New methods for detecting characters wrongly deleted and inserted in Japanese strings and their applicability to DNA chains

摘要

This paper proposes methods to detect and to correct the characters wrongly inserted and deleted in natural language. Natural language is physically different from DNA, however it has a lot of common characteristics in point of medium representing information. Accordingly the methods proposed here are expected to be applied to detect errors in DNA chains. In optical character recognition and continuous speech recognition of a natural language, it has been difficult to detect error characters which are wrongly deleted and inserted. In order to detect and correct these errors, this paper proposes new methods using m-th order Markov chain model for Japanese syllables and "kanji-kana" characters, assuming that Markov probability of a correct chain of syllables or "kanji-kana" characters is greater than that of erroneous chains. From the results of the experiments, it is concluded that the method is useful for detecting as well as correcting these errors.