In the filled-pause skip experiments, the number of filled-pauses in the vocabulary was 109. The experimental results are shown in Table 4 .
speaking | order | original | Filled-pauses | Phone-strings | |
style | word trigram | skip | skip | ||
read | 1 | 90.5% | 82.1% | 92.7% | |
8 | 98.5% | 95.4% | 98.9% | ||
semi | 1 | 42.0% | 63.7% | 65.6% | |
spontaneous | 8 | 43.9% | 80.1% | 72.9% | |
spontaneous | 1 | 25.3% | 42.7% | 42.0% | |
8 | 39.5% | 61.1% | 63.4% |
As can be seen, the phone-strings skip method obtains a higher performance than the filled-pauses skip method. In addition, the sentence recognition rate for spontaneous speech is 42.0% as opposed to 25.3% with no such procedure; including the semantically correct sentences, the sentence recognition rate was about 75%. This means that this algorithm is effective in spontaneous speech recognition.