I have to do deep learning based natural language correction for a project. Given BERT's versatility, can it be used for this task. It uses a MLM based approach which may work well for a task where I have to predict the correct spelling of a word. Currently, I am using a LSTM based encoder-decoder approach. Any suggestions regarding using BERT or MLM approach would be useful.
Most helpful comment
https://github.com/huseinzol05/NLP-Models-Tensorflow/tree/master/spelling-correction