUTION: When two or more combinations of a starting keyword A and a termination keyword B (combination constituting of a starting keyword A whose positional information is "11" and a termination keyword B whose positional information is "20" and combination consisting of a starting keyword A whose positional information is "15" and a termination keyword B whose positional information is "20") are made within a range of the number of characters (10 characters) between keywords, combination in which the starting keyword A and the termination keyword B are approximated most (the combination consisting of the starting keyword A whose positional information is "15" and the termination keyword B whose positional information "20") is adopted as the optimal combination 601b among the two or more combinations and the starting keyword A and the termination keyword B of the optimal combination 601b are enhanced and displayed on a screen. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a document search method, a document search device, a computer program, and a computer-readable recording medium, and is particularly suitable for use in searching for a document file.

  Conventionally, various methods for searching a document file stored in a document database have been proposed. For example, a mechanism for indexing a specific keyword and using the indexed keyword as a key to search for a target document at high speed, or a mechanism for searching a document by performing a full text search (full text search) has been proposed. Yes. In addition, a mechanism has been proposed in which a user specifies a text search condition and the text search condition is searched from a document.

  As described above, various methods have been conventionally employed as a mechanism for searching for a document. In recent years, a so-called neighborhood search is a mechanism for performing a search for a document intended by a user at a higher speed and appropriately. It has been put into practical use. In such a proximity search, first, the user designates two keywords to be searched and the number of characters between the two keywords to the document search apparatus. Then, the document search apparatus searches for a keyword set including two keywords specified by the user. Then, from the searched keyword set, a keyword set in which the number of characters between the keyword sets is within the number of characters specified by the user can be extracted, and the extracted keyword set can be distinguished from others. To highlight. In this way, it is possible to perform a search that takes into account the semantic connection between keywords.

As a technology applying such neighborhood search, Patent Document 1 discloses individual matching information indicating a matching result (search result) of a character string when searching for a complex condition including a neighborhood search with one microcomputer. A mechanism is disclosed that increases processing efficiency by outputting only to programs to be processed (neighboring condition determination program, context condition determination program, logical condition determination program).
Patent Document 2 emphasizes keywords specified by the user by highlighting them according to search conditions (exact match search, neighborhood search, different notation expansion search, etc.) while making them different from other text data. A mechanism is disclosed that allows a user to visually recognize which keyword corresponds to which search condition by changing the display method.

JP-A-6-68152 Japanese Patent Laid-Open No. 10-269233

  However, in the conventional neighborhood search mechanism, all keyword pairs within the range of the number of characters specified by the user are extracted and highlighted so that the extracted keyword pairs can be distinguished from others. For this reason, when there are a plurality of keyword sets within the range of the number of characters specified by the user, all of the plurality of keyword sets are highlighted. Therefore, the conventional technique has a problem that it is difficult for the user to confirm the optimum combination pattern of keywords.

  Further, when the user designates a set of a plurality of keywords as a search condition for neighborhood search, the display becomes even more complicated. For example, A @ 20B (start keyword A and end keyword B exist within 20 characters) and C @ 5D (start keyword C and end keyword D exist within 5 characters) are set as search conditions by the user. When specified, the conventional technique highlights all keyword pairs that match the search condition on the screen as search results. For this reason, for example, there are a large number of keywords that need to be highlighted in a narrow range, such as when both or one of the start keyword C and the end keyword D is included between the start keyword A and the end keyword B. In this case, even if the display color of the corresponding keyword is changed or the corresponding keyword is underlined, the set of the start keyword A and the end keyword B to be noticed, and the set of the start keyword C and the end keyword D Is extremely difficult for the user to identify.

  The present invention has been made in view of the above-described problems, and an object of the present invention is to make it easy to visually recognize a character string as a keyword when displaying a result of neighborhood search.

  The document search method of the present invention is a document search method for searching a document file according to a neighborhood search condition including a plurality of character strings and the existence ranges of the plurality of character strings, and is included in the neighborhood search condition. A search step that searches the document file for a character string that matches a plurality of character strings, and whether the plurality of character strings searched by the search step are within an existence range included in the neighborhood search condition A character string extraction step for extracting a character string to be highlighted based on a result of the determination, and a display step for highlighting the character string extracted by the character string extraction step, The character string extraction step is performed when there are a plurality of character string sets including a plurality of character strings searched in the search step within the existence range included in the neighborhood search condition. Of the set number of strings, and extracts a set of strings that is closest, as a character string to be the highlight.

  The document search device of the present invention is a document search device that searches a document file according to a neighborhood search condition including a plurality of character strings and the existence ranges of the plurality of character strings, and is included in the neighborhood search condition. A search unit that searches the document file for a character string that matches a plurality of character strings, and whether the plurality of character strings searched by the search unit are within an existence range included in the neighborhood search condition A character string extracting unit that extracts a character string to be highlighted based on the determination result, and a highlight display for highlighting the character string extracted by the character string extracting unit Display information creating means for creating information, and the character string extracting means includes a plurality of character strings searched by the search means within the existence range included in the neighborhood search condition. Multiple pairs If the standing, of the set of the plurality of character strings, and extracts a set of strings that is closest, as a character string to be the highlight.

