JPH08314950A - Retrieval method and device for text - Google Patents

Abstract

PURPOSE: To systematically and comprehensively retrieve data through the use of a selected keyword and to easily and speedily extract a text fitted to a retrieval target. CONSTITUTION: A keyword memory area 11 storing a keyword group, a retrieval instruction part 15 taking out the keyword one by one and instructing to perform retrieval, aprimary retrieval part 33 retrieving the text, a retrieval set memory area 13 storing the text which hit, secondary retrieval part 34 executing secondary retrieval by AND obtained by adding another keyword when the number of hitting exceeds a set number of hitting, a hit number memory area 14 recording the number of hitting times of the text when the number of hitting times is less than the set number of hitting, an extraction part 20 extracting the text of the setting number of extraction in the order from one having the largest number of hitting times and outputting respective contents and an output part 21 are provided.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method and apparatus for retrieving a text containing a designated keyword from a plurality of texts such as Japanese and English.

[0002] In the field support work of a computer, in order to quickly resolve a new problem,
BACKGROUND ART It has been practiced to search a database for trouble cases that have occurred in the past and refer to their solutions.
In other words, when a problem occurs, after it has been resolved, describe the content of the problem that has occurred, the action taken in response to it, the cause of the problem, and other reference matters in a sentence, and register it. By doing so, a database of trouble cases is constructed. When a trouble newly occurs, the database is searched using a keyword according to the content of the trouble, and similar trouble cases are extracted and used for solving the trouble.

In order to effectively implement such a system, it is necessary to be able to efficiently extract a trouble case that actually serves as a reference.

[0004]

2. Description of the Related Art Conventionally, when extracting a target text from a database, a user selects a keyword, inputs the selected keyword from a keyboard or the like, and then searches by instructing execution. . The target text is extracted according to the keyword used for the search and the hit situation due to it. Searches are performed with various keywords and the content of the extracted text is examined to determine which text better matches the search objectives. When a plurality of keywords can be selected, the search is performed by a logical sum or a logical product of those keywords.

Generally, in a high-speed text search device (also referred to as a document search system or a text search system), a large amount of text stored in a database is searched at high speed for a text satisfying a specified search condition. An inverted file is provided for this purpose (for example, JP-A-3-294963). Due to the existence of the inverted file, most search conditions can be processed at high speed only by the operation on the inverted file.

[0006]

However, according to the conventional text search device, since the search is repeated by trial and error by repeating the search by the keyword selected by the user and the examination of the text extracted by the search, It was not possible to perform an exhaustive search, and there was a risk of omission of search or deviation of search content.

In particular, when it is found that the number of designated keywords is large and the number of hits is large in the middle of the search by those keywords, the initially planned search procedure is significantly changed. There is a problem that it is actually difficult to perform a systematic search.

The present invention has been made in view of the above-mentioned problems, and it is possible to perform a systematic and exhaustive search using selected keywords, and to easily and quickly extract a text suitable for the search purpose. It is an object of the present invention to provide a text search method and device capable of performing text.

[0009]

A method according to the invention of claim 1 is a method comprising a text database composed of a plurality of texts, wherein the text database is searched by a designated keyword group, which comprises: a) The specified keyword group is stored in the keyword memory area, b) one keyword is extracted from the keyword memory area, and c) the text including the extracted keyword is searched from the text group in the text database, and the hit text is searched. The search set, which is a set of, is stored in the search set memory area, and d) when the number of hits by the search exceeds a preset number of hits, the next one keyword is extracted from the keyword group, and the extracted keyword is extracted. The secondary search is performed according to Step d is repeated until the number of hits is less than or equal to the number of hits, and f) When the number of hits by the search is less than or equal to a preset number of hits, the number of hits for each hit text is counted, and g) Step b) is repeatedly executed and a search is performed using all the keywords of the keyword group at least once, and h) The preset number of set extractions from the text counted by the hits in descending order of the number of hits. The number of texts corresponding to is extracted, and the content of the extracted text is output.

