The Technique of English Word Syllable Division in Speech Synthesis Based on Neural Network

Table 2

Comparison of recognition rate between the decision tree model and the method in this paper.

Training data size	Tree model (%)	KASWT model (%)
400	72.25	72.60
800	87.45	93.75
1200	87.40	88.65
1600	87.75	89.65
2000	90.35	91.80