| 摘要 |
PURPOSE:To easily register accent information in a dictionary with efficiency by excluding beats which never become the core of an accent among beats indicating the rendering of an input character string and narrowing down and selecting candidates for an accent type, and displaying them and selecting the accent. CONSTITUTION:This device is equipped with a means 3 which analyzes the input character string and narrows down accent type candidates regarding the character string when only the character string is inputted and a means 4 which generates and outputs a synthesized voice corresponding to the character string matching the narrowed-down and selected accent type candidate and prompts the input of confirmation information. Then, a proper accent type is selected among the narrowed-down and selected accent type candidates and its synthesized voice is only confirmed. Consequently, the accent information can easily be registered in the dictionary with efficiency. |
