JPH0528199A - Document retrieval device - Google Patents

Abstract

PURPOSE:To prevent a retrieval omission and to enable an accurate document retrieval by providing a means selecting a required word from a word related to a retrieval word. CONSTITUTION:On a retrieval word inputted by a retrieval word input part 1, the related word is developed in a related word development part 3. With a related word dictionary part 4, the synonyms, high and low-order words of the retrieval word held in retrieval word holding part 2 are obtained. Then, the list of the related words is displayed on a related word display part 6, and a user selects the related word. Thus selected related word and the original retrieval word are held in the retrieval word holding part 2. A retrieval processing part 9 retrieves the retrieval word and synonyms. The retrieval result is held in a retrieval result holding part 10. Thus, the retrieval omission at the time of retrieval can be prevented, further, excess retrieval result generation can be prevented, resulting in accurate document retrieval.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a document retrieving device for retrieving by a computer.

[0002]

2. Description of the Related Art Due to the spread of document databases and the improvement of computer processing capability, a document retrieval apparatus for retrieving a document containing a designated retrieval word from a large number of document databases has come to be used. In such a search device,
Conventionally, it has been a general practice to give a keyword to each document and use the keyword when searching.

In recent years, in order to further improve the flexibility in searching, a full text search method has been used in which a keyword is not given in advance and a free keyword can be designated.

According to this method, since the keyword specified by the user is searched from the document, there is a problem in that the search may be omitted depending on how the keyword is specified.
In order to prevent this, the related words (synonyms, high-rank words, low-rank words, etc.) of the keyword given by the user are obtained and all of them are searched.

[0005]

However, if the related words of the word designated as the search word are searched in this way, the related words do not always match the user's search intention at that time. There was a problem that the results could be obtained.

Therefore, an object of the present invention is to eliminate the above-mentioned drawbacks and provide a document retrieval device which is easy to use.

[0007]

A document search device according to the present invention includes a search word input means for inputting a word to be searched, a related word dictionary describing a related word for each word, and the search word. A related word expanding means for searching a related word of a power word from the related word dictionary, and a means for selecting a related word when obtaining a related word to be used for a search from the inputted words to be searched. .

A document search device according to the present invention includes a search word input means for inputting a word to be searched, and a related word dictionary in which synonyms, high-rank words, and low-rank words are described for each word in a hierarchical structure. , Related word expanding means for searching a related word of the word to be searched from the related word dictionary, word number setting means for setting the number of related words to be expanded, and condition setting for setting a condition for expanding the related word Means for expanding related words according to a preset value when obtaining related words to be used for the search from the inputted words to be searched.

A document search device according to the present invention includes a search word input means for inputting a word to be searched, and a related word dictionary in which synonyms, high-rank words, and low-rank words are described for each word in a hierarchical structure. , Related word expanding means for searching a related word of the word to be searched from the related word dictionary, and when searching for a related word to be used from the input word to be searched, the meaning of the search word is In the case where the number is large, the evaluation means for instructing to specify the search word again is provided.

[0010]

In the present invention, a means for obtaining a related word of a search word and a means for selecting a necessary word from the obtained related words are provided, and as a result, an appropriate search word is used at the time of search. This is to prevent extra search results.

According to the present invention, a means for obtaining a related word of a search word, a means for setting the number of words of the related word, and a condition setting means for limiting the range of the related word to be developed are provided, and as a result, an extra amount is added in the search. This is to prevent the retrieval of various search results.

According to the present invention, a means for obtaining a related word of a search word and a means for evaluating the breadth of the meaning of the related word are provided, and as a result, an unnecessary search result is not produced during the search. is there.

[0013]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described in detail below with reference to the drawings.

