Hey! There, I am pursuing MTech and as part of the academic curriculum, I have to publish a research paper in Scopus Indexed Journal.
So the Problem is like my guide suggested writing a paper on Data Augmentation in NLP where we can use an ensemble approach using 3 - 4 text data augmentation techniques and later using some similarity filter we can choose the best augmentation technique for the dataset, and also to use different types of datasets too, for comparison.
I couldn't find how to start that's why I landed here and found it helpful in this. I've attached a Visual concept of that.