An apparatus according to a second aspect of the present invention comprises a text database composed of a plurality of texts, searches the text database by a designated keyword group, and stores the designated keyword group. A keyword memory area, retrieval keywords one by one from the keyword memory area, retrieval instruction means for retrieving stored keywords at least once and instructing retrieval, and executing a text retrieval from the text database by the keywords. A primary search means, a search set memory area for storing a search set which is a set of hit texts, a determination means for determining whether or not the number of hits by the search exceeds a preset number of hits, and a hit by the search A number of preset hits When exceeding the number, and the secondary retrieval means for performing a search of the text by the logical product plus other one keyword extracted from the keyword group,
When the number of hits by the search is less than or equal to the preset number of hits, a recording means for recording the number of hits for each hit text, and the text counted by the hits are arranged in descending order of the number of hits in advance. And an output means for extracting the number of texts corresponding to the set number of extracted settings and outputting the respective contents.

According to a third aspect of the present invention, an apparatus according to a third aspect of the present invention retrieves a keyword memory area for storing a designated keyword group and one keyword from the keyword memory area, and retrieves the stored keyword at least once. A search instruction means for instructing a search, a primary search means for executing a text search from the text database by a keyword, a search set memory area for storing a search set which is a set of hit texts, and the number of hits by the search are previously A determination means for determining whether or not the set number of hits exceeds a set number, and a logic including one keyword extracted from the keyword group when the number of hits by the search exceeds a preset number of set hits. A secondary search means for performing a text search by multiplication, Instructs to add 1 to the hit count in the hit count memory area for the hit text when the hit count memory area that records the hit count of the kist and the hit count by the search is less than or equal to the preset hit count. Addition instruction means for extracting, from the text recorded in the hit count memory area, extraction means for extracting a number of texts corresponding to a preset number of set extractions in descending order of the number of hits; Content output means for outputting the content.

In the apparatus according to the fourth aspect of the present invention, when the retrieved keyword is used for the secondary search and the number of hits in the secondary search is zero, the search instructing means performs the primary search for the keyword. To be used again as a keyword for.

[0013]

An example of the search procedure of the present invention will be described with reference to the flowchart shown in FIG. The specified keyword is
It is stored in the keyword memory area 11 (# 12). One keyword is retrieved from the keyword memory area 11 (# 13). A command for performing a search is generated using the retrieved keyword (# 14), and the text database 12 is searched by executing the command (# 15).

The hit text TX is stored in the search aggregate memory area 13 (# 16), and when the number of hits is less than or equal to the set number of hits (Yes in # 17), the text TX is registered in the hit count memory area 14. Then, "1" is written in the number of hits (# 18).

If the number of hits exceeds the set number of hits (No in # 17), the procedure returns to step # 13, and the next keyword is fetched from the keyword memory area 11,
A secondary search is performed.

If there is a keyword stored in the keyword memory area 11 that has never been used (NO in # 19), the procedure returns to step # 13 and the next keyword is retrieved from the keyword memory area 11. , A primary search is performed. When the number of hits in the secondary search is “0”, the primary search is newly performed with the same keyword.

If all the keywords stored in the keyword memory area 11 are used, (#
(Yes in 19), the text TX recorded in the hit count memory area 14 is sorted (# 20), and the identification name IDs whose number is equal to the set number of extracted items are extracted (# 2).
1), the contents are output (# 22).

[0018]

1 is a block diagram functionally showing the configuration of a text search device 1 according to the present invention, FIG. 2 is a block diagram showing an example of the hardware configuration of the text search device 1, and FIG. 3 is an example of a text database 12. FIG.

In FIG. 1, the text search device 1 includes a keyword memory area 11, a text database 12,
Search set memory area 13, hit count memory area 14,
The search instruction section 15, the search section 16, as the search instruction means of the present invention, and the hit number determination section 1 as the hit number determination means
7. Addition instruction unit 1 as recording means and addition instruction means
8, a sorting unit 19, an extraction unit 20 as an extraction unit, an output unit and an output unit 21 as a content output unit, and the like. The hit count memory area 14 also has a function as a recording unit of the present invention.

The hit number determination unit 17 is provided with a set hit number memory area 35 for storing the set hit number which is a control constant. The keyword memory area 11 is
This is a memory area for storing a specified keyword group.
Eight keywords KW are stored in the keyword memory area 11 shown in FIG. A keyword group, which is a set of keywords KW, is input from a keyboard or other input device (not shown). Alternatively, as in the embodiment described later, the keyword KW is automatically extracted from the input natural sentence CM, and the extracted keyword KW is automatically input to the keyword memory area 11.

The text database 12 is composed of the inverted file FI and a large number of text TX data. As shown in FIG. 3, the data length of each text TX is indefinite. An identification name ID such as text1, text2 ... Is attached to the text TX. Text database 1
2, the text TX can be updated by adding, changing, deleting, or the like. The inverted file FI is a file in which a large number of index words and the identification name ID of the text TX including each index word are stored in association with each other.