A computer program of the present invention is a computer program for causing a computer to perform a search for a document file according to a neighborhood search condition including a plurality of character strings and the existence ranges of the plurality of character strings, A search step for searching the document file for a character string that matches a plurality of character strings included in the neighborhood search condition, and a plurality of character strings searched by the search step are included in the neighborhood search condition. A character string extraction step for determining whether the character string is within the existing range, and extracting a character string to be highlighted based on the determined result, and a display step for highlighting the character string extracted by the character string extraction step And the character string extraction step executes the search string within the existence range included in the neighborhood search condition. When there are a plurality of character string pairs made up of a plurality of character strings searched in the pop-up, the closest character string set of the character string pairs is selected as the character string to be highlighted. It is characterized by extracting.
A computer-readable recording medium according to the present invention records the computer program.

  According to the present invention, it is possible to prevent as much as possible a large number of character strings to be highlighted and displayed in a narrow range. Therefore, the user can identify a noteworthy character string at a glance.

Next, embodiments of the present invention will be described with reference to the drawings.
FIG. 1 is a diagram showing an example of a schematic configuration of a document search system according to an embodiment of the present invention.
1, the document search system includes a document search apparatus 101, a network 105, and a client terminal 106. The document search apparatus 101 and the client terminal 106 are connected to be able to communicate with each other via the network 105. Configured.

The document search apparatus 101 is a server for performing a neighborhood search for document files. The document search apparatus 101 according to this embodiment includes a neighborhood search / display program 102, a document database (document DB) 103, and a full-text index 104.
The document database 103 stores a plurality of document files to be subjected to proximity search. In the present embodiment, the document file is based on a standard generalized markup language (SGML) format, an extensible markup language (XML) format, or the like. However, it goes without saying that the format of the document file is not limited to these.

The neighborhood search / display program 102 performs neighborhood search on the document file stored in the document database 103, edits the corresponding document file so that the user can determine the result of the neighborhood search, and creates the edited document file. It is a program for creating display document data based on the above.
The full-text index 104 is for speeding up the neighborhood search, and stores the position information of the keyword designated by the user in the neighborhood search.

The client terminal 106 is a computer device owned by the user, and is equipped with a program or a Web browser for inputting neighborhood search conditions and displaying neighborhood search results. More specifically, the client terminal 106 transmits a neighborhood search condition input based on a user's operation of a keyboard, a mouse, or the like to the document search device 101 using a common gateway interface (CGI) or the like. Thereafter, display document data including the result of the proximity search performed by the document search apparatus 101 as described later is received from the document search apparatus 101 as an HTML format data file and displayed by the browser application.
The network 105 may be anything such as the Internet or a LAN (Local Area Network) as long as it enables communication between the document search apparatus 101 and the client terminal 106.

FIG. 2 is a diagram illustrating an example of a hardware configuration of the document search apparatus 101.
The document search apparatus 101 includes a CPU 201, a RAM 202, a ROM 203, a system bus 204, an input controller 205, a video controller 206, a memory controller 207, a communication I / F controller 208, a keyboard 209, a CRT (cathode). ray tube) display (CRT) 210 and external memory 211.
The CPU 201 comprehensively controls the devices 202 and 203 connected to the system bus 204 and the controllers 205 to 208.

  The ROM 202 or the external memory 211 has various basic functions necessary for realizing the functions executed by the BIOS (Basic Input / Output System), the operating system program (OS), and the document search device 101, which are control programs of the CPU 201. Programs and so on are stored.

  The RAM 202 functions as a main memory, work area, and the like for the CPU 201. The CPU 201 implements various operations by loading a program necessary for execution of processing into the RAM 202 and executing it.

  The input controller 205 controls input from a pointing device such as a keyboard 209 or a mouse (not shown). The video controller 206 controls display on the CRT display 210. The input controller 205 and the video controller 206 perform control in cooperation with the CPU 201. For example, when the CPU 201 executes outline font rasterization processing on a display information memory area in the RAM 203, for example, display on the CRT display 210 becomes possible. In addition, a user instruction with a mouse cursor (not shown) on the CRT display 210 can be performed. In place of the CRT display 210, a liquid crystal display or the like may be used as a display device.

  The external memory 211 stores a boot program, browser software, various applications, font data, user files, editing files, various data, and the like. The external memory 211 is, for example, a hard disk (HD), a flexible disk (FD), or a compact flash (registered trademark) memory connected to a PCMCIA card slot via an adapter.

  In the present embodiment, the neighborhood search / display program 102 shown in FIG. 1 is recorded in the external memory 211 and loaded into the RAM 202 as necessary to be executed by the CPU 201. The document database 103 and the full text index 104 are also recorded in the external memory 211.

