Browse Prior Art Database

Method to Supplement a Period into Japanese Text Based on the Space Density and the Regularity for Correct Speech Synthesis

IP.com Disclosure Number: IPCOM000112898D
Original Publication Date: 1994-Jun-01
Included in the Prior Art Database: 2005-Mar-27
Document File: 2 page(s) / 63K

Publishing Venue

IBM

Related People

Ichikawa, O: AUTHOR

Abstract

A program is disclosed that supplements a Japanese period into a Japanese text by using two parameters of the space density and the regularity so that a speech synthesizer can speak the text with fewer errors.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 52% of the total text.

Method to Supplement a Period into Japanese Text Based on the Space
Density and the Regularity for Correct Speech Synthesis

      A program is disclosed that supplements a Japanese period into
a Japanese text by using two parameters of the space density and the
regularity so that a speech synthesizer can speak the text with fewer
errors.

      Most of the speech synthesizers regards a period as a separator
of the sentences.  Therefore if a period between two consecutive
sentences is omitted, they may be unexpectedly concatenated and
spoken incorrectly.  For example, a title, a header, a destination
address and a spread sheet do not have a period as a terminator.  In
this program, two parameters are introduced to supplement a period
into the space between two sentences where a period was omitted.
They are defined at each space and each line in the Japanese text.

Space Density (K):

 K = (length of the space) / ( MIN( length of the character before
      the space, length of the character after the space,
      CONSTANT1 ) )

o   Where, MIN() is an operation to take the minimum item from the
    candidates.  CONSTANT1 = 5.

o   A space at the end of the line and a space at the beginning of
    the next line should be concatenated for the length counting.
    Carriage Return and Line Feed at the end of the line must be
    ignored.

o   Length of the character after the space should be zero if the
    space is at the end of the line.

Regularity (P):

P = MAX (P+ , P-), (If the line is between other lines)
  or P+,            (If the line is the top line)
  or P-,            (If the line is the last line)
  or P0             (If the line is both the top and the last line)

          P+ = (Coincidence Count of the line and the next line)
          / ( MAX( length of the line, length of the next line) )
          x STANDARD_LINE_LENGTH

          P- = (Coincidence Count of the line and the before l...