JPH05324728A - Information retrieving device - Google Patents

Abstract

PURPOSE:To improve the operability even when information having many relative words of a key word is retrieved by taking a relative word relating to the key word out of a relative word dictionary and then retrieving object information in an information storage device by using a new key word generated by combining the relative word and key word. CONSTITUTION:When the key word is inputted from an input device 1, a relative word processor 5 takes the relative word relating to the key word out of the relative word dictionary 6 and passes the relative word to a retrieval processor 3, which performs a retrieval process based upon the new key word obtained by combining the relative word and key word together under OR conditions. The relative word is stored while paired with the relativity to the key word and the relative word processor 5 takes out relative words having relativity larger than a set value and uses them for the information retrieval. Consequently, even when many key word candidates are predicted, the need to investigate relative words is eliminated and relative words need not be inputted, one by one, so the operability is improved.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an information search apparatus, and more particularly to an information search apparatus suitable for searching information having a plurality of related words as keywords.

[0002]

2. Description of the Related Art Generally, when searching for desired information from an indexed information storage medium such as a database, a keyword representing the information to be searched is input, and information having the same index as this keyword is read from the information recording medium. Although the search is performed, if the selection of the keyword is not appropriate, the keyword must be input many times before the target information is obtained, which causes a problem that the search work is extremely troublesome.

Therefore, as disclosed in Japanese Patent Application Laid-Open No. 1-72231, when a plurality of candidate information are searched as a result of inputting a keyword, the index of the candidate information is used. An information retrieval apparatus has been proposed in which the selection of a keyword required for the next retrieval is not required by performing the next retrieval by the next retrieval and the efficiency of the retrieval work is improved.

[0004]

By the way, when performing information retrieval, the index attached to the target information is predicted and the keyword is selected. For example, for information such as "floppy disk" Often wonders which of the predicted keywords such as "floppy", "FD", "disk" may be selected.

However, in the above-mentioned conventional information retrieval apparatus, for example, the keyword "floppy" is not considered for the related words such as "FD" and "disk". , "Floppy"
If the target information cannot be retrieved by the first keyword, the keywords such as "FD" and "disk", which are related words, must be input word by word to perform the retrieval. For this reason, there is a problem that a search work for information having many related words of a keyword is troublesome and takes time.

An object of the present invention is to eliminate the need to input a related word of a keyword when the related word of the keyword exists, and particularly to improve the operability in searching information having many related words of the keyword. To provide.

[0007]

In order to achieve the above object, the present invention has a related word dictionary in which related words related to a keyword used for information retrieval are stored, and a keyword to be searched for is attached to a search target. An information storage device that stores information and a related word related to a keyword are extracted from the related word dictionary, a new keyword is created by combining the related word and the keyword, and the new keyword is used to create a new keyword from the information storage device. A search processing device that searches for information to be searched is provided.

[0008]

According to the above means, the related word related to the keyword is taken out from the related word dictionary, the related word and the keyword are combined to create a new keyword, and the new keyword is used to search the information storage device for retrieval. Even if many keyword candidates are predicted, it is not necessary to search for related words for that keyword, and when searching information that has many related words as keywords, related words are searched. There is no need to input each word. As a result, the operability is improved and the search can be completed in a short time.

[0009]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described in detail below with reference to illustrated embodiments.

FIG. 1 is a block diagram of an overall configuration showing an embodiment of an information retrieval apparatus of the present invention. A keyword relating to information to be retrieved is input device 1, a display device 2 for displaying retrieved information, etc. A search processing device 3 for searching information having a keyword input from the device 1, an information storage device 4 storing information to be searched with a keyword for search, and a related word related to the keyword input from the input device 1. When the user has a dictionary 6 and inputs a keyword, the related word related to the keyword is taken out from the related word dictionary 6, the related word is passed to the search processing device 3, and the search processing device 3 receives the related word and the keyword. The related word processing device 5 is configured to execute a search process based on a new keyword combined under a logical sum condition. In this case, the search processing device 3
The related word processing device 5 may be configured as an integrated processing device.