The search set memory area 13 is a memory area for storing a search set which is a set of hit text TX. The hit count memory area 14 registers the identification name ID of the text TX whose hit count is equal to or less than the set hit count by the primary search, the secondary search, or the repetition of the secondary search, and the hit count A memory area for recording.

The search instruction section 15 is provided with a keyword extracting section 31 and a search command generating section 32.
The keyword extracting unit 31 sequentially extracts the keywords KW from the keyword memory area 11 one by one. The search command generation unit 32 generates a command for performing a search using the retrieved keyword. The command is generated in the format of (command word + keyword). The command words include "SEARCH" for instructing the primary search and "AND" for instructing the secondary search.

A primary search is instructed for the keyword retrieved first from the keyword memory area 11. When the number of hits in the primary search exceeds the preset number of hits, the secondary search is instructed for the keyword extracted next. When the secondary search is instructed for the keyword and the number of hits in the secondary search is zero, the primary search is newly instructed by the keyword.

The search unit 16 executes the search according to the command received from the search instruction unit 15. The search unit 16 is provided with a primary search unit 33 as a primary search unit and a secondary search unit 34 as a secondary search unit. When the command word is "SEARCH", the primary search unit 33 executes the primary search of the text TX from the text database 12 by the keyword. The identification name ID of the text TX that is the search set is stored in the search set memory area 13. When the command word is “AND”, the secondary search unit 34 executes the secondary search of the text TX from the search set memory area 13 using the keyword. That is, in the secondary search, the text TX is searched by the logical product of the primary search keyword and the secondary search keyword. The secondary search is repeated until the number of hits falls below the set number of hits. The search set by the secondary search is also stored in the search set memory area 13.

The hit number determination unit 17 determines whether the number of hits by the search exceeds a preset number of hits. That is, the hit count determination unit 17 checks the hit count every time the search by the search unit 16 is performed.

The addition instruction unit 18 instructs the text TX registered in the hit count memory area 14 to add 1 to the hit count each time the number of hits falls below the set number of hits. However, "1" is set when the text TX is first registered.

The sorting unit 19 uses the hit count memory area 1
The text TX registered in No. 4 is sorted in descending order of the number of hits. The extraction unit 20 extracts the text TX registered in the hit count memory area 14 after sorting,
Text T with a number equal to the number of preset extractions
Extract X. As a result, the hit count memory area 1
From the text TX recorded in No. 4, the set number of extracted text TX are extracted in descending order of the number of hits.

The output unit 21 extracts the content of the text TX extracted by the extraction unit 20 from the text database 12 and outputs it to an external device. In FIG. 2, the hardware of the text search device 1 includes a processing device 41, a storage device 42, a mass storage device 43, a keyboard 44, a display device 45, a printer device 46, and the like.

The storage device 42 stores the text database 12 described above. The large capacity storage device 43 stores a large capacity text database 12. The processing device 41 includes a keyword memory area 11, a search set memory area 13, a hit count memory area 14, and a memory area serving as a set hit count memory area 35, a memory area to which all or part of the text database 12 is transferred, and a work area. It has a memory area used as an area. The keyboard 44 is used by the user to input the keyword KW when performing a search, and when the search is performed or when the text TX is registered or updated in the text database 12, the text T is displayed.
It is also used to input X, commands, data, or the like. The search results and other information are displayed on the display device 4.
5 or printed on paper by the printer device 46. Also, it is transferred to another computer or a file device by a communication line.

Next, the processing content or operation of the text search device 1 will be described. Figure 4 is a natural sentence CM for searching
FIG. 5 is a diagram showing an example of the contents of the keyword memory area 11, FIG. 6 is a diagram showing the execution history of the search, FIG. 7 is a diagram showing an example of the contents of the hit count memory area 14, and FIG. 3 is a flowchart showing the procedure of a search process. In the example shown in the figure, the number of set hits is “12” and the number of set extracted cases is “9”.

First, when a natural sentence CM as shown in FIG. 4 is input, the keyword KW is automatically extracted from the natural sentence CM (# 11) and stored in the keyword memory area 11 as shown in FIG. 5 ( # 12). In the example shown in FIG. 5, the natural sentence CM is divided for each character type, and consecutive words of the same character type become the keyword KW. However, the word Kana is not used as the keyword KW.

