Don't be a Fool: Pooling Strategies in Offensive Language Detection from User-Intended Adversarial Attacks

Open in new window