JPH0567152A - Document retrieving device - Google Patents

Abstract

PURPOSE:To easily retrieve a sentence including synonyms by converting an inputted word into a word number in accordance With a dictionary and retrieving a document data base by the word number. CONSTITUTION:This device consists of a dictionary 101 where words as the object of retrieval are collected, a dictionary retrieval part 103 which retrieves a document data base 102 where words are substituted with word numbers to convert words into word numbers, and a document data base retrieval part 104 which extracts a sentence including converted word numbers from the document data base 102. An inputted word is matched with words of the dictionary 101 in consideration of the ambiguity, and the word is substituted with the word number in the document data base 102, and the document including this word number is extracted from the document data base 102. Consequently, the difference and the fluctuation of expression between the inputted word and words used in the document data base 102 are absorbed in the stage where the dictionary 101 is used to convert the inputted word into the word number, and the sentence including synonyms is easily retrieved.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a sentence search apparatus for extracting a sentence or sentence containing a specific word from a sentence database such as an electronic dictionary.

[0002]

2. Description of the Related Art In a conventional text search apparatus, a system for extracting a text (text) such as an example including a specific character string from a text database has already been created.

[0003]

However, in terms of katakana notation, such as a foreign word, words such as "undine" and "ondine" or "gnome" and "gnome" have the same meaning but different notations (hereinafter Called synonymous different notation)
There is. Therefore, the user has to input the concept to be searched by the notation used in the text database. Also, it is difficult for the search system to match the input word with the sentence database while assuming various notational fluctuations.

The present invention has been made in consideration of such circumstances, and has a dictionary that collects words to be searched in a text database, and replaces the words in the text database with the word numbers of this dictionary. Every time, the input word is converted into a word number by the dictionary, and the sentence database is searched by the word number, thereby providing a sentence search device that facilitates the search of sentences including synonymous notation words. Is.

[0005]

The structure of the present invention will be described with reference to FIG. 1. In the present invention, the notation is replaced by the dictionary 101 in which the words to be searched are collected and the word number of the dictionary 101. A sentence database 102, a dictionary search unit 103 that converts the input word into a word number by searching the dictionary 101, and a sentence that extracts a sentence including the word number converted by the dictionary search unit 103 from the sentence database 102 The database search unit 104 and.

[0006]

According to the present invention, the input word and the dictionary 101
If the input word matches a word included in the dictionary 101, it is replaced with the word number of the word used in the sentence database 102, and the word is used. A sentence including a number is extracted from the sentence database 102. Therefore, at the stage of converting the input word into a word number using the dictionary 101, it is possible to absorb the difference and fluctuation of the notation with the word used in the document database 102.

[0007]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described in detail below based on the embodiments shown in the drawings. The present invention is not limited to this.

FIG. 2 is a block diagram showing the configuration of an embodiment of the present invention. In this figure, 1 is a control unit,
It contains a dictionary table for kana-kanji conversion, which is necessary for inputting words. Reference numeral 2 denotes a display unit including a CRT or LCD used for displaying the extracted text. An input unit 3 includes a keyboard for inputting a word to be searched. Reference numeral 4 is a dictionary search unit for converting a word into a word number. Reference numeral 5 is a sentence restoration unit that replaces the word number in the extracted sentence with a word. Reference numeral 6 is a sentence DB search unit that searches the sentence database by word number. Reference numeral 7 is an index for narrowing the search range of the dictionary by the consonants of the input word. Reference numeral 8 is a word dictionary (hereinafter simply referred to as a dictionary) that associates words with word numbers. Reference numeral 9 is a sentence database including sentences in which words to be searched are replaced with word numbers.

FIG. 3 is an explanatory view showing the structure of the index 7. As shown in the figure, the index 7 is a table 71 of different combinations of consonant Katakana words included in the dictionary 8 and a table of word numbers of the dictionary 8 corresponding thereto.
It consists of 72. The "1" part of the table 71 indicates that it has the consonant shown above. The example in the upper row of the table 71 indicates that the only consonants of Katakana are “d” and “n”, and the corresponding contents of the table 72 are the consonants “d” and “n” in the dictionary 8. It is a Katakana word number that includes only.

