An Unsupervised Method for Detecting Grammatical Errors

We present an unsupervised method for detecting grammatical errors by inferring negative evidence from edited textual corpora. The system was developed and tested using essay-length responses to prompts on the Test of English as a Foreign Language (TOEFL). The errorrecognition system, ALEK, performs with about 80% precision and 20% recall.

• Type: Conference
• Year: 2000
• Where: ANLP
• Authors: Martin Chodorow, Claudia Leacock
• Keywords: ANLP 2000, Essay-length Responses, Grammatical Errors, Unsupervised Method