One keyword KW is fetched from the keyword memory area 11 (# 13). A command for performing a search is generated using the retrieved keyword KW (# 14), and the text database 12 is searched by executing the command (# 15).

In the example shown in the figure, first the keyword "JE
"F" is taken out and the command "SEARCH JEF"
Is generated and a search is performed by the command "SEARCH JEF".

If the identification name ID of the hit text TX is stored in the search aggregate memory area 13 (# 16), and the number of hits is less than or equal to the set number of hits (YES in # 17), the identification name ID of the text TX. Is registered in the hit count memory area 14 and "1" is written in the hit count (# 18).

If the number of hits exceeds the set number of hits (NO in # 17), the process returns to step # 13 to retrieve the next keyword KW from the keyword memory area 11 and perform the secondary search.

The keyword KW stored in the keyword memory area 11 has not been used even once.
If there is (No in # 19), the process returns to step # 13 and the next keyword KW is read from the keyword memory area 11.
Is retrieved and a primary search is performed. However, when the number of hits in the secondary search is “0”, the primary search is newly performed with the same keyword KW.

In the example shown in the figure, the command "SEARCH
Although 1927 hits were found by the primary search by "JEF" (see Fig. 6), this number is "1" of the set hit count.
2 ”is exceeded, a secondary search is performed with the next keyword KW,“ code ”(see FIGS. 5 and 6). However, since the number of hits by the secondary search is 0, it is not recorded in the hit count memory area 14. Since the number of hits is 0, the primary search is performed with the same keyword "code", and the number of hits "1" is obtained (see FIG. 6). This one text T
X is recorded in the hit count memory area 14. The next keyword KW, “6EA4”, is retrieved from the keyword memory area 11 and the primary search is performed. Since the number of hits is “14”, the secondary search is performed with the next keyword “expansion”, and the number of hits “9” is obtained. 9
The text TX in question is recorded in the hit count memory area 14.

When all the keywords KW stored in the keyword memory area 11 are used (Yes in # 19), the identification name IDs recorded in the hit count memory area 14 are sorted (# 20), Distinguished name IDs whose number is equal to the set number of extracted items are extracted (# 2
1), the contents are output (# 22).

In the example shown in FIG. 7, in the hit count memory area 14, text4 is 3 times, text1 is 3 times, and text2 is 3 times.
Times, text5 is recorded twice, and so on. The top 9 items are extracted from the text TX recorded in the hit count memory area 14, and the contents are displayed on the display device 45.
And is printed by the printer device 46.

According to the above-described embodiment, when a plurality of keywords KW are designated, the text database 12 is searched using all of the keywords KW, and the required number of texts TX are extracted and their contents are retrieved. Is output. Moreover, the number of hits for the keyword KW is statistically processed and automatically extracted in descending order of the number of hits. Therefore, the target text TX can be searched easily and quickly by performing an exhaustive search without depending on the user's ability to combine keywords as in the conventional case.

Since the keyword KW is automatically generated by inputting the natural sentence CM, it is easy to specify the keyword KW. Generation of the keyword KW for designating the keyword KW, and the text database 12
If the same keyword KW is obtained by interlocking with the extraction of the keyword KW for creating the inverted file FI in the above, the text TX having a high degree of similarity to the natural sentence CM that is the source of the search is extracted. can do.

Further, when the number of hits in the secondary search is 0, the primary search is newly performed with the same keyword KW, so that the keyword KW is effectively used, and the omission of the search can be reduced. it can.

In the above-described embodiment, the word itself used in the text TX or the natural sentence CM is used as the keyword KW. However, all katakana and the like are converted into English characters, and the keyword KW is an English character. You may make it use many. Secondary search Search set memory area 1
However, the secondary search may be performed by searching the text database 12 also using the keyword KW of the primary search. The command format and the like can be changed as appropriate. Although the text TX recorded in the hit count memory area 14 is sorted, the text TX may be extracted in order of the hit count without sorting. Method, keyword memory area 11 for selecting type, content, character type, quantity, and keyword KW of text TX
The extraction order of the keywords KW from the, the configuration of the text search device 1, the processing content, the processing order, and the like can be appropriately changed in accordance with the gist of the present invention.

[0045]

According to the inventions of claims 1 to 4,
It is possible to perform a systematic and exhaustive search using the selected keywords, and easily and quickly extract a text that matches the search purpose.