The memory controller 207 controls access to the external memory 211. The communication I / F controller 208 is for communicating with an external device such as the client terminal 106 via the network 105, and executes data communication control processing via the network 105. The communication I / F controller 208 executes, for example, Internet communication using TCP / IP.
Since the hardware configuration of the client terminal 106 is the same as that shown in FIG. 2, detailed description of the client terminal 106 is omitted.

Here, an outline of an operation when performing a neighborhood search in the document search system of the present embodiment will be described.
First, the user operates the client terminal 106 to specify a neighborhood search condition including a start keyword and an end keyword, and the number of characters between the start keyword and the end keyword (hereinafter abbreviated as the number of characters between keywords). Then, the client terminal 106 transmits information indicating the vicinity search condition to the document search apparatus 101.

  The document search apparatus 101 that has received the information including the neighborhood search condition searches the document file and derives an optimum combination from the combinations of the start keyword and end keyword specified by the user. Then, the document search apparatus 101 creates display document data for highlighting the start keyword and the end keyword of the optimal combination, and transmits the display document data to the client terminal 106. The client terminal 106 that has received the display document data highlights the start keyword and the end keyword of the optimal combination based on the display document data.

In particular, in the document search apparatus 101 of the present embodiment, when there are a plurality of combinations of start keywords and end keywords in the number of characters between keywords specified by the user as a result of searching for a document file, The closest combination is set as the optimum combination.
In the neighborhood search performed in the present embodiment, when the start keyword and the end keyword corresponding to the neighborhood search condition are searched, the start keyword and the end keyword are used until the neighborhood search based on the next neighborhood search condition is performed. It will not be reused.

  Hereinafter, an example of an operation when performing a neighborhood search in the document search system of this embodiment will be described with reference to the flowcharts of FIGS. 3, 4, and 6. Note that the processing based on the flowcharts of FIGS. 3, 4, and 6 is mainly executed by the neighborhood search / display program 102 stored in the document search apparatus 101.

FIG. 3 is a flowchart for explaining an example of a schematic operation when performing a neighborhood search in the document search system. FIG. 4 is a flowchart following FIG.
First, when the user inputs a neighborhood search condition using the search condition designation screen 901 (see FIG. 5A) displayed on the display device of the client terminal 106, the client terminal 106 indicates the neighborhood search condition. Information is transmitted to the document search apparatus 101. In FIG. 5A, the first neighborhood search condition (FW = A @ 10B) in which the start keyword is “A”, the end keyword is “B”, and the number of characters between keywords is “10”, and the start keyword is “ In this example, a neighborhood search condition including a second neighborhood search condition (FW = C @ 6D) of “C”, the end keyword “D”, and the number of characters between keywords is “6” is shown.

  When the information indicating the proximity search condition is transmitted from the client terminal 106 in this way, the CPU 201 of the document search apparatus 101 inputs the proximity search condition via the communication I / F controller 208 (step S301), and the external memory The inputted neighborhood search condition is stored in the neighborhood search condition table 902 (see FIG. 5B) stored in 211 etc. (step S302). As described above, the neighborhood search condition includes a start keyword and an end keyword, and the number of characters between keywords (specified distance). However, it goes without saying that conditions other than these may be included as neighborhood search conditions.

  Thus, when the neighborhood search condition is stored in the neighborhood search condition table 902, the CPU 201 selects one of the neighborhood search conditions stored in the neighborhood search condition table 902 and matches the start keyword of the selected neighborhood search condition. Information on the position where the keyword to be present (hereinafter referred to as start keyword position information) is acquired from the full-text index 104 and stored in the RAM 202 (step S303). In this embodiment, the number of characters from the beginning of the document file to the first character of the start keyword is used as start keyword position information.

  In the example of the neighborhood search condition shown in FIG. 5, as shown in FIG. 7, the start keyword A of the first neighborhood search condition (A @ 10B) is the second character, the eleventh character from the top of the document file, Since they are in the 15th character, the 34th character, etc., these numbers are stored in the RAM 202 in the ascending order as the start keyword position information 501a. In addition, since the start keyword C of the second neighborhood search condition (C @ 6D) is in the fourth character, the 32nd character, etc. from the top of the document file, these numbers are in ascending order as the start keyword position information 501b. Stored in the RAM 202.

  Then, the CPU 201 confirms whether or not a keyword matching the start keyword exists in the document file in the process of step S303 (step 304). As a result of the confirmation, if the keyword that matches the start keyword does not exist in the document file, the CPU 201 refers to the neighborhood search condition table 902 to determine whether or not there is a next neighborhood search condition (step S305).

  As a result of this determination, if the next neighborhood search condition does not exist, the process is terminated, and if it exists, a process of acquiring start keyword position information for the next neighborhood search condition is performed (step 303).

