DataAug4NLP

Why we included this project

Training NLP models on small datasets usually means wrestling with data augmentation, and this curated index collects the papers that matter for that problem. The entries are grouped by task, so whether you are doing text classification, translation, summarization, or question answering, you can jump straight to the techniques that fit instead of wading through a generic survey. Most entries link to the original paper and, when available, its code, which makes it easy to try a method quickly. The repository grew out of an ACL 2021 survey paper, which supplies the conceptual framing, while the list itself stays browsable and keeps the references organized. It is a research resource rather than a deployable tool, but for anyone weighing which augmentation strategy to test, it cuts down the literature hunting considerably.