Constituent-based Accent Prediction

Near-perfect atttomatic accent assignment is attainable,for citation-s~le speech, but better computational models art, needed to predict accent in extended, spontaneous discourses. This paper presents an empirically motivated theory of the discourse ]bcttsing nature of accent in spontaneous speech. Hypotheses based on this theory lead to a new approach to accent prediction, in which patterns o f deviation from citation.fbrm accentuation, defined at the constituent or noun phrase level, are atttomatically learned from an annotated corpus. Machine learning experiments on 1031 noun phrases from eighteen spontaneous direction-giving monologues show that accent assignment c~tn be signoqcantly improved by up to 4%-6% relative to a hypothetical baseline system that would produce only citation-form accentuation, giving error rate reductions of I 1%-25%.