JPS6355624A - Collation device for address dictionary - Google Patents

Abstract

PURPOSE:To shorten an address collating time when an address collating range can limited by registering and storing addresses after sorting previously them by ZIP codes. CONSTITUTION:The addresses sorted previously by ZIP codes are sorted again by the number of characters and several leading characters are registered and stored in an address dictionary 6. In a collation mode the ZIP codes 11 and the input words 21 are outputted and set to a ZIP code input register 1 and an input word register 2. Based on the ZIP code 11, the chain information 31 are outputted via a ZIP code index table pointer 3. Then the address chain information 41 are outputted from a word length group index table pointer 4 with reference to those words 21 and information 31. While the chained addresses 61 are outputted to a dictionary reading part 5 from the dictionary 6. In such a way, the territory properties are controlled by the ZIP codes, therefore the read character strings can be limited. Then the address collating time is shortened.

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Applications] The present invention relates to an address dictionary matching device that recognizes addresses written on paper sheets such as mail by using a large-capacity address dictionary.

[Conventional technology]

Conventionally, this type of address dictionary matching device classifies all addresses by paying attention to the number of characters in the address or the first few characters and registers and stores them in advance. The registered location of the address was serialized and verified based on the number of characters or the first few characters. As a conventional technique, there is Japanese Patent Application No. 59-244544 proposed by the present inventors as a recognition and collation device for Roman addresses.

[Problem that the invention seeks to solve]

The conventional address dictionary matching device described above classifies addresses using the number of characters and the first few characters, so the number of registered addresses is averaged out regardless of regionality, and the range of address matching is limited in advance. Even if it is possible, there is a drawback that the collation time is not shortened.

SUMMARY OF THE INVENTION An object of the present invention is to eliminate such drawbacks and provide an address dictionary matching device that can shorten matching time when the range of address matching can be limited.

[Means for problem decomposition method]

The configuration of the present invention is to provide an address dictionary matching device that recognizes an address by comparing a read character string outputted from a character reading means and consisting of an input character string and a postal code with an address dictionary. A registration storage means that classifies and registers according to the postal code to which the address belongs, and reads out the address registration position in a chain based on the postal code and number of characters of the address registered in this registration storage means, and reads out the mail from the character reading means. The present invention is characterized by comprising address verification means for verifying numbers and input character strings.

[Effect]

According to the configuration of the present invention, after the address is classified by postal code in the registration storage means that becomes the address fJP, the classification is performed by paying attention to the number of characters and the first few characters, and the postal code of the input character string is Since the registered positions of the address are chained by the number of characters and the first few characters, it is possible to efficiently match the input character string with the chained address.

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

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

In the diagram, 1 is Postal I! ! Number input register, 2 is input word register, 3 is postal code index table pointer, 4 is word length group index table pointer, 5 is dictionary reading section, 5
is an address dictionary, and 7 is an address matching unit. Addresses classified in advance by postal code are further classified and registered and stored in the address dictionary 6 by the number of characters and the first few characters. Based on the result of character reading of the postal code and input word read by the character reading means, the postal code 11 and input word 21 are set in the postal code register 1 and the input word register 2 and output. Chain information 31 is output from the zip code index table pointer 3 according to the zip code 11, and address chain information 41 is output from the word length group index table pointer 4 by referring to these input words 21 and chain information 31. The chained addresses 61 from the address dictionary 6 are outputted to the dictionary reading section 5, and since regionality is controlled by postal code in this way, it is possible to limit the read character strings and shorten the collation time.

I. Effects of the Invention] As explained above, the present invention makes it possible to limit the range of verification for read character strings by classifying addresses roughly by postal code and then registering and storing the addresses. This has the effect of significantly shortening the matching time.

[Brief explanation of drawings]

FIG. 1 shows a professional address dictionary matching device according to an embodiment of the present invention.
This is a diagram. 1... Postal code input register, 2... Input word register, 3... Postal code index table pointer, 4...
word length group index table pointer, 5... dictionary reading section, 6... address dictionary, 7... address matching section, 11...
- Postal code, 21... Input word, 31... Chain information, 41... Address chain information, 51... Verification information, 61
... Address output, 7] ... Light output.
, 71, No Kaya / Flash

Claims (1)

[Claims] In an address dictionary matching device that recognizes an address by comparing a read character string outputted from a character reading means and consisting of an input character string and a postal code with an address dictionary, addresses are classified by the number of characters in the address and the postal code to which the address belongs. The address registration position is read out in a chain based on the registration storage means that has been registered and stored, and the postal code and number of characters of the address registered in this registration storage means, and the postal code and input character string from the character reading means are read out. An address dictionary matching device comprising: address matching means for matching.