FIG. 2 is an explanatory diagram showing an example of the contents of the related word dictionary 6 included in the related word processing device 5. The related word dictionary 6 is composed of a keyword 7, a related word 8 and a degree of association 9. Of these, the keyword 7 stores a keyword input from the input device 1, and the related word 8 stores a word related to the keyword. In this case, as shown in the figure, for one keyword such as "floppy",
When there are a plurality of related words such as “FD” and “disk”, the keyword 7, the related word 8 and the degree of association 9 are stored as a set and the number of related words is stored.

The degree of association 9 expresses the degree of association between the keyword 7 and the associated word 8 in the range of 0 to 100, and is a value subjectively determined by the creator of the associated word dictionary 6, and the larger the value, the more related It shows that the degree is high.

Next, the operation of the above configuration will be described with reference to the flowchart of FIG.

First, in step 31, the information storage device 4
From the input device 1, the keyword 36 of the information to be retrieved from is input. For example, the keyword 36 "floppy"
Enter. The input keyword 36 is passed to the search processing device 3.

Therefore, in step 32, the search processing device 3 determines whether or not the processing of the related word for the input keyword 36 has been completed. In this determination, if the processing of the related word for the keyword 36 is not completed, the processing of extracting the related word for the keyword 36 input in step 33 from the related word dictionary 6 is performed. That is, the input keyword 36 is transferred to the related word processing device 5. Then, the related word 38 for the keyword 36 input by the related word processing device 5 is extracted from the related word dictionary 6 and returned to the search processing device 3.

In this case, the search processing device 3 uses the input device 1 as a condition of the degree of association of the related word, "relevance degree = 75 or more"
When the condition 37 such as is set, the condition 37 of the degree of association is also transferred to the related word processing device 5. Then, when a plurality of related words exist as related words related to the keyword 36 in the related word dictionary 6, the related word processing device 5 satisfies the condition 37 such as “relevance = 75 or more”.
Is retrieved and returned to the search processing device 3.

According to the example of the related word dictionary 6 shown in FIG. 2, the related word "FD" for the keyword 36 "floppy" excluding "storage device" having a degree of association of 50 and "medium" having a degree of association of 70. , “Disk”, “diskette” are taken out and returned to the search processing device 3.

When there are a plurality of keywords input in step 31, extraction of related words satisfying the ending condition 37 for processing related words in step 32 is repeated for the number of keywords.

FIG. 4 is an explanatory diagram for explaining the details of the processing for extracting the related words in step 33, in which the keyword 3 is used.
When the related word processing device 5 retrieves the related word related to 6 and returns it to the search processing device 3, the search processing device 3 causes the display device 2 to display the keyword 36, the related word 42, and the degree of association 4
3. The information including the re-search instruction 44 is displayed on the display device 2 as the first stage related word list 45.

When the related word 38 displayed in the related word list 45 in the first stage is to be searched again, Yes is specified in the field of the re-search instruction 44 from the input device 1; otherwise, No is selected. specify.

With regard to the related word designated as Yes, the related word processing device 5 extracts the related word again by using the related word as a keyword.

Similarly, the second stage related word list 46,
The related word list 47 of the third stage is displayed.

In FIG. 4, the search conditions at the final stage are displayed. Then, the keyword 39 which is the current newly set search condition is displayed. At this time, re-search instruction item 4
Only those with Yes in 4 are new keywords 39
Is a component of.

Note that the keyword for which Cancel is specified is removed from the new search condition.

As described above, when the related word 38 for the input keyword 36 is extracted from the related word processing 6, the retrieval processing device 3 in step 34 retrieves the keyword 36 and the related word 38 extracted in step 33. A new keyword 39 is created by combining with the logical sum condition.

In this example, "floppy" or "F"
A new keyword 39 "D" or "disk" or "diskette" is created.

Next, in step 35, the information retrieval device 3 accesses the information storage device 5 by using this new keyword 39, retrieves information corresponding to the keyword 39, and displays the retrieval result on the display device 2. To do.

As described above, one keyword 36
If there is a related word 38 for, the related word 38
Is extracted from the related word dictionary 6, the related word 38 and the keyword 36 are combined to create a keyword 39 indicating a new search condition, and the information storage device 4 is searched by using this new keyword 39, whereby When searching information having a plurality of related words, it is not necessary to input the related words one by one, the operability is improved, and the search can be completed in a short time.

