An automatic approach to “De, Di, De” of spelling errors detection in Chinese text

“De, Di, De” is related to the part-of-speech of the first previous word and the first following word, so their spelling errors can bring trouble to Chinese information processing. This paper propose an automatic approach to “De, Di, De” of spelling errors check. The new method can construct the statistical model of the context, and use this model to detection “De, Di, De” of spelling errors in Chinese text. Experiments are done on one noise free simplified Chinese dataset, one noise free traditional Chinese dataset and one traditional Chinese dataset with noise. The results show that the proposed approach based on the statistical model can achieve the better performance than the detection method based rules.