JPH04670A - Information retrieving device - Google Patents

Abstract

PURPOSE:To quickly select a keyword for executing a reduction retrieval by counting the retrieved information by a keyword counting means for reducing a result of retrieval by an information retrieving means, and presenting only an effective related keyword by a related keyword presenting means. CONSTITUTION:An information storage means 1 stores information and a keyword by relating them to each other, and a retrieval condition input means 2 designates the contents of the information to be retrieved, in the stored information by combining the keyword and an operator. Subsequently, an information retrieving means 3 interprets a retrieval condition designated by the retrieval condition input means 2, checks up the contents of information storage means 1 and fetches information being applicable to its retrieval condition, and a keyword counting means 4 counts a keyword corresponding to the information retrieved by the information retrieving means 3. Next, by a related keyword presenting means 5, only an effective related keyword is presented. In such a way, a user examines its selected information, and can select quickly the keyword for executing a reduction retrieval.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of the Invention The present invention relates to an information retrieval device that stores information such as documents, images, audio, etc. in association with keywords, and outputs information that matches search conditions input by a user.

Conventional Technology Conventionally, when searching for desired information from a large amount of information, keywords are assigned to each piece of information and stored, and when searching, a conditional expression that includes the keyword is input, and a keyword that matches it is searched. A common method is to output the information that it has. For example, “A and B”
If you enter the search condition, l A + and B'
Only information with both keywords will be searched.

This type of search method can be used in two ways depending on how the keywords are given.
It can be divided into The first one is called the "controlled keyword method J," which is a method in which the words that can be used as keywords are limited in advance.
The second method is called the ``free keyword method,'' which does not limit the keywords that can be used. In this case, it is particularly difficult to understand which words should be used as keywords in the free keyword method.

Furthermore, it is difficult for users to express their search requests using conditional expressions that include "keywords" and "operators."
It is rare to be able to create an appropriate conditional expression in one go. Conventionally, users input such a conditional expression, look at the list of search results returned, and change the conditional expression over and over again to obtain a set of information that closely matches their search request. That is the reality. Therefore, in conventional systems, there is almost no guidance from the system at the stage of changing the conditional expression (only displaying all registered keywords), and the burden on the user is enormous. In many cases, users are unable to obtain the information they are looking for or give up on using the information retrieval system altogether.

Problems to be Solved by the Invention In conventional devices such as those described above, generally, the first conditional expression entered manually searches for information in a wider range than the information desired by the user. If the search results are broader than the range of information desired, the user must further narrow down the range of information using conditional expressions from the search results.

However, conventional systems only provide a function to present a list of keywords registered in the system as a guide to the user at this stage. This list contains an extremely large number of keywords (several hundred to several thousand), and it is virtually impossible for the user to scrutinize them all. Moreover, most of the keywords included in the list are unnecessary keywords (keywords unrelated to the search results) when narrowing down the information, and originally do not need to be presented to the user.

In this way, with conventional information retrieval systems, users have no way of knowing which keywords to use to narrow down their search results; Search formulas were constructed using keywords that were likely to be related, which required a great deal of effort and time.

Means for Solving the Problems Therefore, in order to solve such problems, the present invention provides an information storage means for storing information and keywords in association with each other, and the information stored in the information storage means is provided. A search condition input means is provided for specifying the content of the information to be searched for by combining a keyword and an operator, and the search condition input means interprets the search condition specified and examines the content of the information storage means to create the search condition. an information retrieval means for retrieving information applicable to the above, a keyword counting means for counting keywords corresponding to the information searched by the information retrieval means, and a related keyword presentation means for presenting the keywords counted by the keyword counting means. Established.

As a result, when narrowing down the results searched by the information search means, the keyword counting means counts the information searched for, and the related keyword presentation means presents only valid related keywords, so the user can It becomes possible to quickly select keywords for narrowing down the search by examining the information obtained.

Example An embodiment of the present invention will be described based on the drawings.

In this embodiment, as shown in FIG. 1, an information storage means 1,
It consists of a search condition input means 2, an information search means 3, a keyword counting means 4, and a related keyword presentation means 5. Therefore, the configuration of each of these parts will be described below.

First, the information storage means I stores information and keywords in association with each other. That is, information to be searched (documents, images, audio, etc.) and keywords expressing the content are stored in association with each other. In other words, there are generally multiple keywords that correspond to one piece of information, and by specifying a certain keyword, all the corresponding information can be retrieved immediately, and conversely, by specifying a certain piece of information, all the corresponding information can be retrieved immediately. It can be taken out.