Embodiment 1 FIG. 1 is a block diagram showing the configuration of an apparatus according to an embodiment of the present invention. In the figure, 1 is a search word input unit for inputting a search word, 2 is a search word holding unit for holding a search word input from the search word input unit 1, and 3 is a search word held in the search word holding unit 2. A related word expansion unit for obtaining related words such as synonyms, upper words, and lower words from 4; a related word dictionary used in the related word expansion unit 3 for expanding related words; 5 a related word selection unit for selecting related words; 6 is a related word display unit for displaying when selecting a related word, 7 is a selection input unit for inputting when selecting a related word, 8 is a document holding unit that stores documents, and 9 is A search processing unit 10 for searching the search word held in the search word holding unit 2 from the document in the document holding unit 8 is a search result holding unit for holding the result searched by the search processing unit 9.

FIG. 2 is a flow chart showing a processing procedure of the operation in the apparatus shown in FIG. The operation of the embodiment of the present invention will be described with reference to the figure.

First, in step S1, the search word input unit 1 inputs a search word. The input search word is held in the search word holding unit 2.

Next, in step S2, the related word expansion unit 3
Related words are expanded in. Here, using the related word dictionary 4, the synonyms of the search word held in the search word holding unit 2
Search for high-rank words and low-rank words. Then, the obtained result and the original search word are held in the search word holding unit 2.

Next, in step S3, step S2
The list of related words obtained in step 3 is displayed on the related word display unit 6. Then, in step S4, the user selects a related word. The selection input is performed by the selection input unit 7, but this may be performed, for example, by pointing the displayed related word with a pointing device or the like.

The end of the selection is indicated by an input instructing the end. In step S5, it is checked whether or not there is such an input from the user. Steps S4 and S5 are repeated until there is an input.

When there is an input, the related word selected and the original search word are held in the search word holding unit 2 in step S6.

Next, in step S7, the retrieval processing unit 9 performs retrieval processing for retrieving the retrieval word held in the retrieval word holding unit 2 from the document holding unit 8. The search processing here may be performed by a method such as an AC method for simultaneously searching for a plurality of character strings. The result of the search processing is held in the search result holding unit 10.

Next, the present embodiment will be further described by taking the case where the search word shown in FIG. 3 is input in the search word input section 1 as an example.

As a result of expanding the search word "murder" in the related word dictionary 4, as shown in FIG. 4, "striking" and "poisoning" as related words.
"Assassination,""suicide,""murder,""stabbed," etc. can be obtained. These are displayed on the related word display unit 6. Here, the user selects a related word from these. The selection method may be performed, for example, by indicating the word to be selected with a pointing device.

FIG. 5 shows the selected results. The selected words are shaded here. In the actual related word display portion, it is indicated by reverse display or the like. These words and the original search word “murder” are held in the search word holding unit 2 and used for the actual search. In the search, "assassination" and "suicide" that the user did not select are not searched, and unnecessary search results cannot be obtained.

Next, other embodiments will be listed.

1. In the above embodiment, the pointing device is used to select the related word, but this may be performed by specifying the number of the related word.

2. Although one word is used as the search word in the above-mentioned embodiment, a plurality of words can be dealt with in the same manner as in the embodiment of the present invention by individually processing the related word selection.

3. In the above-mentioned embodiment, the related word used for the search is instructed, but, on the contrary, the unnecessary related word may be instructed. In this case, the unselected word is used for the search.

4. In the above embodiment, the user always selects the related word. However, a threshold may be set and the user may select the related word only when the number of related words exceeds the threshold.

5. In the above embodiment, the AC method was used for the search, but other pattern matching algorithms may be used, or a method such as sentence analysis may be used. The present invention can be implemented by any search method.

Embodiment 2 FIG. 6 is a block diagram showing the arrangement of an apparatus according to another embodiment of the present invention. In the figure, 21 is a search word input unit for inputting a search word, 22 is a search word holding unit for holding a search word input from the search word input unit 21, and 23 is a search word held in the search word holding unit 22. To synonyms, broader terms,
A related word expansion unit for obtaining related words such as subordinate words, 24 is a related word dictionary used by the related word expansion unit 23 for expanding related words, and 25
Is a related word number setting unit for setting the number of related words to be expanded, 26
Is a development condition holding unit that holds a condition for expanding related words, 27 is a document holding unit that stores documents, 28 is a search word held in the search word holding unit 22 from a document in the document holding unit 27 Reference numeral 29 denotes a search result holding unit for holding the result searched by the search processing unit 28.

