Cooccurrence smoothing for stochastic language modeling

@inproceedings{citeulike:3015959, abstract = {Training corpora for stochastic language models are virtually always too small for maximum-likelihood estimation, so smoothing the models is of great importance. The authors derive the cooccurrence smoothing technique for stochastic language modeling and give experimental evidence for its validity. Using word-bigram language models, cooccurrence smoothing improved the test-set perplexity by 14\% on a German 100000-word text corpus and by 10\% on an English 1-million word corpus}, author = {Essen, U. and Steinbiss, V. }, booktitle = {Acoustics, Speech, and Signal Processing, 1992. ICASSP-92., 1992 IEEE International Conference on}, citeulike-article-id = {3015959}, doi = {http://dx.doi.org/10.1109/ICASSP.1992.225947}, journal = {Acoustics, Speech, and Signal Processing, 1992. ICASSP-92., 1992 IEEE International Conference on}, keywords = {toread-tribes}, pages = {161--164 vol.1}, posted-at = {2008-07-18 01:39:10}, priority = {2}, title = {Cooccurrence smoothing for stochastic language modeling}, url = {http://dx.doi.org/10.1109/ICASSP.1992.225947}, volume = {1}, year = {1992} }

TY - CONF ID - citeulike:3015959 L3 - citeulike-article-id:3015959 TI - Cooccurrence smoothing for stochastic language modeling T2 - Acoustics, Speech, and Signal Processing, 1992. ICASSP-92., 1992 IEEE International Conference on JF - Acoustics, Speech, and Signal Processing, 1992. ICASSP-92., 1992 IEEE International Conference on VL - 1 SP - 161 EP - 164 vol.1 N2 - Training corpora for stochastic language models are virtually always too small for maximum-likelihood estimation, so smoothing the models is of great importance. The authors derive the cooccurrence smoothing technique for stochastic language modeling and give experimental evidence for its validity. Using word-bigram language models, cooccurrence smoothing improved the test-set perplexity by 14% on a German 100000-word text corpus and by 10% on an English 1-million word corpus KW - toread-tribes AU - Essen, U AU - Steinbiss, V PY - 1992/// UR - http://dx.doi.org/10.1109/ICASSP.1992.225947 ER -