Algorithms for statistical translation of spoken language

In this paper, we describe three approaches to statistical translation and present experimental results. The statistical translation approach uses two types of information: a translation model and a language model. The langua ge model used is a bigram or general m-gram model. The translation model is decomposed into a lexical model and an alignment model. There are three ap proaches that are presented and tested in detail: the quasimonotone alignme nt approach, the inverted alignment approach, and the alignment template ap proach. For each of these three approaches, a suitable search method is pre sented. The system has been tested on a Limited-domain spoken-language task for which a bilingual corpus is available: the Verbmobil task (German-Engl ish, 7000-word vocabulary). We present experimental results for each of the three approaches. The experimental tests were performed on both the text t ranscription and the speech recognizer output.