A Statistical Approach to Grammatical Error Correction

Abstract

In this Ph.D. thesis, we pursue a statistical approach to grammatical error correction based on machine learning methods that advance the field in several directions. First, the NUS Corpus of Learner English, a one-million-word corpus of annotated learner English was created as part of this thesis. Based on this data set, we present a novel method that allows for training statistical classifiers with both learner and non-learner data and successfully apply it to article and preposition errors. Next, we focus on lexical choice errors and show that they are often caused by words with similar translations in the native language of the writer. We show that paraphrases induced through the native language of the writer can be exploited to automatically correct such errors. Fourth, we present a pipeline architecture that combines individual correction modules into an end-to-end correction system with state-of-the-art results. Finally, we present a novel beam-search decoder for grammatical error correction that can correct sentences which contain multiple and interacting errors. The decoder further improves over the state-of-the-art pipeline architecture, setting a new state of the art in grammatical error correction.