  Thus, when it is confirmed in step S304 that a keyword that matches the start keyword exists in the document file, the CPU 201 determines that a keyword that matches the end keyword of the neighborhood search condition selected in step S303 exists. Information related to the above (hereinafter referred to as end keyword position information) is acquired from the full-text index 104 and stored in the RAM 202 (step S306). In the present embodiment, the number of characters from the beginning of the document file to the first character of the end keyword is used as end keyword position information.

  In the example of the neighborhood search condition shown in FIG. 5, as shown in FIG. 7, the end keyword B of the first neighborhood search condition (A @ 10B) is the eighth character, the 20th character from the beginning of the document file, Since it is in the 40th character, etc., these numbers are stored in the RAM 202 as the end keyword position information 502a in ascending order. Further, since the end keyword D of the second neighborhood search condition (C @ 6D) is at the ninth character, the twelfth character, the 37th character, etc. from the top of the document file, these numbers are in ascending order in the start keyword. The position information 502b is stored in the RAM 202.

  In step S306, the CPU 201 confirms whether a keyword that matches the end keyword exists in the document file (step S307). As a result of the confirmation, if the keyword that matches the end keyword does not exist in the document file, the CPU 201 refers to the neighborhood search condition table 902 and determines whether or not there is a next neighborhood search condition (step S308). ).

As a result of this determination, if the next neighborhood search condition does not exist, the process is terminated, and if it exists, a process of acquiring start keyword position information for the next neighborhood search condition is performed (step 303).
Then, the CPU 201 sets the start keyword position information and the end keyword position information in the order from the smallest position information for the document file in which the start keyword position information and the end keyword position information are acquired as described above. Merging (merging) is performed in ascending order, and the merged position information (hereinafter referred to as merge position information) is stored in the RAM 202 (step 309).

  Then, the CPU 201 refers to the neighborhood search condition table 902 to determine whether or not there is a next neighborhood search condition (step S310), and if so, for all the neighborhood search conditions in the neighborhood search condition table 902, Until the merge position information is created, the processes in steps S303 to S310 are repeated.

  In the example of the neighborhood search condition shown in FIG. 5, as shown in FIG. 7, first, start keyword position information 501a and start keyword “A” in the first neighborhood search condition are associated with each other, and end keyword position information 502a. Is associated with the end keyword “B”. The start keyword position information 501a associated with the start keyword “A” and the end keyword position information 502a associated with the end keyword “B” are merged in ascending order from the smallest position information. Position information 503 is created.

  Further, in the example of the neighborhood search condition shown in FIG. 5, since there is the second neighborhood search condition in addition to the first neighborhood search condition, start keyword position information 501b in the second neighborhood search condition and The start keyword “C” is associated with the end keyword position information 502b and the end keyword “D”. The start keyword position information 501b associated with the start keyword “C” and the end keyword “D” are associated with each other. The associated end keyword position information 502b and merge position information 503 are merged in ascending order from the smallest position information to create final merge position information 504 used for neighborhood search. Thus, in the present embodiment, only one merge position information is created even when a plurality of neighborhood search conditions exist.

  Next, the CPU 201 performs an adjacent position acquisition process in merge information using the merge position information 504 created in step S309 (step S311). In this merge information adjacent position acquisition process, first, the CPU 201 reads the merge position information 504 from the RAM 202 and checks whether there are a plurality of combinations of start keywords and end keywords within the number of characters between keywords. As a result of this check, if there are multiple combinations of the start keyword and the end keyword within the number of characters between keywords, the closest start keyword and end keyword among the combinations of these start keywords and end keywords Is the optimal combination in the neighborhood search. On the other hand, when there is only one combination of the start keyword and the end keyword within the number of characters between keywords, the combination is determined as the optimum combination in the neighborhood search.

  In the example of the first neighborhood search condition (A @ 10B) shown in FIG. 5, as shown in FIG. 8, the start keyword position information is the number of characters between keywords from the start keyword “A” with “2”. An end keyword “B” having end keyword position information “8” exists at a position within 10 ”characters. Further, an end keyword “B” having end keyword position information “40” exists at a position within “10” characters from the start keyword “A” having start keyword position information “34”. Therefore, the CPU 201 sets the combination 601a and 601c as an optimal combination, and sets flags for the position information of the start keyword “A” and the position information of the start keyword “B” in the combinations 601a and 601c.

  In contrast, the start keyword “A” with the start keyword position information “15” and the end keyword position information are at positions within the “10” characters from the start keyword “A” with the start keyword position information “11”. There is an end keyword “B” of “20”. Then, a combination of the start keyword “A” whose start keyword position information is “11” and the end keyword “B” whose end keyword position information is “20”, and the start keyword “A” whose start keyword position information is “15”. ”And a start keyword“ B ”whose end keyword position information is“ 20 ”.

  As described above, when there are two combinations of the start keyword “A” and the end keyword “B” in an area within “10” characters, which is the number of characters between keywords, when all these keywords are highlighted, Many keywords are highlighted in a narrow range. Then, it may be extremely difficult for the user to determine which keyword should be focused on. Therefore, in the present embodiment, the CPU 201 sets the combination 601b in which the start keyword “A” and the start keyword “B” are closer to each other among these two combinations as the optimum combination, and the start keyword in the combination 601b. A flag is set for the position information of “A” and the position information of the start keyword “B”.

