For people with visual disabilities, reading Braille text is an important way to acquire information. There are great challenges for Chinese-Braille translation due to the characteristics of word segmentation and tone marking in Chinese Braille. In this paper, a novel scheme of Chinese-Braille translation is proposed. Unlike current methods which use heuristic rules defined by experts for Braille word segmentation, the proposed method performs Chinese-Braille translation based on a Braille Corpus without experts on Braille. Under the scheme, a Braille word segmentation model based on statistical machine learning is trained on a Braille corpus, and Braille word segmentation is carried out using the statistical model directly without the stage of Chinese word segmentation. Tone marking and some special treatment are also performed based on word and rule mining on the Corpus. This method avoids manually establishment of rules concerning syntactic and semantic information and uses statistical model to learn the rules by stealthily and automatically. Experimental results show the effectiveness of the proposed approach.
Improved source-channel models for Chinese word segmentation