Grammar Error Correction using Seq2Seq

Abstract: Grammatical Error Correction (GEC) in English language is a challenging topic among the emerging works. GEC is a process of converting the erroneous sentences to a corrected sentence by using Sequence2Sequence (Seq2Seq) method. Usually the system focused on correcting the grammars based on the 20 rules in English language and it includes punctuation, grammatical and word choice errors. Deep learning method is used to work behind the system. Long Short-Term Memory (LSTM) Encoder - Decoder model is used in the conversion of incorrect sentence to a grammatically corrected sentence. This is a supervised learning system which includes incorrect and corrected sentences in the GEC dataset and thus gives better results.

Keywords: GEC, Deep Learning, seq2seq, LSTM.