  Similarly, in the example of the second neighborhood search condition (C @ 6D) shown in FIG. 5, as shown in FIG. 9, the characters “6” to “6” from the start keyword “C” where the start keyword position information is “4”. The end keyword “D” whose end keyword position information is “9” is present at the position within. Further, an end keyword “D” having end keyword position information “37” exists at a position within “6” characters from the start keyword “C” having start keyword position information “32”. Therefore, the CPU 201 sets the combination 701a, 701b as an optimal combination, and sets a flag for the position information of the start keyword “C” and the position information of the start keyword “D” in the combination 701a, 701b.

As described above, after extracting the combinations 601a to 601c, 701a, and 701b optimum for the neighborhood search, the CPU 201 acquires the document data for display of the corresponding document file from the document database 103 (step S312).
Next, the CPU 201 acquires the position information for which the flag is set in the process of step S311 and edits the display document data (step S313). Specifically, the CPU 201 displays the document data for display so that the background of characters between the start keyword and the end keyword extracted as the optimal combinations 601a to 601c, 701a, and 701b is displayed in gray. To edit.

  Next, the CPU 201 edits the display document data so that the start keyword and the end keyword extracted as the optimal combinations 601a to 601c, 701a, and 701b are white characters (step S314).

  Finally, the CPU 201 transmits the display document data edited in steps S313 and S314 to the client terminal 106 via the communication I / F controller 208. The client terminal 106 that has received the display document data displays the document on the display device so that the start keyword and the end keyword extracted as the optimum combination are emphasized based on the display document data. Details of document highlighting performed as described above will be described later with reference to FIG.

Details of the merge information adjacent position acquisition processing in step S311 of FIG. 4 will be described with reference to the flowchart of FIG.
First, the CPU 201 acquires one neighborhood search condition stored in the neighborhood search condition table 902 (step S401). Then, the CPU 201 sets a search pointer at the head of the merge position information 504 created in step S309 in FIG. 3 (step S402), and the keyword indicated by the search pointer satisfies the neighborhood search condition acquired in step S401. It is determined whether or not the start keyword is included (step S403).

  If these keywords do not match as a result of the determination, the CPU 201 moves the position of the search pointer to the next keyword (step S404), and determines whether the position of the search pointer is the last of the merge position information 504. (Step S405). If the result of this determination is that the position of the search pointer is the last of the merge position information 504, the CPU 201 refers to the neighborhood search condition table 902 to determine whether or not the next neighborhood search condition exists (step S416). . If the result of this determination is that the next neighborhood search condition exists, the search pointer is returned to the beginning of the merge position information 504, and the next neighborhood search condition is acquired (step S401).

  In step S403, when the keyword indicated by the search pointer matches the start keyword acquired in step S401, the CPU 201 stores the position information of the search pointer in the RAM 202 as start keyword position information (step S406). ), The search pointer is moved to the next keyword (step S407). At this time, the CPU 201 determines whether or not the position of the search pointer is the last of the merge position information 504 (step S408), and if the position of the search pointer is the last of the merge position information 504, the neighborhood search condition table 902 is stored. With reference to this, it is determined whether or not the next neighborhood search condition exists (step S416).

  On the other hand, if it is determined in step S408 that the position of the search pointer is not the end of the merge position information 504, the CPU 201 determines whether or not the keyword indicated by the search pointer matches the start keyword acquired in step S401 ( Step S409). As a result of this determination, if these keywords match, the CPU 201 rewrites the recorded contents of the RAM 202 and updates the position information of the search pointer (step S406).

  On the other hand, if the keyword indicated by the search pointer does not match the start keyword included in the neighborhood search condition acquired in step S401, the CPU 201 determines that the keyword indicated by the search pointer is the neighborhood search acquired in step S401. It is determined whether or not the end keyword included in the condition matches (step S410). If these keywords do not match as a result of this determination, the CPU 201 moves the search pointer to the next keyword (step S407).

  On the other hand, when the keyword indicated by the search pointer matches the end keyword included in the neighborhood search condition acquired in step S401, the CPU 201 stores the current search pointer position information in the RAM 202 as end keyword position information. (Step S411).

  In the example shown in FIG. 8, for example, since there is a start keyword A whose start keyword information is “2” at the beginning of the merge position information 504, the start keyword position information (“2”) is stored in the RAM 202 (step S406). Then, the merge position information 504 is sequentially checked until the start keyword A or the end keyword B is found. Then, since the next keyword to be found is the end keyword B whose end keyword position information is “8”, the end keyword position information (“8”) is stored in the RAM 202 (step S411).

