A Survey on Data Augmentation for Text Classification

Data augmentation, the artificial creation of training data for machine learning by transformations, is a widely studied research field across disciplines. While it useful increasing model's generalization capabilities, can also address many other challenges and problems, from overcoming limited amount to regularizing objective, limiting used protect privacy. Based on precise description goals applications augmentation taxonomy existing works, this survey concerned with methods textual classification aims at providing concise comprehensive overview researchers practitioners. Derived taxonomy, we divide more than 100 into 12 different groupings give state-of-the-art references expounding which are highly promising relating them each other. Finally, perspectives that may constitute building block future work provided.