A Backdoor Attack Against LSTM-Based Text Classification Systems

Abstract

With the widespread use of deep learning system in many applications, the adversary has strong incentive to explore vulnerabilities of deep neural networks and manipulate them. Backdoor attacks against deep neural networks have been reported to be a new type of threat. In this attack, the adversary will inject backdoors into the model and then cause the misbehavior of the model through inputs including backdoor triggers. Existed research mainly focuses on backdoor attacks in image classification based on CNN, little attention has been paid to the backdoor attacks in RNN. In this paper, we implement a backdoor attack against LSTM-based text classification by data poisoning. After the backdoor is injected, the model will misclassify any text samples that contains a specific trigger sentence into the target category determined by the adversary. The backdoor attack is stealthy and the backdoor injected in the model has little impact on the performance of the model. We consider the backdoor attack in black-box setting, where the adversary has no knowledge of model structures or training algorithms except for a small amount of training data. We verify the attack through sentiment analysis experiment on the dataset of IMDB movie reviews. The experimental results indicate that our attack can achieve around 96% success rate with 1% poisoning rate.