Application of neural networks into Japanese-to-Braille translation

Braille is an important communication method for the visually disabled. In translating Japanese-to-Braille, it is necessary to insert spaces among words. In this paper, we applied neural networks, and determined where the spaces were inserted among words. Data was input into neural networks by the unit of two words with a part of speech and battuta of each word. The experimental results showed that the proposed system has high accuracy for inserting spaces among words.