Many of learners may not be accustomed to viewing speech in the form of waveforms and speech rhythm plots. To solve such a problem, our rhythm training system has adopted an approach that sees rhythm as a visualization of the learners' utterances. Furthermore, we have implemented our system based on user-friendly graphic interface and a graphical feedback method so that they can manipulate it at their own pace. Figure 5.16 shows an example of the user interface for the mora rhythm instruction module using the Soft Template ``UaNUe:'', while a non-native learner of Japanese pronounced the [kaNke:] (relation) for the target rhythm pattern.
図 5.16: Example of user interface for training based on templates
There are two main windows in this interface: the target template
window, and the learner window (from top to bottom). The means
of each mora ratio for the target rhythm pattern ``UaNUe:'' are
represented with a rectangle in the template window. A learner can
record his (or her) utterance and then confirm whether his (or her)
speech rhythm is good or not on the spot. If abnormal, the
corresponding rhythm rectangle is filled with a shadow. By clicking
the PLAY button on the bottom of the window, a learner can listen to
the utterance again for self-feedback. A proper rhythm range can be
adjusted by changing the n of
according to the
intensity of instruction.
Using such templates, learners are expected to be able to acquire the same rhythm pattern words without any difficulty, e.g., the rhythm template ``UaNUe:'' covers [saNse:] (agreement), [kaNse:] (sensitivity), and [haNte:] (judgement), etc. For rare mora rhythm patterns, our CALL system also offers a word-by-word instruction module using model rhythm speech.