  As described above, when the start keyword position information is stored in step S406 and the end keyword position information is stored in step S411, the CPU 201 determines the distance between the start keyword position information and the end keyword position information ( The number of characters is determined, and it is determined whether the determined distance (number of characters) is equal to or less than the number of characters between keywords included in the neighborhood search condition acquired in step S401 (step S412).

  In the example shown above, the distance (number of characters) between the start keyword position information “2” and the end keyword position information “8” is “6”, which is equal to or less than “10” which is the number of characters between keywords. . Therefore, the start keyword A whose start keyword position information is “2” and the end keyword B whose end keyword position information is “8” are an optimal combination.

  When the distance between the start keyword position information and the end keyword position information is equal to or less than the number of characters between keywords as a result of the determination in step S412, the CPU 201 stores the start keyword position information stored in step S406. Then, the merge position information 504 is marked so that the end keyword position information stored in step S411 can be identified (step S413).

  Then, the position of the search pointer is moved to the next keyword, and it is determined whether or not the position of the moved search pointer is the last of the merge position information 504 (step S415). If the result of this determination is that the position of the moved search pointer is not the end of the merge position information 504, the processing of steps S403 to S415 is repeated until the position of the search pointer reaches the end of the merge position information 504.

  In the example of FIG. 8 shown above, the end keyword B whose end keyword position information is “8” is not the last keyword of the merge position information 504, so the next keyword is checked. Since the start keyword A whose start keyword position information is “11” is found next to the end keyword B whose end keyword position information is “8”, the start keyword position information (“11”) is stored in the RAM 202. Store (step S406).

  Then, the merge position information 504 is sequentially checked until the start keyword A or the end keyword B is found. Then, since the next keyword to be found is the start keyword A whose start keyword position information is “15”, the start keyword position information stored in the RAM 202 is rewritten from “11” to “15” (Yes in step S409). Step S406). Thereafter, when the check of the merge position information 504 is resumed in the same manner, the next keyword to be found is the end keyword B whose end keyword position information is “20”, so this end keyword position information (“20”) is stored in the RAM 202. Store (step S411). The distance between the stored start keyword position information “15” of the start keyword A and the end keyword position information “20” of the end keyword B is “5”, and the number of characters between keywords is “10” or less. is there. Therefore, the start keyword A whose start keyword position information is “15” and the end keyword B whose end keyword position information is “20” are an optimal combination.

  Thus, in the present embodiment, for example, when there are a plurality of sets of the start keyword A and the end keyword B within the range of the inter-keyword character count “10” included in the neighborhood search condition (start keyword position information) A combination of a start keyword A with "11" and an end keyword B with end keyword position information "20", a start keyword A with start keyword position information "15", and an end keyword position information "20" Combinations with the end keyword B), and combinations that are closer to each other among these combinations are set as the optimum combinations.

  On the other hand, if the position of the moved search pointer is the last of the merge position information 504 in step S415, the CPU 201 refers to the neighborhood search condition table 902 to determine whether or not the next neighborhood search condition exists. (Step S416), and the processing of steps S401 to 416 is performed until processing is performed for all the neighborhood search conditions. In this way, when all the neighborhood search conditions have been processed, the process proceeds to step S312 in FIG. 4 (returns).

Next, an example of a method for displaying the result of the proximity search performed by the document search apparatus 101 as described above on the display device of the client terminal 106 will be described. FIG. 10 is a diagram showing an example of a document display screen in which the result of the neighborhood search is reflected. Specifically, the display screen when the neighborhood search condition is “connection @ 10 means” and “site @ 40 communication” is shown.
In the area 801 of the display screen 800, the start keyword “connection” and the end keyword “means” are displayed as white characters, and the background of the characters between these “connection” and “means” is grayed out. Is displayed. In the area 803, the start keyword “site” and the end keyword “communication” are displayed in white characters, and the background of the characters between the “site” and “communication” is displayed in gray. is doing.

  As a method of realizing such a display, the position information of the start keyword that is the optimum combination, the position information of the end keyword, and the number of characters between the start keyword and the end keyword (the number of characters between keywords) Each time it is acquired from the merge position information 504. Then, based on the acquired information, it is determined whether there is another optimum combination among the optimum combinations. Then, display of the display screen 800 is controlled as follows according to the determined result.

  In the display screen 800 shown in FIG. 10, there is only one optimum combination in the areas 801 and 802. In this case, the position after the acquired start keyword position information by the number of characters of the start keyword is set as the first color designation start point. Also, the position information of the end keyword that is the optimal combination is acquired from the merge position information 504, and the position one character before the position information of the end keyword is set as the first color designation end point. Then, the background in the range from the first color designation start point to the first color designation end point is displayed in gray.

  Further, the position information of the acquired start keyword is set as a second color designation start point. Further, the position after the number of characters obtained by subtracting 1 from the number of characters of the start keyword from the position information of the start keyword is set as the second color designation end point. Then, characters in the range from the second color designation start point to the second color designation end point are displayed in white.