Therefore, it is extremely effective when searching for information having many related words as keywords.

Note that the keyword 36 without combining related words
When it is desired to perform a normal search only by itself, if the condition of step 32 is always satisfied, the same search as that of a conventional information search apparatus having no related word processing function becomes possible.

The relevance degree condition 37 may be changed every time a re-search is performed.

Although the re-search is instructed each time the related word is displayed, if the end condition 37 of the related word processing is determined to be the degree of relevance n or more and the re-search is up to the m-th stage, the process may be interrupted. It can be processed automatically without waiting for the instructions in.

Further, although the related words are taken out from the related word dictionary 6, the related words may be taken out from an apparatus for automatically generating the related words instead.

Alternatively, the new search condition 39 may be directly editable.

Next, a method of accumulating related words and a learning method will be described with reference to FIG.

First, the related word dictionary 6a in FIG. 5A is a related word dictionary in the initial state, and the related word dictionary 6a in the initial state.
The utilization degree 10a, which is one of the items, has a variable range set to an intermediate value "50" of "0 to 100".

If a keyword 36a as shown in the figure is input to the related word dictionary 6a in the initial state, the related words "FD" and "disk" related words 38a are input.
a is taken out.

Therefore, when the extracted related word 38a is actually used for creating a new search condition in step 34 of FIG. 3, the related word "FD" in the related word dictionary 6a corresponding to the related word 38a is used. "And" disk usage 10a
"1" is added to.

The dictionary in a new state in which the utilization value is updated is the related word dictionary 6b in FIG. 4 (b), and the updated utilization degree 10b.
The value of is "51".

Similarly, when the second search is performed, the keyword 36b is input, and the related word 38b to this is input.
Is retrieved from the updated related word dictionary 6b.

In the second search, only the related word "FD" corresponding to the keyword "floppy" 36b is used, and the related word "disk" is not used.

Therefore, in this case, "1" is subtracted from the usage rate of the related word "disk" which is not used. As a result, the related word dictionary 6c updated to the usage 10c as shown in FIG. 5C is created.

The values of the degrees of utilization 10a, 10b, 10c are set to the degrees of relevance in consideration of the degree of utilization, for example, by the following formula 1 when determining the condition for extracting the related word in step 32 of FIG. It is corrected, and the related words satisfying the corrected degree of relevance are extracted and used.

[0044]

[Equation 1] Relevance with utilization rate = relevance degree × utilization degree / 100 For learning related words, addition and subtraction of this utilization value are performed, and related words with high utilization values are preferentially displayed. With this, the related words having a high degree of use can be preferentially taken out.

When a new search condition is directly edited, the items connected by the logical sum condition (or condition) are regarded as related words, and the items connected by the or condition are regarded as related words. Are extracted and checked as a pair of a keyword and a related word to be registered in the related word dictionary 6. If not registered, the word is registered as a related word. At this time, the degree of association and the degree of use are assumed to be, for example, “50”. In this way, this newly registered related word can also be used in the next search.

Next, another learning method of the related word dictionary will be described with reference to FIG.

The related word dictionary 6d shown in FIG. 6 is an example in which the learning information 14 is independently held in the dictionary 6d. The number 11 is uniquely registered for each keyword and related word pair. The learning information 14 is composed of a number 11, a user name 12 and a degree of use 13. The degree of use 13 for a certain related word is retrieved by searching the learning information 14 using the number 11 and the user name 12 as keys.

The user name 12 of the learning information 14 is input from the input device 1 at the start of information retrieval.

The method of updating the usage 13 is the same as in the case of FIG. The usage updated at that time is only the usage for the user name input at the start of the information search. By doing this, you can manage the degree of association of multiple users separately,
The optimum related words can be extracted and used for each user.

Next, another method of constructing the related word dictionary will be described with reference to FIG.

The related word dictionary 6e of FIG. 7 has three values of the degree of relevance of the related word with respect to the keyword: the degree of relevance 9a, the degree of relevance 9b, and the degree of relevance 9c. These three types of relevance are, for example, a relevance of 9 in computer-related information retrieval.
a, the degree of relevance 9b is used for the advertisement-related information search, and 9c is used for the electronic-related information search.

