1996 Fiscal Year Final Research Report Summary
Development of Quantitative Prosodic Analysis System Utilizing the F_0 and Power Data Generated by "Onsei-Rokubunken"
| Project/Area Number |
07680314
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Japanese language education
|
|---|
| Research Institution | University of Marketing and Distribution Sciences |
|---|
Principal Investigator |
AZUMA Junichi Faculty of Commerce, University of Marketing and Distribution Sciences Professor, 商学部, 教授 (90202621)
|
|---|
| Project Period (FY) |
1995 – 1996
|
| Keywords | Japanese speech / Prosody / Onsei-Rokubunken / Quantitative Analysis |
|---|
| Research Abstract |
A program was developed to quantitatively analyze the power and fundamental frequency (F0) measures of the data files generated by "Onsei-Rokubunken, " a speech signal analysis system developed at the Faculty of Medicine, University of Tokyo. The programing was developed by Microsoft Visual C++ so that the prosodic analysis could be done on the platform of Windows95, the most prevailing operating system of today. The program, for example, contains the routines to
・calculate the mean and variance of the F0 and power data
・culculate the correlation coefficient between F0 and power
・culculate major quantitative data representing the dynamic F0 pattern
・display the long-term F0 and power data
and so on.
Data files produced by "Onsei-Rokubunken" of the reading of the nearly identical script in eight different Japanese dialects were analyzed by the newly developed program and the results of the quantitative analyzes were examined to study the dialectal differences in prosody.
|
