Abstract:Relation extraction has become a crucial step for the automatic construction of Knowledge Graph (KG). Recently, researchers leverage Sequence-to-Sequence (Seq2Seq) models for Joint Entity and Relation Extraction (JERE). Nevertheless, traditional decoding methods entail the generation of the target sequence incrementally from left to right by the decoder, without the ability to revise earlier predictions when errors occur. This limitation becomes evident when decoding errors manifest prior to the current decoding step. Furthermore, the interrelations among triplets originating from the same sentence exhibit a robust correlation, which has been overlooked. In this paper, we propose B idirectional D ecoding with Co -graph re presentation (BDCore) to address the issues mentioned above. Specifically, we first introduce a backward decoder to decode the target sequence in a reverse order. Then, the forward decoder introduces two attention mechanisms to simultaneously considering the hidden states of the encoder and the backward decoder. Thus, the backward decoding information helps to alleviate the negative impact of the forward decoding errors. Besides, we construct a relation co-occurrence graph (Co-graph) and exploit Graph Convolutional Network (GCN) to capture the relation correlation. The extensive experiments demonstrate the benefits of the proposed bidirectional decoding and co-graph representation for relation extraction. Compared to the previous methods, our approach significantly outperforms the baselines on the NYT benchmark.

Graph-Based Bootstrapping for Coreference Resolution

Improving Coreference Resolution by Leveraging Entity-Centric Features with Graph Neural Networks and Second-order Inference

Semi-supervised Bootstrapping approach for Named Entity Recognition

Graph Mutual Reinforcement Based Bootstrapping

A New Graph Clustering Algorithm for Chinese Noun Phrase Coreference Resolution

Bootstrapping Object Coreferencing on the Semantic Web

Unsupervised Entity Resolution with Blocking and Graph Algorithms

A Graph-Theoretic Fusion Framework for Unsupervised Entity Resolution

CorefQA: Coreference Resolution As Query-based Span Prediction.

Link Type Based Pre-Cluster Pair Model for Coreference Resolution.

A Bootstrapping Approach to Entity Linkage on the Semantic Web.

Incorporating Singletons and Mention-based Features in Coreference Resolution via Multi-task Learning for Better Generalization

Learning Semantic Lexicons Using Graph Mutual Reinforcement Based Bootstrapping

Coreference Resolution as Query-based Span Prediction

Coreference Resolution with Supervised Correlation Clustering

Integrating K+ Entities into Coreference Resolution on Biomedical Texts

Filling in the Gaps: Efficient Event Coreference Resolution using Graph Autoencoder Networks

ETS: An Error Tolerable System for Coreference Resolution.

BDCore: Bidirectional Decoding with Co-graph Representation for Joint Entity and Relation Extraction

Coreference Resolution Based on High-Dimensional Multi-Scale Information

Coreference Resolution in Research Papers from Multiple Domains