FIG. 7 is a flow chart showing the processing procedure of the operation in the apparatus shown in FIG. The operation of the embodiment of the present invention will be described with reference to the figure.

First, in step S21, the search word input unit 2
Enter the search word at 1. The input search word is held in the search word holding unit 22.

Next, in step S22, the related word expansion unit 23 expands the related words. Here, the related word dictionary 24 is used to find synonyms, high-rank words, and low-rank words of the search word held in the search word holding unit 22. The obtained result and the original search word are held in the search word holding unit 22.

Next, in step S23, step S
Check the number of related words found in 22. In step S24, it is determined whether this number is less than or equal to the value set in advance by the related word number setting unit 25.

If the number of related words is less than or equal to the set value, the search process of step S26 is performed.

On the other hand, if the number of related words is larger than the set value, the process proceeds to step S25.

In step S25, the range in which the related words are expanded is limited. For example, assuming that the conditions held as the initial conditions in the expansion condition holding unit 26 are “synonyms, high-rank words of distance 1, low-rank words of distance 1”, this is narrowed down to “synonyms only”.

Here, the distance is the distance in the related word dictionary described in a hierarchical structure.

The new condition is held in the expansion condition holding unit 26, and related words are expanded again in step S23.

Next, in step S26, the search processing unit 2
In 8, a search process for searching the search word held in the search word holding unit 22 from the document holding unit 27 is performed. The search processing here may be performed by a method such as an AC method for simultaneously searching for a plurality of character strings. The result of the search process is held in the search result holding unit 29.

Next, the present embodiment will be further described by showing an actual example.

FIG. 8 is a diagram showing a part of the related word dictionary 24. "Crime" is a broader term of "murder", "murder" and "murder" are synonymous terms, and "stabbing" is a subordinate term.
There are "poisoning" and "stroking". Further, the related word dictionary has a hierarchical structure such that “crime” has an upper word and “stabbed” and “poison” have a lower word. From the viewpoint of “crime”, “murder” is a subordinate word of distance 1 and “stabbed” is a subordinate word of distance 2.

The related word number setting unit 25 sets the number of related words to 5
Is set. Here, consider the case where "murder" is input as a search word in the search word input unit 21 as shown in FIG.

First, when the expansion condition holding unit 26 holds "synonyms, upper word of distance 1, lower word of distance 1" as initial conditions, the one shown in FIG. 10 is obtained as a related word of "murder". To be The number of related words obtained here is 5 which is the set value.
Since it is larger, the expansion condition is changed to "synonyms only".

As a result, as a related term of "murder", as shown in FIG.
As shown in 1, "kill" and "murder" remain. Since it is less than the set value, this is used for search.

Next, other examples will be listed.

1. In the above embodiment, one word is used as the search word, but a plurality of words can be dealt with in the same manner as the embodiment of the present invention by individually processing the related word expansion.

2. In the above embodiment, "synonyms, high-rank words of distance 1 and low-rank words of distance 1" are shown as initial conditions for related word expansion, and "synonyms only" are shown as restricted conditions, but these conditions are different. It doesn't matter. Further, the restriction of conditions may be divided into three or more stages.

3. In the above embodiment, after expanding the related words, the number of them is checked, and the expansion condition is changed when the number is larger than the set value. However, when the related word is expanded, the expansion condition is changed and the set value is changed. The following may be set.

4. In the above embodiment, the system sets the number of related words, but this may be set by the user.

5. If the number of related words is set to 0, the related words will not be expanded. Therefore, an embodiment of providing a mode in which the related words are not expanded is also included in the present invention.

