Paper: A Method for Unsupervised Broad-Coverage Lexical Error Detection and Correction

ACL ID W09-2108
Title A Method for Unsupervised Broad-Coverage Lexical Error Detection and Correction
Venue Innovative Use of NLP for Building Educational Applications
Session
Year 2009
Authors

We describe and motivate an unsupervised lexical error detection and correction algo- rithm and its application in a tool called Lex- bar appearing as a query box on the Web browser toolbar or as a search engine inter- face. Lexbar accepts as user input candidate strings of English to be checked for accept- ability and, where errors are detected, offers corrections. We introduce the notion of hy- brid n-gram and extract these from BNC as the knowledgebase against which to compare user input. An extended notion of edit dis- tance is used to identify most likely candi- dates for correcting detected errors. Results are illustrated with four types of errors.