However, the three types of relevance are weighted by a calculation formula such as Formula 2, and this value is used as the value of the relevance of the related word extraction condition in step 32 of FIG. By doing so, it is possible to simultaneously use information on the degree of association of different fields.

[0053]

## EQU00002 ## Association degree = association degree 9a.times.0.5 + association degree 9b.times.0.1 + association degree 9c.times.0.4 Next, another method of constructing the related word dictionary will be described with reference to FIG.

The related word dictionary 6f of FIG. 8 is an example in which the related degree information 9d and the learning information 14 are independently provided in the related degree word dictionary 6f.

Number 11 for each keyword-related word pair
Is uniquely attached and registered.

The related word information 9d includes the number 11 and the field 15.
And the degree of association 16 in the field 15.

The learning information 14 is the same as that shown in FIG.

In such a structure, the degree of relevance and the degree of use of a certain related word are retrieved by retrieving the degree of relevance information 9 using the number 11 for the related word and the field 15 of the information to be searched as a key. Which field the degree of association is to be used is input from the input device 1 when the keyword is input in step 31 of FIG.

Next, a method of using a plurality of related word dictionaries will be described with reference to FIG.

In FIG. 9, a plurality of related word dictionaries such as a related word dictionary 6g for the computer field, a related word dictionary 6h for the advertising field, and a related word dictionary 6i for the electronic field are provided in the related word processing device 5a. When a plurality of related word dictionaries are provided in this way, the related word dictionary used in step 31 of FIG. 3 is selected. At this time, if only one of the related word dictionary 6g is selected, the related word dictionary 6
Using the degree of relevance 9a of g, a process of extracting a related word for a keyword is performed.

However, the related word dictionary 6h and the related word dictionary 6i
If two are selected, the keyword "floppy"
When searching for a related word for the
"D, 100" and "diskette, 100", and from the related word dictionary 6i "media, 50" and "diskette, 9"
0 "is taken out. Of these, the degree of association is "75 or higher"
In the case of retrieving the word, the relevance is calculated by Expression 3 based on the relevance of each related word dictionary.

[0062]

## EQU00003 ## Relevance = (relevance 9b + relevance 9c) / 2 Then, "FD, 50", "diskette, 95",
It becomes "medium, 25", and only "diskette" is taken out as a related term. At this time, it is assumed that the related degree of the related word only in one dictionary (6h or 6i) is "0".

Although the related word dictionaries are provided for each field,
It is also possible to have a separate related word dictionary for each user of information retrieval.

Regarding the related word dictionary, Japanese Patent Laid-Open No. 59-59
As described in Japanese Patent Laid-Open No. 197923 “Document Creation Device”, a method of using a related word dictionary in which words having different meanings related to a selected word are sequentially displayed, and a desired word can be easily input or changed. There is. Also, JP-A-62-19786
As described in Japanese Patent Laid-Open No. 3 “Language Information Providing Device”, there is a method of arranging data having similar meanings close to each other and realizing a related word dictionary with a structure in which they are arranged sequentially.

However, here, no consideration is given to the ease of dynamically changing the degree of relevance of the meaning of the related word, and it is necessary to reconnect the pointers and replace the positions of the related words.

Further, no consideration is given to the case where a plurality of related word dictionaries are used at the same time, which makes it difficult to judge the priority of data in the plurality of related word dictionaries.

On the other hand, in the present invention, as is apparent from the above description, the value of the degree of association is provided in the related word dictionary,
Since the meaning of the related word can be substantially changed by changing the value of the degree of association, there is an advantage that the meaning of the related word can be easily changed. Further, even when a plurality of related word dictionaries are used at the same time, there is an advantage that it is possible to easily determine the priorities among the data in the plurality of related word dictionaries by sequentially extracting the ones having a high degree of relevance.

[0068]

As described above, according to the present invention, the related word dictionary storing the related words related to the keyword used for information retrieval is provided, and the related word related to the keyword is extracted from the related word dictionary. By combining the related word and the keyword to create a new keyword and searching the information to be searched with this new keyword, it is not necessary to search the related word for the keyword, and many related words are used as the keyword. It is not necessary to input the related words one by one when searching for information having "," operability is improved, and the search can be completed in a short time. Therefore, it is extremely effective when searching for information having many related words as keywords.