6. In the above embodiment, the AC method was used for the search, but other pattern matching algorithms may be used, or a method such as sentence analysis may be used. The present invention can be implemented by any search method.

Embodiment 3 FIG. 12 is a block diagram showing the arrangement of an apparatus according to another embodiment of the present invention. In the figure, 31 is a search word input unit for inputting a search word, 32 is a search word holding unit for holding the search word input from the search word input unit 31, and 33 is a search word held in the search word holding unit 32. A related word expansion unit for finding related words such as synonyms, upper terms, and lower terms from 34, a related word dictionary used by the related word expansion unit 33 for expanding related words, 3
5 is a related word number setting unit for setting the number of related words to be expanded, 3
Reference numeral 6 is a related word evaluation unit that evaluates the number of related words and, if the number is large, requests to specify the search word again, 37 is a display unit used by the related word evaluation unit 36 for display, and 38 is a document storage A document holding unit 39, a search processing unit 39 for searching the search word held in the search word holding unit 32 from the document in the document holding unit 38,
Reference numeral 40 is a search result holding unit for holding the results searched by the search processing unit 39.

FIG. 13 is a flow chart showing the processing procedure of the operation in the apparatus shown in FIG. The operation of the embodiment of the present invention will be described with reference to the figure.

First, in step S31, the search word input unit 3
Enter the search word at 1. The input search word is held in the search word holding unit 32.

Next, in step S32, the related word expansion unit 33 expands the related words. Here, the related word dictionary 34 is used to obtain synonyms, high-rank words, and low-rank words of the search word held in the search word holding unit 32. The obtained result and the original search word are held in the search word holding unit 32.

Next, in step S33, the number of related words obtained in step S32 is checked. In step S34, it is determined whether or not this number is less than or equal to the value set in advance by the related word number setting unit 35.

If the number of related words is less than or equal to the setting, step S3
Move on to the search process of 6.

On the other hand, if the number of related words is larger than the set value, the process proceeds to step S35.

In step S35, the user is requested to input again a search word having a more limited meaning. This message is displayed on the display unit 37. Next in step S31
Return to and enter the search term again.

Next, in step S36, the search processing unit 3
In 9, a search process for searching the search word held in the search word holding unit 32 from the document holding unit 38 is performed. The search processing here may be performed by a method such as an AC method for simultaneously searching for a plurality of character strings. The search result holding unit 40 holds the result of the search processing.

Next, this example will be further described by showing an actual example.

FIG. 14 is a diagram showing a part of the related word dictionary 34. The subordinate terms of "vehicle" include "airplane", "automobile", "train", "ship", and the subordinate terms of "airplane" include "passenger aircraft", "fighter plane", and "bomber plane". Further, the related word dictionary has a hierarchical structure such that “vehicle” has a higher-ranking word and “passenger aircraft”, “fighter”, “bomber” etc. have a lower-ranking word.

The number of related words is set to 2 in the related word number setting unit 35.
It is assumed that 0 is set. Here, consider a case where "vehicle" is input as a search word in the search word input unit 31 as shown in FIG.

Here, in the hierarchical structure of the related word dictionary,
If all lower-order words are selected as related words, the number of obtained related words is larger than the set value of 20. Therefore, in FIG.
As shown in 6, an instruction is given to specify a word having a more limited meaning as a search word. On the other hand, the user designates “airplane” as a new search term, as shown in FIG. 17, for example.

Next, other examples will be listed.

1. In the above embodiment, one word is used as the search word, but a plurality of words can be dealt with in the same manner as the embodiment of the present invention by individually processing the related word expansion.

2. In the above example, the number of related words was used to evaluate the breadth of the meaning of the search word, but as an evaluation measure, the position of the search word in the hierarchical structure of the related word dictionary was used. Good. In this case, in the case of a word located in the upper part of the hierarchical structure, it is requested to specify the search word again.

