JPH0528132A - Word dictionary collating device - Google Patents

Abstract

PURPOSE:To attain the optimum retrieval of words in response to the word features. CONSTITUTION:A field is provided in a word record registered in a word dictionary 6 to store the retrieving conditions. This retrieving condition field stores the proper retrieving conditions of words. At a word collating part 4, the character string inputted from a character input part 2 is collated with the character string read out of the dictionary 6. Then the result of this collation is outputted to a retrieving condition collating part 7. The part 7 performs the collation to decide whether the word collating result satisfies or not the retrieving conditions proper to the words. Then only the word collating results that satisfy the retrieving conditions are outputted to a word deciding part 8. The part 8 outputs the code showing the word that is most coincident with an input character string out of those input word collating results.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a word dictionary collation device for recognizing an input character string by collating a character string input from a character input device with a word registered in a word dictionary.

[0002]

2. Description of the Related Art FIG. 2 is a block diagram showing a conventional word dictionary matching device. The character string data input from the character input unit 2 is converted by the character code conversion unit 3 into a character code suitable for matching with a read code from the word dictionary 25. The read control unit 24 reads a word from the word dictionary 25 and transfers it to the word matching unit 23. In the word matching unit 23,
The character string code transferred from the character code conversion unit 22 is collated with the character string code transferred from the read control unit 24, and the collation result is transferred to the word determination unit 26. The word determination unit 26 outputs the word that most matches the input character string according to the matching result of the word matching unit 23.

[0003]

In the above-mentioned conventional word dictionary collating apparatus, all word records in the word dictionary 25 are searched under the same condition, so that a unique search process according to the feature of the word is performed. It has a drawback that it cannot be executed only for word records having characteristics. For example, if you want to improve the accuracy of the search for specific words,
It is not possible to perform a search by making the matching condition of the word different from other words.

[0004]

The apparatus of the present invention is a word dictionary for searching and matching a word that is the maximum match with the input character string among the words registered according to the character string input from the character input means. In the matching device, a field in which a search condition unique to the word is provided in a word record registered in the word dictionary, and the word dictionary in which the search condition is stored together with the word, and the word dictionary and the character input And a word determining unit that outputs only the matching result corresponding to the word record satisfying the search condition in accordance with the word matching result and the search condition in the word record. .

[0005]

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

FIG. 1 is a block diagram showing an embodiment of the present invention. In this embodiment, the area name in the address notation written on the mail piece is detected, and the mail piece is sorted according to the area name.

The character input unit 2 is provided with data obtained by reading characters written on a mail piece with the character reading device 1. The character code conversion unit 3 converts the character string data input from the character input unit 2 into a character code suitable for matching with a read code from the word dictionary 6. Read control unit 5
Reads a word record from the word dictionary 6 in response to a command from the word matching unit 4. The word dictionary 6 is a dictionary in which a large number of area names are stored in advance, and the search condition field of a specific word record stores a search condition that the matching of the character string must be exactly the same. ..
The word collating unit 4 collates the character string code input from the character code converting unit 3 with the character string code of the area name read from the word dictionary 6 via the reading control unit 5. A “DP matching method” is used as a matching method, and it is possible to deal with errors, omissions, and mixing of input character strings.
The word matching unit 4 outputs only the word record information having a matching result equal to or higher than a predetermined similarity to the search condition matching unit 7. When the area name of the matching result output from the word matching unit 4 corresponds to the area name given the search condition in the word record, the search condition matching unit 7 determines whether the matching of the character strings is completely matched. If they do not match, the output of the matching result information of the word record to the word determination unit 8 is prohibited. If the search conditions match or if the word record is a region name with no search condition,
The information of the matching result of the word record is output to the word determination unit 8.

The word determination unit 8 outputs a code for displaying the maximum matching area name with the character string input from the character reading device 1 from the matching results input from the search condition matching unit 7. The sorting machine 9 sorts the mail pieces according to the area name code input from the word determination unit 8.

Next, the operation of this embodiment will be described with reference to a concrete example. A large number of area names are registered in the word dictionary 6, of which “Chai Wan” and “Causewa” are included.
"yBay", "Yuen Long", "Yuen K"
"Ong" and the like are included.