Next, the search condition input means 2 specifies the content of the information to be searched among the information stored in the information storage means 1 by combining a keyword and an operator. Examples of such operators include:

AandB: Information with both the keywords A and B. AorB: Information with at least one of the keywords A or B. notA: Information without the keyword A. can also be used in combination, such as ((A and B) or C).

Note that the search conditions are entered using the keyboard.

Next, the information retrieval means 3 interprets the search conditions specified by the search condition input means 2, examines the contents of the information storage means 1, and extracts information that applies to the search conditions.

Next, the keyword counting means 4 counts the keywords corresponding to the information searched by the information searching means 3. In other words, we find all the keywords that correspond to each piece of information retrieved for inputting a certain search condition (hereinafter referred to as search information), and calculate how many types of information in the search information each keyword corresponds to. Calculate the number of animals. A specific example of this counting will be described based on Table 1.

Table 1 shows an example of a correspondence table between keywords and information. Now, assume that information 3, information 4, and information 5 are search information.

Since the keyword A-CE corresponds to any of information 3, 4, and 5, it is determined how many types of search information the keyword corresponds to.

Keyword A corresponds to information 3, 4, and 5, keyword C corresponds to information 4, and keyword E corresponds to information 4 and 5. Therefore, keyword A corresponds to three types of search information, keyword C corresponds to one type, and keyword E corresponds to two types of search information.
This corresponds to each type of search information.

Next, the related keyword presentation means 5 presents only the keywords counted by the keyword counting means 4. In other words, only keywords that are effective in narrowing down search information are presented. Various presentation orders can be considered in this case; for example, the following methods are available.

The first method is to present the search information in the order in which the number of search information corresponding to the keyword is close to half of the total number of search information.

Now, if the total number of search information is N and the number of search information corresponding to a certain keyword is X, then the effect of narrowing down the search information is
The keyword corresponding to the number of pieces of search information and the keyword corresponding to the number of pieces of search information are exactly the same. The reason is that by using the not operator, it is possible to narrow down the search information to X pieces or N−X pieces using the same keyword.

Therefore, if keywords are presented in the order in which the number of corresponding search information is close to half of the total number of search information, the keywords can be presented in order of similar effectiveness in narrowing down the search information.

The second method is to present search information in descending order of the number of pieces of information corresponding to a keyword.

As a practical matter, it is thought that there are few keywords that correspond to more than half of the total number of search information.

In that case, if the keywords are presented in order of the number of corresponding search information items, the processing is simpler than the method described above, and the same effect can be obtained.

Also, when presenting a keyword, if you also present not only the keyword but also what percentage of the total search information corresponds to the keyword,
This will be of great help to users in predicting how many types of new search information will be obtained by new conditional expressions.

Next, FIG. 2 shows the flow of processing when searching for information using the various means described above. First, a user creates a search expression that expresses the information he or she wants by combining keywords and operators. At this stage, refer to the list of all keywords registered in the system. Second, look at the list of search results and, if the information you want is included, suggest related keywords. Third, find keywords that can effectively narrow down your search results and create search conditions. By repeating the second and third steps, you can obtain only the information you want.

Effects of the Invention The present invention provides an information storage means that stores information and keywords in association with each other, and specifies the content of the information to be searched among the information stored in the information storage means by combining the keyword and an operator. Provide search condition input means to
Information retrieval means is provided which interprets the search conditions specified by the search condition input means, examines the contents of the information storage means, and retrieves information that meets the search conditions, and searches for keywords corresponding to the information retrieved by the information retrieval means. Since a keyword counting means for counting is provided and a related keyword presenting means for presenting the keywords counted by this keyword counting means, keyword counting is used to narrow down the search results by the information retrieval means among these various means. It is possible to count the information searched by the means, and to present only valid related keywords by the related keyword presentation means.This allows the user to carefully examine the selected information and narrow down the search. Keywords can be selected quickly.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing an embodiment of the present invention, and FIG. 2 is a flowchart showing the flow of information search processing. 1... Information storage means, 2... Search condition input means, 3
... Information retrieval means, 4... Keyword counting means, 5
... Related keyword presentation means Applicant: Rico Co., Ltd.

Claims (1)

[Claims] information storage means for storing information and keywords in association with each other; search condition input means for specifying the content of the information to be searched among the information stored in the information storage means by combining keywords and operators; information retrieval means for interpreting the search conditions specified by the search condition input means, checking the contents of the information storage means and retrieving information that meets the search conditions; and counting keywords corresponding to the information retrieved by the information retrieval means. An information retrieval device comprising a keyword counting means and a related keyword presenting means for presenting the keywords counted by the keyword counting means.