Paper: Improved Iterative Correction for Distant Spelling Errors

ACL ID P14-2028
Title Improved Iterative Correction for Distant Spelling Errors
Venue Annual Meeting of the Association of Computational Linguistics
Session Main Conference
Year 2014
Authors

Noisy channel models, widely used in modern spellers, cope with typical mis- spellings, but do not work well with infre- quent and difficult spelling errors. In this paper, we have improved the noisy chan- nel approach by iterative stochastic search for the best correction. The proposed al- gorithm allowed us to avoid local minima problem and improve the F 1 measure by 6.6% on distant spelling errors.