Browse Prior Art Database

Learning of Contiguous Words in Kana-To-Kanji Conversion

IP.com Disclosure Number: IPCOM000122239D
Original Publication Date: 1991-Nov-01
Included in the Prior Art Database: 2005-Apr-04
Document File: 3 page(s) / 54K

Publishing Venue

IBM

Related People

Ono, K: AUTHOR [+3]

Abstract

Disclosed is a new Learning Logic of Kana-to-Kanji Conversion that can learn contiguous two words, each of which may be converted individually, as one word. Conversion logic of prior art

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 100% of the total text.

Learning of Contiguous Words in Kana-To-Kanji Conversion

      Disclosed is a new Learning Logic of Kana-to-Kanji
Conversion that can learn contiguous two words, each of which may be
converted individually, as one word.
Conversion logic of prior art

      Leaning logic of prior art cannot learn "TSURUMA" (place name)
shown in Fig.1 as one word.  Therefore, the user cannot always
convert "TSURUMA" to "t/M" without word registering.
 - YOMI of "TSURU" has 'T','t','TT','tT' as SEISHO.
 - YOMI of "MA" has 'M','m','mm' as SEISHO.
 - The SEISHO of "TSURUMA" is "t/M".
 - '/' indicates the boundary of the phrase.
 - SEISHO is the expected conversion result.
Objective and Advantages of new learning logic

      This new learning logic learns contiguous two words as one
word.  But, the word which is not the result of the initial
conversion must be included in these two words.

      This new learning logic is useful when the user converts a
proper noun, such as a place name, a company name and a person's
name.
The way of learning

      The new way of learning is shown in Fig. 2.
Flow chart

      Kana-to-Kanji conversion program using the new learning logic
is shown in Fig. 3.