Automatic assessment of foreign speakers' pronunciation of dutch

Abstract

relates to similar studies carried out elsewhere. The aim of the research reported on here is to develop a system 2. PREVIOUS STUDIES for automatic assessment of foreign speakers’ pronunciation of Dutch. In this paper similar studies carried out for English are 2.1. Automatic scoring of pronunciation quality first examined. Subsequently, suggestions are made for partly improving the methodology that is usually adopted in research In the various methods for automatic pronunciation assessment on automatic pronunciation assessment. Finally, an experiment developed so far (e.g. [1, 4]) different machine measures have is presented in which automatic scores of telephone speech been used for automatic scoring: HMM log-likelihood scores, produced by native and nonnative speakers are compared with timing scores, phone classification error scores and segment scores assigned by human raters. The approach used in this duration scores. Recently, also phone log-posterior probability experiment is compared with those of previous studies. scores have been investigated by [5].