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.

An Entity Relation Extraction Method for Few-Shot Learning on the Food Health and Safety Domain

RETRACTED: an Entity Relationship Extraction Model Based on BERT-BLSTM-CRF for Food Safety Domain (retracted Article)

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

Few-Shot Relation Extraction with Hybrid Visual Evidence

Learning Discriminative Semantic and Multi-view Context for Domain Adaptive Few-Shot Relation Extraction.

Introducing high correlation and high quality instances for few-shot entity linking

MICK: A Meta-Learning Framework for Few-shot Relation Classification with Small Training Data

Mutually Guided Few-shot Learning for Relational Triple Extraction

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

Self-distillation framework for document-level relation extraction in low-resource environments

FewRel: A Large-Scale Supervised Few-Shot Relation Classification Dataset with State-of-the-Art Evaluation

A Few-Shot Learning Focused Survey on Recent Named Entity Recognition and Relation Classification Methods

Document-Level In-Context Few-Shot Relation Extraction via Pre-Trained Language Models

Semantic-consistent learning for one-shot joint entity and relation extraction

Towards Realistic Few-Shot Relation Extraction: A New Meta Dataset and Evaluation

Chem-FINESE: Validating Fine-Grained Few-shot Entity Extraction through Text Reconstruction

Few-shot named entity recognition with hybrid multi-prototype learning

Empowering Few-Shot Relation Extraction with The Integration of Traditional RE Methods and Large Language Models

SaberNet: Self-attention based effective relation network for few-shot learning

Effective Few-Shot Named Entity Linking by Meta-Learning

A Sequence Tagging based Framework for Few-Shot Relation Extraction