The area name "Chai Wan" is the personal name "C
It is a character string similar to “han Wan”, “Chan Won”, etc., and moreover, these Chinese names appear frequently. If the address writing area, not the area name writing area, is scanned on the mail by the character reading device 1, as a result, "Chan Wan" and "Chan Won" are input to the character input unit 2.
In the conventional word dictionary collating device, the input character strings are "Chi Wa
It will be erroneously determined to be "n".

In this embodiment, the area name is "Chai Wa".
For an area name such as “n” that does not exactly match but is similar and has a frequently appearing person name, the search condition field of the word record must be completely matched by matching the character string. The specific search condition is not stored in the search condition field of the word record of the other area name.

(1) Now, for example, the character string "Yuan L
When “ong” is input from the character input unit 2, the word matching unit 4 extracts the input character string “Yuan Long” and the character string of the area name read from the word dictionary via the reading control unit 5. , “DP matching method”. In the case of this embodiment, “Yuen Kong”, “Yu”
The information of the matching result is transferred to the search condition matching unit 7 as an area name having a similarity equal to or more than a predetermined value as “en Long”.

The search condition matching unit 7 refers to the search condition field together with the area name transferred from the word matching unit 4, and the area names "Yuen Long" and "Yuen K" are referenced.
Since no specific search condition is stored in the search condition field of "ong", both area names are transferred to the word determination unit.

In the word determination section 8, the input character "Yuan"
The area name that most matches "Long" is "Yuen Lon"
g ", and then" Yuen Long "
Output the code that displays.

(2) Next, for example, the character string "Chan W
When “an” is input from the character input unit 2, as a result of the same operation as the above item (1), the information of the area name “Chai Wan” is input to the search condition matching unit 7 as the matching result of the word matching unit 4. Transferred.

In the search condition matching unit 7, the area name "Chai"
The search condition field is referred to together with “Wan”, but since the condition that the character strings must be matched completely is stored here, it is checked whether or not this condition is satisfied. In this case, the character string "C
"Hai Wan" and the input character string "C from the character input unit 2
Since one character is different and does not completely match with “han Wan”, output of the information of the area name “Chai Wan” to the word determination unit 8 is prohibited.

In the word determination unit 8, since the matching result information is not input from the search condition matching unit 7, the input character string "Cha
A code indicating that the area name matching "n Wan" is not in the word dictionary 6 is output.

As can be seen from the explanations in (1) and (2) above, the character string "Yuan" input from the character input unit 2 is input.
“Long” and “Chan Wan” are both character strings “Yuen Long” and “C” stored in the word dictionary 6.
Although it is a character string that is one character different from "hai Wan", the area name search is successful for the character string "Yuan Long", and the code for displaying the correct area name "Yuen Long" is output from the word determination unit 8. . However, the word determination unit 8 outputs a code that the character string “Chan Wan” of the person's name is not searched as the area name.

In this way, different input character strings having similar degrees of similarity can be judged under different conditions to improve search accuracy.

[0020]

As described above, according to the present invention, the word record stored in the word dictionary is provided with a search condition peculiar to the word, and the matching result of the word record satisfying this search condition is provided. By transferring only the word to the word determination unit, it is possible to perform a unique search process according to the feature of the word, and improve the search accuracy.

[Brief description of drawings]

FIG. 1 is a block diagram of an embodiment of the present invention.

FIG. 2 is a block diagram of a conventional device.

[Explanation of symbols]

 1 Character reading device 2 Character input unit 3 Character code conversion unit 4 Word matching unit 5 Read control unit 6 Word dictionary 7 Search condition matching unit 8 Word judging unit 9 Sorting machine 10 Word dictionary matching device

Claims (1)

Claim: What is claimed is: 1. A word dictionary matching device for searching and matching a word that matches the input character string at a maximum among words registered according to the character string input from the character input means. The word record registered in the word dictionary is provided with a field to which a search condition specific to the word is provided, and the word dictionary in which the search condition is stored together with the word, and the word dictionary and the character input means are used for input. And a word determining unit that outputs only the matching result corresponding to the word record satisfying the search condition according to the word matching result and the search condition in the word record. Word dictionary matching device.