JPH0256068A - Chaining search system for related word - Google Patents

Abstract

PURPOSE:To obtain high quality retrieved result by searching the related word of a retrieving term, further regarding the found related word as another retrieving term, successively searching the related word of the retrieving term, and creating the term as a retrieving expression. CONSTITUTION:An inquiry 5 as the retrieving vocabulary to be searched is inputted to a dictionary reference part 4, and the dictionary reference part 4 examines whether the retrieving term exists in a related word dictionary 3 or not. When the retrieving term is found, the dictionary reference part 4 repeatedly executes chain processing to regard each existing related word as each retrieving term, further examine whether the related word of each retrieving term exists or not, and store the found related word in a vocaburaly retaining table until a point comes where no related word exists. Further, a term assembling part 6 assembles the term inputted to the inquiry 5 and the term stored in the term retaining table into a retrieving expression 7. The retrieving expression 7 is created by combining the retrieving vocabulary and the related word with an OR operator.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention repeatedly searches a related word dictionary for a search term, its related words, and related words using the related word as a search term. Master the chain search method for related words generated as expressions.

[Conventional technology]

Traditionally, this type of related word search method has two methods: if the database to be searched has a search terminology glossary, the searcher searches for the related term needed in it, and the other is the search method that requires a search terminology glossary. For example, there is a method in which a necessary related word is found through trial and error using an operator specific to the search system.

[Problem to be solved by the invention]

Among the conventional methods described above, the former method has the disadvantage that it is impossible to search for a corresponding term unless one is familiar with the search terminology list included in the database. In addition, with the latter method, results may not be obtained unless the search is repeated many times, or the terms selected will differ depending on the searcher, as the searcher may not be familiar with the area of the database being searched.

The disadvantage is that search results may be omitted.

[Means to solve the problem]

The present invention aims to provide a related word chain search method that can solve the above-mentioned problems, and is a dictionary creation means that inputs registered terms consisting of a search term and its related words and stores them in a related word dictionary. and.

It is checked whether or not a word related to the search term inputted as one query exists in the related word dictionary.

Furthermore, each of the related words found is considered as a search term, and whether or not the related word exists is determined.

The apparatus includes a dictionary reference means for repeatedly searching until there are no related words to be considered as search terms, and a term assembling means for assembling a term input as a query and related words found in the related word dictionary as a search formula.

〔Example〕

Next, the present invention will be explained with reference to the drawings.

FIG. 1 is a longitudinal sectional view of an embodiment of the present invention.

FIG. 2 is an example of search terms and related words. The term "search term" used herein is a keyword used in a search, and is a key term for extracting related words.

FIG. 3 shows the contents of a related word dictionary configured to allow related words to be found from search terms. The related word dictionary uses search terms as an index.

It consists of an index section configured to allow related words to be searched from index data when found, and a data section that stores the related words themselves. The index data includes the physical address of the location where the related word is stored. In other words, the search term and the related word are associated with each other by the address at which the related word, which is the data part, is stored in the dictionary.

FIG. 4 shows the contents of a term storage table that stores found related terms.

In Fig. 1, 1 is a registered term consisting of a search term and related words stored in the related word dictionary 3, and 2 is a registered term that reads registered term 1 and stores the related term in the data section as shown in Fig. 3. , is a dictionary creation unit for storing search terms in an index unit using their physical addresses as data. Reference numeral 4 denotes a dictionary reference section in which a query 5, which is a search term to be searched for, is inputted, and it is checked whether the search term exists in the related word dictionary 3.

When a search term is found, the dictionary reference unit 4 considers each existing related word as a search term, further checks whether the related word exists, and stores the found related words in the term storage table. This process is repeated in a chain until there are no more related words to be considered as search terms. Reference numeral 6 denotes a term assembling unit that creates a search expression 7 from the terms input in the query 5 and the terms stored in the term storage table. Search formula 7 is created by combining the search term and related terms using an OR operator.

For the sake of explanation, it is assumed that registered term 1 is shown in FIG.

Create a beginning related word dictionary 3. The dictionary creation unit 2 uses the first search term "car" and related words "car operation", "
Load "Automobile Road".

Two related words are stored together in the data section. When stored, the physical address in the data section is clear.

Next, the search term "automobile" is stored in the index section.

At this time, the physical address of the related word previously stored is stored as index data. In the same manner, the search terms "car operation" and "motor road" and related words are stored in the index section and the data section, respectively.

In this way, the related word dictionary 3 is created, and the dictionary can be referenced. When "automobile" is input as inquiry 5, dictionary reference section 4 searches whether or not "automobile" is included in the index section of related word dictionary 3. If found, the physical address stored as index data is traced and the data section is referenced. As shown in Figure 3, the corresponding address No. 1 contains the related word "car operation".
, "motor road" is stored. The found related words are sequentially stored in the term storage table as shown in FIG. continue.

The related terms "automobile operation" and "automobile road" are each regarded as search terms, and a search is made to see if the terms exist in the index section. It can be seen that the search term "vehicle operation" has the related words "oversteer" and "understeer", and the search term "automobile road" has the related words "expressway". In this way, the related word dictionary 3 is searched until there are no related words belonging to the search term, and the found related words are stored in the term storage table.

When the search in the related word dictionary 3 is completed, the term storage table is passed to the term assembling unit 5. here.

The individual terms are combined using an OR operator with the search term input in 9-query 5 at the beginning, and a search expression 7 is generated as follows.

Further, as an example, assume that "automatic gear shift" is input in inquiry 5. In this case, it can be seen that the search term "automatic transmission" does not exist in the index section of the related word dictionary 3.

The term assembling unit 6 generates only the term "automatic transmission" in the query 5 as the search expression 7.

〔Effect of the invention〕

As explained above, the present invention searches for words related to a search term, regards the found related terms as search terms, sequentially searches for the related terms, and creates the terms as a search expression.

Check the search glossary to find the related words you need,
This eliminates the need for trial-and-error searches using a search system operator, and has the effect that all searchers can obtain search results of the same high quality.

[Brief explanation of the drawing]

Fig. 1 is a longitudinal sectional view of an embodiment of the present invention, Fig. 2 is a sectional view of a search term and related words, Fig. 3 is a sectional view of a related word dictionary, and Fig. 4 is a part of a term storage table. FIG. 1...Registered term, 2...Dictionary creation section, 3...Related word dictionary, 4...Dictionary reference section, 5...Inquiry, 6.
...Term assembly section, 7...Search formula. Figure 1 Figure 4

Claims (1)

[Claims] 1. A dictionary creation unit that inputs registered terms consisting of search terms and their synonyms and related words (hereinafter referred to as related terms) and stores them in a related word dictionary; ,
It is checked whether a related word exists for the search term entered as a query, and each related word found is considered as a search term, and whether the related term exists or not is compared to the search term. The present invention is characterized by comprising a dictionary reference section that repeatedly searches until there are no related words to be considered, and a term assembly section that assembles a term input as an inquiry and related words found in the related word dictionary as a search formula. A chain search method for related words.