Abstract:Extracting entities and relations from text is a significant task of information extraction. Existing extraction models often straightforwardly produce their confident prediction results without any reconsideration or double-checking, resulting in avoidable mistakes and sub-optimal performance. In this paper, we propose a novel coarse-to-fine extraction framework, which first extracts high-potential relations as well as entities via knowledge distillation, and then rechecks the predictions via handcrafted natural language inference (NLI) task in a fine-grained manner. Specifically, based on the knowledge distillation mechanism, we train multiple teacher models iteratively through an adaptive loss function for making one teacher concentrate more on the data that others are incompetent for. Then, these complementary teacher models are utilized to provide valuable soft-label information for training a considerate student model, enabling it to generate reliable preliminary predictions. Further, these generated potential relations and entities are formulated as hypotheses, together with the original sentences as premises, serving as the input for an NLI model. Considering the linguistic diversity of relational expression, we automatically generate various semantic templates for hypotheses through an <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathcal{N}$</tex> -gram mining strategy. Moreover, due to the existence of multi-fact sentences, a relation-guided Gaussian attention is designed to reduce the gap between the single-relation hypothesis and the multi-relation premise. To implement efficient training, we also develop several ways to generate high-quality negative samples, which help the NLI model learn to identify errors. Experimental results show that the proposed method is effective and outperforms other strong baselines on public benchmarks.

A joint model for entity and relation extraction based on BERT

Bridge Relation Extraction: New Chinese Dataset and Model

RETRACTED: an Entity Relationship Extraction Model Based on BERT-BLSTM-CRF for Food Safety Domain (retracted Article)

Enriching Pre-trained Language Model with Entity Information for Relation Classification

Construction and Application of Text Entity Relation Joint Extraction Model Based on Multi-Head Attention Neural Network

A Coarse-to-Fine Framework for Entity-Relation Joint Extraction.

Research on the Chinese Named-Entity–Relation-Extraction Method for Crop Diseases Based on BERT

Entity relation joint extraction model combining pointer network and attention mechanism based on relative position embedding

BERT-BIGRU-CRF: A Novel Entity Relationship Extraction Model

Joint Extraction of Triple Knowledge Based on Relation Priority.

Siamese BERT Model with Adversarial Training for Relation Classification.

Joint extraction of entity relations from geological reports based on a novel relation graph convolutional network

Joint extraction of entities and relations via an entity correlated attention neural model

Joint entity and relation extraction model based on directed-relation GAT oriented to Chinese patent texts

Entity relation joint extraction method for manufacturing industry knowledge data based on improved BERT algorithm

Joint Extraction model of Chinese Recipe Entities and Relations Based on Adversarial Training and Cascaded Binary Tagging

Joint extraction model of entity relations based on decomposition strategy

A Method of Relation Extraction Using Pre-training Models

An Entity-Relation Joint Extraction Method Based on Two Independent Sub-Modules From Unstructured Text

A neural joint model for entity and relation extraction from biomedical text

An Entity Relationship Extraction Model Based on BERT-BLSTM-CRF for Food Safety Domain