For language modeling of spontaneous speech recognition, we propose a style transformation approach, which transforms written texts to a spoken-style language model. Since these two styles are largely different and thus direct transformation is difficult, we cascade two transformation methods; rule-based transformation to rewrite written-style texts to intermediate polite-style texts, and statistical transformation of language model from polite style to faithful style which is suitable for ASR. In an experimental evaluation on real lecture speech, the proposed transformation approach realized higher performance than conventional linear interpolation method.

Index Terms: automatic speech recognition, lecture speech, language model, style transformation