  On the other hand, in the area 802 of the display screen 800, a combination of “connection” and “means” exists between the combination of “site” and “communication”. In this case, first, the start keyword (“connection”) and the end keyword (“means”) located inside are excluded from the target of color designation, and the combination of “site” and “communication” located outside is As described above, the background of characters between “site” and “communication” is displayed in gray, and “site” and “communication” are displayed in white. After that, the start keyword (“connection”) and the end keyword (“means”) located inside are displayed in white as described above, and the start keyword (“connection”) located inside is the end keyword. The background of the characters up to ("means") is returned to the normal background color.

The method for highlighting the start keyword and the end keyword is not limited to the method shown in FIG. For example, not only the background color of the character between the start keyword and the end keyword, but also the background color of the start keyword and the end keyword may be displayed in gray.
Also, the display method may be different for each combination of the start keyword and the end keyword. Specifically, for example, the start keyword (“connection”) and the end keyword (“means”) located inside are displayed in red, and the start keyword (“site”) and end keyword (“site”) located outside ( The combination with “communication”) is displayed in blue.

  In addition, in order to improve the visibility, the start keyword (“connection”) and the end keyword (“means”) located inside are kept excluded from the color designation, and the “site” located outside is designated as “site”. Only the combination with “communication” may be highlighted. Furthermore, the combination of the start keyword (“site”) and the end keyword (“communication”) located outside is excluded from the color specification target, and only the combination of “connection” and “means” located inside is emphasized. You may make it display. In other words, when there are a plurality of neighborhood search conditions, the display area of the search result by the neighborhood search may be inclusive as in the area 802. Visibility can be further improved by controlling the display so that the display by “communication” is different from the display by “connection” and “means” existing inside.

  As described above, in the present embodiment, when there are two or more combinations of the start keyword and the end keyword within the range of the number of characters between keywords, the combination of the two or more start keywords and the end keyword Among them, the combination in which the start keyword and the end keyword are closest to each other is regarded as the optimum combination, and the start keyword and the end keyword of the optimum combination are highlighted on the display screen 800. It is possible to prevent the displayed keywords from existing in a narrow range as much as possible. As a result, the user can identify at a glance keywords that are strongly linked (noticeable). Therefore, by using the document search system of the present embodiment, for example, it is possible to improve the accuracy and efficiency of prior art searches and rights searches when filing patent applications as much as possible.

  Also, one merge position information 504 is created for the start keyword and end keyword searched based on a plurality of neighborhood search conditions, and the optimum combination of the start keyword and end keyword is created using the created merge position information 504. Therefore, the optimal combination of the start keyword and the end keyword can be extracted at a higher speed than when the merge position information is created for each neighborhood search condition. Even when the display target of the neighborhood search condition is in an inclusive relationship, data management and display control for performing display suitable for visual recognition can be easily performed.

  In this embodiment, as shown in FIG. 1, one document search apparatus 101 and one client terminal 106 are provided. However, a plurality of at least one of the document search apparatus 101 and client terminal 106 are provided. Thus, a document search system may be configured.

In the present embodiment, the document database 103 and the full-text index 104 are stored in the document search apparatus 101, but other apparatuses that can communicate with the document search apparatus 101 can store the document database 103 and the full-text index 104. You may make it store in (server).
Furthermore, in this embodiment, the number of characters between keywords is used as the neighborhood search condition. However, the number of characters between keywords may not necessarily be used as the neighborhood search condition as long as the information indicates a range where the start keyword and the end keyword exist.
In the present embodiment, the case where a document is searched for in the vicinity in the forward direction has been described as an example. However, the document may be searched in the vicinity in the direction opposite to the forward direction.

(Other embodiments of the present invention)
In order to operate various devices to realize the functions of the above-described embodiments, program codes of software for realizing the functions of the above-described embodiments are provided to an apparatus or a computer in the system connected to the various devices. What is implemented by operating the various devices according to a program supplied and stored in a computer (CPU or MPU) of the system or apparatus is also included in the scope of the present invention.

  In this case, the program code of the software itself realizes the functions of the above-described embodiments, and the program code itself and means for supplying the program code to the computer, for example, the program code are stored. The recorded medium constitutes the present invention. As a recording medium for storing the program code, for example, a flexible disk, a hard disk, an optical disk, a magneto-optical disk, a CD-ROM, a magnetic tape, a nonvolatile memory card, a ROM, or the like can be used.

  Further, by executing the program code supplied by the computer, not only the functions of the above-described embodiments are realized, but also the OS (operating system) or other application software in which the program code is running on the computer, etc. It goes without saying that the program code is also included in the embodiment of the present invention even when the functions of the above-described embodiment are realized in cooperation with the embodiment.

  Further, after the supplied program code is stored in the memory provided in the function expansion board of the computer or the function expansion unit connected to the computer, the CPU provided in the function expansion board or function expansion unit based on the instruction of the program code Needless to say, the present invention also includes a case where the functions of the above-described embodiment are realized by performing part or all of the actual processing.

