Wen Wang and D. Vergyri, “The Use of Word N-Grams and Parts of Speech for Hierarchical Cluster Language Modeling,” 2006 IEEE International Conference on Acoustics Speech and Signal Processing Proceedings, 2006, pp. I-I, doi: 10.1109/ICASSP.2006.1660206.
Abstract
We present extensions to the work of backoff hierarchical class n-gram language modeling of Zitouni et al. (2003) by studying t he efficacy of exploring the use of parts of speech (POS) information in hierarchical word clustering. We propose two approaches. One is to u se POS n-gram contextual distributions of a target word for clustering. The other is to generate a class tree for each group of words sharing the same POS. The resulting class tree and a set of class trees, from the two approaches, respectively, are then employed in the hierarchica l cluster language modeling. We evaluate the two approaches on SRI Arabic conversational telephone speech recognition system and show that th e approach of building a set of POS-specific class trees achieves a 3 pct. relative improvement on perplexity compared to the model of Zitouni et al. and a 8 pct. relative improvement on perplexity over the baseline standard word n-grams. When used for N-best rescoring, our approach also outperforms the model of Zitouni et al. and the baseline and achieves significant word error rate (WER) reductions.