Further, if the conditions for extracting the related words such as the degree of association are set in advance, no response is required when extracting the related words for the keyword, and the information can be searched more effectively.

Further, when the utilization status of the related words used in the information search is stored as learning information, the related words for the keyword can be effectively used in the next information search, and the appropriate search result can be obtained. Can be obtained in a short time.

Further, since the relation between the meaning of the keyword and the related word is managed by the degree of relevance, the meaning of the related word can be changed only by changing the value of the degree of relevance, and the related word dictionary can be maintained. It can be done easily.

Further, even when a plurality of related word dictionaries are provided, it is only necessary to calculate the degree of relevance in the plurality of related word dictionaries by a specific calculation formula, and even if a plurality of related word dictionaries are simultaneously used, a plurality of related word dictionaries are used. There is an effect that the priority of the related word between the word dictionaries can be easily determined, and the information can be searched by effectively using the plurality of related word dictionaries.

[Brief description of drawings]

FIG. 1 is a block diagram showing an overall configuration of an embodiment of an information search device according to the present invention.

FIG. 2 is an explanatory diagram of a related word dictionary.

FIG. 3 is a flowchart for explaining the operation of the information search device of the embodiment of FIG.

FIG. 4 is an explanatory diagram of a process of extracting a related word related to a keyword.

FIG. 5 is an explanatory diagram of a related word learning method.

FIG. 6 is an explanatory diagram showing another embodiment of a related word learning method.

FIG. 7 is an explanatory diagram of a related word dictionary having a plurality of related degrees.

FIG. 8 is an explanatory diagram of a related word dictionary that independently has relatedness information and learning information.

FIG. 9 is an explanatory diagram of a method of using a plurality of related word dictionaries.

[Explanation of symbols]

1 ... Input device, 2 ... Display device, 3 ... Search processing device, 4 ...
Information storage device, 5 ... Related word processing device, 5a ... Related word processing device having a plurality of related word dictionaries, 6 ... Related word dictionary, 6a ...
Initial state related word dictionary, 6b ... Related word dictionary after first search, 6c ... Related word dictionary after second search, 6d ... Related word dictionary having learning information independently, 6e ... Multiple related degrees Related word dictionary to have, 6f ... Related word dictionary having relevance information and learning information independently, 6g ... Related word dictionary for computer field, 6h ... Related word dictionary for advertising field, 6i ... Related word dictionary for electronic field , 7 ... Keywords in the related word dictionary, 8 ... Related words for keywords, 9 ... Relevance between keywords and related words, 11 ... Numbers uniquely assigned to pairs of keywords and related words, 12 ... Related words in learning information User name, 13
... Use of related words in learning information, 14 ... Learning information of related word dictionary, 15 ... Utilization field of related words in related information, 16
… The degree of use of the relevance in the related word information.

Claims (5)

[Claims] 1. A related word dictionary that stores related words related to a keyword used for information search, an information storage device that stores search target information with a search keyword attached, and a keyword from the related word dictionary. And a processing device for retrieving information to be searched from the information storage device by using the new keyword by extracting the related word related to, creating a new keyword by combining the related word and the keyword. Characteristic information retrieval device. 2. The related word is stored as a pair with a degree of relevance to a keyword, and the processing device retrieves a related word having a degree of relevance higher than a set value and uses it for information retrieval. 1. The information retrieval device described in 1. 3. A plurality of related word dictionaries, each of which stores a related word paired with a degree of relevance to a keyword, calculate the stored related degree by a predetermined calculation formula to obtain the related degree, and calculate the related degree The information retrieval apparatus according to claim 1, wherein the related words satisfying the degree are extracted from a plurality of related word dictionaries. 4. The related word is stored as a pair with a degree of association with a keyword and an actual degree of utilization, and the processing device refers to a degree of association after correction in which the degree of association is corrected with the degree of utilization, and is equal to or more than a set value. The information retrieval apparatus according to claim 1, wherein a related word having a degree of relevance of is retrieved and used for information retrieval. 5. The information retrieving apparatus according to claim 4, wherein the degree of use is updated each time information is retrieved.