A Decoupling and Aggregating Framework for Joint Extraction of Entities and Relations

Named Entity Recognition and Relation Extraction are two crucial and challenging subtasks in the field of Information Extraction. Despite the successes achieved by the traditional approaches, fundamental research questions remain open. First, most recent studies use parameter sharing for a single subtask or shared features for both two subtasks, ignoring their semantic differences. Second, information interaction mainly focuses on the two subtasks, leaving the fine-grained informtion interaction among the subtask-specific features of encoding subjects, relations, and objects unexplored. Motivated by the aforementioned limitations, we propose a novel model to jointly extract entities and relations. The main novelties are as follows: (1) We propose to decouple the feature encoding process into three parts, namely encoding subjects, encoding objects, and encoding relations. Thanks to this, we are able to use fine-grained subtask-specific features. The experimental results demonstrate that our model outperforms several previous state-of-the-art models. Extensive additional experiments further confirm the effectiveness of our model. A Decoupling and Aggregating Framework for Joint Extraction of Entities and Relations Introduction Named Entity Recognition (NER) and Relation Extraction (RE), as two essential subtasks in information extraction, aim to extract entities and relations from semi-structured and unstructured texts. They are used in many downstream applications in different domains, such as knowledge graph construction [38, 39], Question-Answering [36, 37], and knowledge graph-based recommendation system [40, 41]. Most traditional models and some methods used in specialized areas [9,35,43,46] construct separate models for NER and RE to extract entities and relations in a pipelined manner. This type of method suffers from error propagation and unilateral information interaction.