FIG. 4 is an explanatory view showing the structure of the dictionary 8.
As shown in the figure, the dictionary 8 includes a writing section 81 and a representative word number 82.
Consists of. In the notation unit 81, the words to be searched for in the document database 9 and the words having the same meaning as the word but different in notation (hereinafter referred to as synonymous notation words) are arranged in ascending order of the character codes. As the word number of each word included in the dictionary 8, the arrangement order in the dictionary 8 is used.
For example, "undine" is the 31st word, so
The word number for "undine" is 00003. Regarding the synonymous different notation word, the word number of the notation used in the sentence database 9 (hereinafter referred to as the representative notation) is entered in the corresponding representative word number 82 column. In the example shown, word number 00
Regarding "Gnome" in 082, "Nome" of word number 00254 is a representative notation, and "Gnome" is a synonymous different term of "Nome".

FIG. 5 is an explanatory view showing an example of sentence data in the sentence database 9. In the sentence database 9, a sentence like (a) in the figure is stored in the form shown in (b) by replacing a word to be searched with a coded word number. [...] in the figure represents the coded word number.

The operation of this embodiment will be described below with reference to FIG. 6 using "ondine" and "gnome" as examples of search words.

In the case of "Ondine", first, after inputting a word (step 21), a simple match with the notation 81 of the dictionary 8 is made (dictionary search 1: step 22). Here, since there is no matching word (step 23), the consonants “d” and “n” included in “Ondine” are used to index 7.
To obtain word numbers "00010" and "00031" .... The words "Undead" and "Undine" in the dictionary 8 corresponding to this word number are sequentially searched, and a similar word "Undine" is obtained from the order of consonants (dictionary search 2: steps 24 and 25). Since "Undine" is a representative notation because the representative word number 82 is empty (step 26), the sentence database is searched with the word number 00003 (sentence DB search: step 27), as shown in FIG. Since the sentence having the word number 00003 is extracted, the sentence restoring unit 5 replaces the word number with the corresponding notation and outputs it (result display: step 28).

In the case of "Gnome", first, the notation 81 in the dictionary 8
And simply try to find a match (Dictionary Search 1: Step 2
2). It matches the word number 00082 "Gnome" in the dictionary 8 (step 23). Since the representative word number 82 of this word is not empty, it is understood that "Gnome" is not a representative notation (step
26). According to the content of the representative word number 82, it is replaced with the word number 00254 of the representative notation "Nome" (word number replacement: step 29). Then, similar to the case of "Ondine", the sentence DB is searched (step 27) and the result is displayed (step 28).

In this way, it is possible to perform a search using a word other than the original notation of the text database. In the above example, when there is a similar word or when the word is represented by a representative word, the user may be informed of the fact.

[0016]

According to the present invention, even if the user does not know the notation of the word used in the sentence database, the sentence can be searched with the notation that he or she knows. Also,
Since the text database is simply matched by word number, high-speed search can be performed.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of the present invention.

FIG. 2 is a block diagram showing the configuration of an embodiment of the present invention.

FIG. 3 is an explanatory diagram showing a structure of an index.

FIG. 4 is an explanatory diagram showing the structure of a word dictionary.

FIG. 5 is an explanatory diagram showing an example of sentence data in a sentence database.

FIG. 6 is a flowchart showing the operation of the embodiment.

[Explanation of symbols]

 1 Control Section 2 Display Section 3 Input Section 4 Dictionary Search Section 5 Text Restoration Section 6 Text DB Search Section 7 Index 8 Word Dictionary 9 Text Database

Claims (3)

[Claims] 1. A dictionary in which words to be searched are collected, a sentence database in which notations are replaced by word numbers in the dictionary, and a dictionary search for converting input words into word numbers by searching the dictionary. And a sentence database search unit that extracts a sentence including a word number converted by the dictionary search unit from a sentence database. 2. The sentence search device according to claim 1, further comprising an index for narrowing down a search target of the dictionary according to a consonant contained in an input word. 3. A word appearing in a sentence database and its synonymous different notation are stored in the dictionary, and word number information of a word appearing in the sentence database is added to the synonymous different notation in the dictionary. The sentence search device according to claim 1, wherein: