Japanese language sentence dividing method and apparatus

• US 5,029,084 A
• Filed: 03/10/1989
• Issued: 07/02/1991
• Est. Priority Date: 03/11/1988
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Morohasi, Masayuki, Umeda, Shigeki
• Primary Examiner(s): MacDonald, Allen R.
• Assistant Examiner(s): NOT, DEFINED
• Application Number:
  US07/321,812
• Time in Patent Office: 844 Days
• Field of Search: 364/200, 364/900, 364/419
• US Class Current: 704/9
• CPC Class Codes: G06F 40/53 Processing of non-Latin text kana-to-kanji conversion G06F40/129; vowelisation G06F40/232

Abstract

A Japanese language sentence containing a word not registered in an electronic dictionary is divided by following a series of predetermined rules. When more than one division of such a sentence is possible, an evaluation is made in order to determine the best division of the sentence containing a word not registered in the dictionary.

Claims

1. A Japanese language sentence dividing apparatus comprising:

dictionary means containing definitions, rules, tables, and words;

first sentence dividing means coupled to said dictionary means for dividing an inputted Japanese language sentence by referring to said dictionary means;

detecting means coupled to said first sentence dividing means for detecting when said first sentence dividing means encounters a word which is not registered in said dictionary means;

temporary dividing means coupled to said detecting means for dividing a character string containing at least a word which is not registered in said dictionary means into one or more realizable forms of partial character strings, each string containing at least one character, at dividing points in response to said detecting means and said dictionary means;

means for matching each of the partial character strings located between dividing points in each of said realizable forms from said temporary dividing means with a word in said dictionary means;

evaluation means for evaluating said realizable forms by counting the number of characters contained in said partial character string having successfully matched a word in said dictionary means, and

a second sentence dividing means for dividing said inputted Japanese language sentence containing words which are not registered in said dictionary means, in a way that results in a best division of said sentence.

2. A Japanese language sentence dividing apparatus as claimed in claim 1, wherein said evaluation means evaluates said realizable forms by counting the number of partial character strings into which said realizable form is divided.

3. A method of dividing a Japanese language sentence comprising the steps of:

providing dictionary means containing definitions, rules, tables and words;

dividing, in a first sentence dividing means, an inputted

Japanese language sentence by referring to said dictionary means;

detecting when said first sentence dividing means encounters a word which is not registered in said dictionary means;

temporarily dividing a character string containing at least one word which is not registered in said dictionary means into one or more realizable forms of partial character strings, each string containing at least one character, at dividing points in response to the detecting of a not registered word;

matching each of the partial character strings located between dividing points in each of said realizable forms with a word in said dictionary means;

evaluating said realizable forms by counting the number of characters contained in said partial character string having successfully matched a word in said dictionary means, and

dividing, in a second sentence dividing means, said inputted Japanese language sentence containing words which are not registered in said dictionary means, in a way that results in the best division of said sentence.

4. A method of dividing Japanese language sentence as claimed in claim 3, wherein said evaluating said realizable forms counts the number of the partial character strings into which said realizable form is divided.

5. A Japanese language sentence dividing apparatus comprising:

memory means;

first sentence dividing means for dividing an inputted Japanese language sentence by character type according to character type definition and division determination rules contained in said memory means and providing a first output;

second sentence dividing means for dividing said first output by function word strings according to a function word table contained in said memory means and providing a second output;

third sentence dividing means for dividing said second output by content word dictionary and function word table according to a content word dictionary and function word table contained in said memory means and providing a third output;

fourth sentence dividing means for dividing said third output according to unregistered word deduction rules according to the content word dictionary, the function word table, unregistered word deduction rules and affix table and connection rules contained in said memory means and providing a fourth output, and

adjustment means for adjusting said fourth output by compound word synthesizing rules according to compound word synthesizing rules contained in said memory means.

6. A Japanese language sentence dividing apparatus as set forth in claim 5, wherein said fourth sentence dividing means includes evaluating means for evaluating said third output for providing a best said fourth output.

7. A Japanese language sentence dividing apparatus as set forth in claim 6, wherein said evaluating means includes counting means for counting the number of characters in said third output having matched a word in the content word dictionary in said memory means.

8. A method of dividing a Japanese language sentence comprising the steps of:

providing a memory means;

dividing an inputted Japanese language sentence by character type according to character type definitions and division determination rules contained in said memory means;

dividing the sentence divided by character type by function word strings according to a function word table contained in said memory means;

dividing the sentence divided by function word strings by content word dictionary and function word table according to a content word dictionary and function word table contained in said memory means;

dividing the sentence divided by content word dictionary and function word table according to unregistered word deduction rules according to the content word dictionary, the function word table, unregistered word deduction rules and affix table and connection rules contained in said memory means, and

adjusting the sentence divided according to unregistered word deduction rules by compound word synthesizing rules according to compound word synthesizing rules contained in said memory means.

9. A method of adjusting a Japanese language sentence as set forth in claim 8, wherein said adjusting according to unregistered word deduction rules divides the sentence by counting the number of partial character strings into which the sentence is divided and performing the dividing based upon said counting.

10. A method of dividing a Japanese language sentence as set forth in claim 8, wherein said adjusting according to unregistered word deduction rules includes evaluating the sentence divided by content word dictionary and function word table for providing a best said sentence divided by unregistered word deduction rules.