According to the invention of claim 4, the keyword is effectively used, and the omission of retrieval can be reduced.

[Brief description of drawings]

FIG. 1 is a block diagram functionally showing the configuration of a text search device according to the present invention.

FIG. 2 is a block diagram showing an example of a hardware configuration of a text search device.

FIG. 3 is a diagram showing an example of a text database.

FIG. 4 is a diagram showing an example of a natural sentence for a search.

FIG. 5 is a diagram showing an example of contents of a keyword memory area.

FIG. 6 is a diagram showing a process of executing a search.

FIG. 7 is a diagram showing an example of the contents of a hit count memory area.

FIG. 8 is a flowchart showing a procedure of search processing.

[Explanation of symbols]

 1 Text Retrieval Device 11 Keyword Memory Area 12 Text Database 13 Search Set Memory Area 14 Hit Count Memory Area (Recording Means) 15 Search Instructing Unit (Search Instructing Means) 17 Hit Number Judgment Unit (Hit Number Judging Means) 18 Addition Indication Unit ( Recording unit, addition instruction unit) 20 extraction unit (output unit, extraction unit) 21 output unit (output unit, content output unit) 33 primary search unit (primary search unit) 34 secondary search unit (secondary search unit)

Claims (4)

[Claims] 1. A method for searching a text database by a designated keyword group, comprising a text database composed of a plurality of texts, comprising: a) storing the designated keyword group in a keyword memory area; b) retrieve one keyword from the keyword memory area, c) retrieve a text containing the retrieved keyword from the text group in the text database, and store a retrieval set, which is a set of hit text, in the retrieval aggregate memory area D) When the number of hits by the search exceeds the preset number of hits, the next one keyword is extracted from the keyword group and a secondary search is performed by the extracted keyword, and e) The number of hits is set by the above. Repeat step d until the number of hits is less than or equal to F) When the number of hits by the search is less than or equal to the preset number of hits, the number of hits for each hit text is counted, and g) Step b and subsequent steps are repeated, and the keyword group is executed. Search using all keywords at least once, and h) From the texts counted by hits, extract the number of texts corresponding to the preset number of extractions in order of the number of hits, and extract A text search method characterized by outputting the contents of the specified text. 2. An apparatus comprising a text database composed of a plurality of texts, for searching the text database by a designated keyword group, the keyword memory area storing the designated keyword group, and the keyword memory. A retrieval instruction means for retrieving keywords one by one from the area and for instructing retrieval by retrieving stored keywords at least once, a primary retrieval means for executing a text retrieval from the text database by the keyword, and a hit text A search set memory area that stores a search set that is a set of, a determination means that determines whether the number of hits by the search exceeds a preset number of hits, and a set hit that the number of hits by the search is preset. If you exceed the number of cases, A secondary search means for performing a text search by a logical product that adds another one keyword extracted from the keyword group, and each hit text when the number of hits by the search is equal to or less than a preset number of hits. From the texts counted by the hits and the recording means that records the number of hits, the number of texts corresponding to the preset number of extractions in the order of the number of hits is extracted and the content of each is output. A text search device comprising: an output unit. 3. A device for searching the text database by a designated keyword group, comprising a text database composed of a plurality of texts, the keyword memory area storing the designated keyword group, and the keyword memory. A retrieval instruction means for retrieving keywords one by one from the area and for instructing retrieval by retrieving stored keywords at least once, a primary retrieval means for executing a text retrieval from the text database by the keyword, and a hit text A search set memory area that stores a search set that is a set of, a determination means that determines whether the number of hits by the search exceeds a preset number of hits, and a set hit that the number of hits by the search is preset. If you exceed the number of cases, A secondary search means for performing a text search by a logical product of another one keyword extracted from the keyword group, a hit count memory area for recording the number of hits of the text, and a hit count by the search are preset. When the number of hits is less than or equal to the set number of hits, an addition instruction means for instructing to add 1 to the number of hits in the hit count memory area for the hit text, and the number of hits from the text recorded in the hit count memory area A text search characterized by comprising: an extraction means for extracting a number of texts corresponding to a preset number of extractions in order of descending number of content; and a content output means for outputting the content of the extracted texts. apparatus. 4. The search instructing means re-uses the keyword as a keyword for the primary search when the retrieved keyword is used for the secondary search and the number of hits in the secondary search is zero. The text search device according to claim 2, wherein the text search device according to claim 2 or 3.