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.

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

Joint extraction of entities and relations using multi-label tagging and relational alignment

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

Joint extraction model of entity relations based on decomposition strategy

Joint Extraction of Entities and Relations Based on a Novel Tagging Scheme

Joint Extraction of Entities and Relations Based on a Novel Decomposition Strategy

A neural network-based joint learning approach for biomedical entity and relation extraction from biomedical literature

Joint Extraction of Entities and Overlapping Relations using Position-Attentive Sequence Labeling

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

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

Joint learning of named entity recognition and relation extraction

Joint Model of Entity Recognition and Relation Extraction with Self-attention Mechanism

Attention As Relation: Learning Supervised Multi-head Self-Attention for Relation Extraction

Extracting Entities and Relations by an End-to-End Neural Model with Multi-Layer Encoder.

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

Joint Extraction of Entity and Relation with Information Redundancy Elimination

Joint extraction of entities and relations based on character graph convolutional network and Multi-Head Self-Attention Mechanism

Joint Extraction of Entities and Relations Based on Multi-feature Fusion

OneRel:Joint Entity and Relation Extraction with One Module in One Step

Joint Extraction of Entities and Relations Based on a Novel Graph Scheme

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