A Review of Relationship Extraction Based on Deep Learning

Relation extraction is a key task in natural language processing. In recent years, deep learning techniques have been widely applied in relation extraction tasks. This paper systematically reviews relation extraction techniques based on deep learning, including the application of convolutional neural networks, recurrent neural networks and Transformer models. Firstly, it introduces the representative applications of these three models. Then, relation extraction methods based on the three deep learning models are comprehensively reviewed and compared. Finally, it discusses challenges faced by relation extraction tasks, including data sparsity, long-distance dependency, and outlooks new techniques like weakly supervised relation extraction. This paper provides a systematic overview of deep learning techniques for relation extraction, aiming to facilitate further research in this field.