Improved relation classification by deep recurrent neural networks with data augmentation

Nowadays, neural networks play an important role in the task of relation classification. By designing different neural architectures, researchers have improved the performance to a large extent, compared with traditional methods. However, existing neural networks for relation classification are usually of shallow architectures (e.g., one-layer convolution neural networks or recurrent networks). They may fail to explore the potential representation space in different abstraction levels. In this paper, we propose deep recurrent neural networks (DRNNs) to tackle this challenge. Further, we propose a data augmentation method by leveraging the directionality of relations. We evaluate our DRNNs on the SemEval-2010 Task 8, and achieve an F1score of 85.81%, outperforming state-of-theart recorded results.