Mandarin-English Information (MEI)

Abstract

We describe a system which supports English text queries searching for Mandarin Chinese spoken documents. This is one of the first attempts to tightly couple speech recognition with machine translation technologies for cross-media and cross-language retrieval. The Mandarin Chinese news audio are indexed with word and subword units by speech recognition. Translation of these multiscale units can effect cross-language information retrieval. The integrated technologies will be evaluated based on the performance of translingual speech retrieval.