Abstract. We report on the design of a system for correcting spelling errors resulting in non-existent words. The system aims at improving edition of medical reports. Unlike tradit...
This paper presents a comparative study of spelling errors that are corrected as you type, vs. those that remain uncorrected. First, we generate naturally occurring online error c...
We study the problem of correcting spelling mistakes in text using memory-based learning techniques and a very large database of token n-gram occurrences in web text as training d...