Deconstructing Classifiers: Towards A Data Reconstruction Attack Against Text Classification Models

Adel Elmahdy, Ahmed Salem

Natural language processing (NLP) models have become increasingly popular in real-world applications, such as text classification. However, they are vulnerable to privacy attacks,…