<![CDATA[Writing errors in Indonesian are often found in various writings made in educational, government and mass media environments. The most dominant error is in spelling. This research proposes a Grammatical Error Correction (GEC) for Indonesian using the Neural Machine Translation (NMT) method, namely seq2seq, which is popularly used for English and has achieved the best performance approaching human capabilities. The model developed is made into a web-based service that is easy for users to access. The datasets used in this experiment are artificial datasets sourced from several studies regarding error analysis in Indonesian. The research results show that with the help of currently available open-source tools such as OpenNMT-py, it is possible to simplify the training process of NMT-based GEC models. Unfortunately, the small number of datasets leads to poor predictions for random sentences.]]>
Copyrights © 2023
