This paper applies Chinese subword representations, namely character and syllable n-grams, into the TextTiling-based automatic story segmentation of Chinese broadcast news. We show the robustness of Chinese subwords against speech recognition errors, out-of-vocabulary (OOV) words and versatility in word segmentation in lexical matching on errorful Chinese speech recognition transcripts. We prop...

This paper deals with the choice of suitable subword units (SWU) for a HMM based speech recognition system. Using demisyllables (including phonemes) as base units, an inventory of domain-specific larger sized subword units, so-called macro-demisyllables (MDS), is created. A quality measure for the automatic decomposition of all single words into subword units is presented which takes into accou...

Hidden Markov model speech recognition systems typically use Gaussian mixture models to estimate the distributions of decorrelated acoustic feature vectors that correspond to individual subword units. By contrast, hybrid connectionist-HMM systems use discriminatively-trained neural networks to estimate the probability distribution among subword units given the acoustic observations. In this wor...

In this work, we introduce and develop a novel framework, angie, for modelling subword lexical phenomena in speech recognition. Our framework provides a exible and powerful mechanism for capturing morphology, syllabi cation, phonology, and other subword e ects in a hierarchical manner which maximizes sharing of subword structures. Angie models the subword structure within a context-free grammar...

This paper investigates the feasibility of using subword unit representations for spoken document retrieval as an alternative to using words generated by either keyword spotting or word recognition. Our investigation is motivated by the observation that word-based retrieval approaches face the problem of either having to know the keywords to search for a priori, or requiring a very large recogn...

This paper deals with the choice of suitable subword units (SWU) for a HMM based speech recognition system. Using demisyllables (including phonemes) as base units, an inventory of domain-specific larger sized subword units, so-called macro-demisyllables (MDS), is created. A quality measure for the automatic decomposition of all single words into subword units is presented which takes into accou...

Kitaev, Liese, Remmel, and Sagan recently defined generalized factor order on words comprised of letters from a partially ordered set (P,≤P) by setting u ≤P w if there is a subword v of w of the same length as u such that the i-th character of v is greater than or equal to the i-th character of u for all i. This subword v is called an embedding of u into w. Generalized factor order is related t...

Vector representations of words that capture semantic and syntactic information accurately is critical for the performance of models that use these vectors as inputs. Algorithms that only use the surrounding context at the word level ignore the subword level relationships which carry important meaning especially for languages that are highly inflected such as Korean. In this paper we compare th...