3. In the above-mentioned embodiment, when the meaning of the search word is wide, the user is prompted to input a new search word. However, a list of words located one step below the search word in the hierarchical structure is shown and the search word is selected from the list. You may

4. In the above embodiment, the AC method was used for the search, but other pattern matching algorithms may be used, or a method such as sentence analysis may be used. The present invention can be implemented by any search method.

[0072]

As described above, according to the present invention,
Since it is possible to prevent the omission of the search at the time of the search and to prevent the generation of extra search results, it is possible to obtain the document search device with high accuracy.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of an apparatus according to a first exemplary embodiment of the present invention.

FIG. 2 is a flowchart showing a processing procedure according to the first embodiment of the present invention.

FIG. 3 is a diagram showing search terms for explaining the first embodiment of the present invention.

FIG. 4 is a diagram showing a display when processing is performed using the search word of FIG.

5 is a diagram showing a display when processing is performed using the search word of FIG.

FIG. 6 is a block diagram showing a configuration of an apparatus according to a second exemplary embodiment of the present invention.

FIG. 7 is a flowchart showing a processing procedure according to a second embodiment of the present invention.

FIG. 8 is a diagram showing a structure of a related word dictionary for explaining a second embodiment of the present invention.

9 is a diagram showing a structure of a related word dictionary shown in FIG. 8;

10 is a diagram showing a structure of a related word dictionary shown in FIG. 8;

11 is a diagram showing a structure of a related word dictionary in FIG.

FIG. 12 is a block diagram showing a configuration of an apparatus according to a third exemplary embodiment of the present invention.

FIG. 13 is a flowchart showing a processing procedure according to the third embodiment of the present invention.

FIG. 14 is a diagram showing a structure of a related word dictionary for explaining a third embodiment of the present invention.

FIG. 15 is a diagram for explaining processing using the related word dictionary of FIG. 14;

16 is a diagram for explaining a process using the related word dictionary of FIG. 14. FIG.

17 is a diagram for explaining a process using the related word dictionary of FIG. 14. FIG.

[Explanation of symbols]

1 Search word input section 2 Search word storage 3 Related term expansion section 4 Related term dictionary 5 Related word selection section 6 Related word display 7 Selection input section 8 Document storage 9 Search processing unit 10 Search result storage 21 Search word input section 22 Search word storage 23 Related term expansion section 24 Related term dictionary 25 Related Word Number Setting Section 26 Deployment condition holding unit 27 Document Holder 28 Search Processing Unit 29 Search result storage 31 Search term input section 32 Search term storage 33 Related term expansion section 34 Related term dictionary 35 Related Word Number Setting Section 36 Related Words Evaluation Department 37 Display 38 Document holding unit 39 Search Processing Unit 40 Search result storage

Claims (3)

[Claims] 1. A search word input means for inputting a word to be searched, a related word dictionary describing the related word for each word, and searching for a related word of the word to be searched from the related word dictionary. A document search device comprising: a related word expanding means; and a means for selecting a related word when a related word used for a search is obtained from the inputted words to be searched. 2. A search word input means for inputting a word to be searched, a related word dictionary in which synonyms / upper words / lower words are described in a hierarchical structure for each word, and a word to be searched. The related word expanding means for searching the related words from the related word dictionary, the word number setting means for setting the number of related words to be expanded, and the condition setting means for setting the condition for expanding the related words are inputted. A document search device, wherein when a related word used for a search is obtained from the words to be searched, the related word is expanded according to a preset value. 3. A search word input means for inputting a word to be searched, a related word dictionary in which synonyms, high-rank words, and low-rank words are described in a hierarchical structure for each word, and the word to be searched. Related word expanding means for searching a related word from the related word dictionary and, when a related word used for a search is obtained from the inputted words to be searched, when the meaning of the search word is wide and the number of the related words is large, A document search device comprising: an evaluation unit that gives an instruction to specify a search word.