SSCAE: A Novel Semantic, Syntactic, and Context-Aware Natural Language Adversarial Example Generator

Abstract

Training a machine learning model with adversarial examples (AEs) improves its robustness against adversarial attacks. Hence, it is crucial to develop effective generative models to produce high-quality AEs. Developing such models has been much slower in natural language processing (NLP). The current state-of-the-art in NLP generates AEs that are somehow human detectable and/or include semantic and linguistic defects. This paper introduces a novel, practical, and efficient adversarial attack model called SSCAE for Semantic, Syntactic, and Context-aware natural language Adversarial Examples generator. SSCAE generates humanly imperceptible context-aware AEs thatpreserve semantic consistency and source language’s syntactical and grammatical requirements. The effectiveness and superiority ofthe proposed SSCAE model are illustrated over eleven comparative experiments, extensive ablation studies, and human evaluations.

Tasks

Reproductions