The structured language model (SLM) of was one of the first to successfully integrate syntactic structure into language models. We extend the SLM framework in two new directions. First, we propose a new syntactic hierarchical interpolation that improves over previous approaches. Second, we develop a general information-theoretic algorithm for pruning the underlying Jelinek-Mercer interpolated LM used in , which substantially reduces the size of the LM, enabling us to train on large data. When combined with hill-climbing the SLM is an accurate model, space-efficient and fast for rescoring large speech lattices. Experimental results on broadcast news demonstrate that the SLM outperforms a large 4-gram LM.
展开▼