It is a figure which shows embodiment of this invention and shows an example of schematic structure of a document search system. It is a figure which shows embodiment of this invention and shows an example of the hardware constitutions of a document search device. 6 is a flowchart illustrating an example of a schematic operation when performing a neighborhood search in the document search system according to the embodiment of this invention. It is a flowchart which shows embodiment of this invention and continues from FIG. It is a figure which shows embodiment of this invention and shows an example of a neighborhood search condition designation | designated screen and a neighborhood search condition table. FIG. 5 is a flowchart illustrating an embodiment of the present invention and explaining details of an adjacent position acquisition process in merge information in the flowchart of FIG. 4. It is a figure which shows embodiment of this invention and shows an example of the concept at the time of creating merge position information. It is a figure which shows embodiment of this invention and shows the 1st example of the concept at the time of extracting the optimal combination of a start keyword and an end keyword using merge position information. It is a figure which shows embodiment of this invention and shows the 2nd example of the concept at the time of extracting the optimal combination of a start keyword and an end keyword using merge position information. It is a figure which shows embodiment of this invention and shows an example of the display screen of the document in which the result of the neighborhood search was reflected.

Explanation of symbols

101 server (document search device)
102 Neighborhood Search / Display Program 103 Document Database (Document DB)
104 Full-text index 105 Network 106 Client terminal 201 CPU
202 ROM
203 RAM
204 System Bus 205 Input Controller 206 Video Controller 207 Memory Controller 208 Communication I / F Controller 209 Keyboard 210 CRT Display 211 External Memory 503, 504 Merge Position Information 901 Search Condition Specification Screen 902 Neighborhood Search Condition Table

Claims (9)

A document search method for searching a document file according to a neighborhood search condition including a plurality of character strings and the existence ranges of the plurality of character strings,
A search step of searching the document file for a character string that matches a plurality of character strings included in the neighborhood search condition;
A character that determines whether or not a plurality of character strings searched in the search step are within the existence range included in the neighborhood search condition, and extracts a character string to be highlighted based on the determined result A column extraction step;
A display step of highlighting the character string extracted by the character string extraction step,
In the character string extraction step, when there are a plurality of character string sets composed of a plurality of character strings searched in the search step within the existence range included in the neighborhood search condition, the character string extraction steps A document search method, wherein a set of character strings that are closest to each other is extracted as the highlighted character string. In the search step, when a character string that matches a plurality of character strings included in the neighborhood search condition is searched, character string position information indicating a position where the plurality of character strings exist is acquired. And
The character string extraction step creates a merge position information by integrating a plurality of character string position information acquired by the search step in an order based on the character string position information, and using the created merge position information, The document search method according to claim 1, wherein the character string to be highlighted is extracted.   3. The document search method according to claim 2, wherein the character string extraction step creates only one merge position information regardless of the number of the neighborhood search conditions. A document search device that searches a document file according to a proximity search condition including a plurality of character strings and the existence ranges of the plurality of character strings,
Search means for searching the document file for character strings that match a plurality of character strings included in the neighborhood search condition;
A character that determines whether or not a plurality of character strings searched by the search means are within an existing range included in the neighborhood search condition, and extracts a character string to be highlighted based on the determined result Column extraction means;
Display information creating means for creating highlighted display information for highlighting the character string extracted by the character string extracting means;
When there are a plurality of sets of character strings made up of a plurality of character strings searched by the search means within the existence range included in the neighborhood search condition, the character string extraction means A document search apparatus, wherein a set of character strings that are closest to each other is extracted as the highlighted character string. The search means, when a character string that matches a plurality of character strings included in the neighborhood search condition is searched, obtains character string position information indicating a position where the plurality of character strings exist. And
The character string extraction means integrates a plurality of character string position information acquired by the search means in order based on the character string position information, creates merge position information, and uses the created merge position information, The document search apparatus according to claim 4, wherein the character string to be highlighted is extracted.   The document search apparatus according to claim 5, wherein the character string extraction unit creates only one merge position information regardless of the number of the neighborhood search conditions.   The document search apparatus according to claim 4, further comprising a transmission unit configured to transmit the highlighted display information created by the display information creation unit to a client terminal. A computer program for causing a computer to perform a search for a document file according to a proximity search condition including a plurality of character strings and the existence ranges of the plurality of character strings,
A search step of searching the document file for a character string that matches a plurality of character strings included in the neighborhood search condition;
A character that determines whether or not a plurality of character strings searched in the search step are within the existence range included in the neighborhood search condition, and extracts a character string to be highlighted based on the determined result A column extraction step;
Causing the computer to execute a display step of highlighting the character string extracted by the character string extraction step;
In the character string extraction step, when there are a plurality of character string sets composed of a plurality of character strings searched in the search step within the existence range included in the neighborhood search condition, the character string extraction steps A computer program, wherein a set of character strings closest to each other is extracted as the highlighted character string.   A computer-readable recording medium on which the computer program according to claim 8 is recorded.

Images