• Corpus ID: 221819244

Finding Influential Instances for Distantly Supervised Relation Extraction

@article{Wang2020FindingII,
  title={Finding Influential Instances for Distantly Supervised Relation Extraction},
  author={Zifeng Wang and Rui Wen and Xi Chen and Shao-Lun Huang and Ningyu Zhang and Yefeng Zheng},
  journal={ArXiv},
  year={2020},
  volume={abs/2009.09841}
}
Distant supervision has been demonstrated to be highly beneficial to enhance relation extraction models, but it often suffers from high label noise. In this work, we propose a novel model-agnostic instance subsampling method for distantly supervised relation extraction, namely REIF, which bridges the gap of realizing influence subsampling in deep learning. It encompasses two key steps: first calculating instance-level influences that measure how much each training instance contributes to the… 

Figures and Tables from this paper

Learning from Noisy Labels for Entity-Centric Information Extraction
TLDR
A simple co-regularization framework for entity-centric information extraction is proposed, which consists of several neural models with identical structures but different parameter initialization, which prevents overfitting on noisy labels.
Logic-guided Semantic Representation Learning for Zero-Shot Relation Classification
TLDR
This paper proposes a novel logic-guided semantic representation learning model for zero-shot relation classification that builds connections between seen and unseen relations via implicit and explicit semantic representations with knowledge graph embeddings and logic rules.
Bridging Text and Knowledge with Multi-Prototype Embedding for Few-Shot Relational Triple Extraction
TLDR
This paper proposes a novel multi-prototype embedding network model to jointly extract the composition of relational triples, namely, entity pairs and corresponding relations and designs a hybrid prototypical learning mechanism that bridges text and knowledge concerning both entities and relations.
Relation Extraction as Open-book Examination: Retrieval-enhanced Prompt Tuning
TLDR
This work regards RE as an open-book examination and proposes a new semi-parametric paradigm of retrieval-enhanced prompt tuning for relation extraction, which not only infers relation through knowledge stored in the weights during training but also assists decision-making by unwinding and querying examples in the open- book datastore.
KnowPrompt: Knowledge-aware Prompt-tuning with Synergistic Optimization for Relation Extraction
TLDR
A Knowledge-aware Prompt-tuning approach with synergistic optimization (KnowPrompt) that injects latent knowledge contained in relation labels into prompt construction with learnable virtual type words and answer words.
PRGC: Potential Relation and Global Correspondence Based Joint Relational Triple Extraction
TLDR
A component to predict potential relations is designed, which constrains the following entity extraction to the predicted relation subset rather than all relations; then a relation-specific sequence tagging component is applied to handle the overlapping problem between subjects and objects; finally, a global correspondence component is designed to align the subject and object into a triple with low-complexity.
NC-DRE: Leveraging Non-entity Clue Information for Document-level Relation Extraction
TLDR
A novel graph-based model NC-DRE is proposed that introduces decoder-to-encoder attention mechanism to leverage non-entity clue information during the inference process of document-level RE and achieves state-of-the-art performance.
PAC-Bayes Information Bottleneck
TLDR
An MCMC-based algorithm is proposed to sample from the optimal weight posterior characterized by PIB, which fulfills the potential of IIW in enhancing NNs in practice and builds an IIW-based information bottleneck on the trade-off between accuracy and information complexity of NNs.
Joint Extraction of Triple Knowledge Based on Relation Priority
TLDR
A new model, called RFTE, based on relation priority is proposed, which decouples entity recognition under different types of relations and significantly reduces the impact of the large search space.
...
...

References

SHOWING 1-10 OF 36 REFERENCES
Neural Relation Extraction with Selective Attention over Instances
TLDR
A sentence-level attention-based model for relation extraction that employs convolutional neural networks to embed the semantics of sentences and dynamically reduce the weights of those noisy instances.
Distant Supervision for Relation Extraction via Piecewise Convolutional Neural Networks
TLDR
This paper proposes a novel model dubbed the Piecewise Convolutional Neural Networks (PCNNs) with multi-instance learning to address the problem of wrong label problem when using distant supervision for relation extraction and adopts convolutional architecture with piecewise max pooling to automatically learn relevant features.
Improving Distantly-Supervised Relation Extraction Through BERT-Based Label and Instance Embeddings
TLDR
RedSandT (Relation Extraction with Distant Supervision and Transformers), a novel distantly-supervised transformer-based RE method that manages to capture a wider set of relations through highly informative instance and label embeddings for RE by exploiting BERT’s pre-trained model, and the relationship between labels and entities, respectively is proposed.
Robust Distant Supervision Relation Extraction via Deep Reinforcement Learning
TLDR
A deep reinforcement learning strategy is explored to generate the false-positive indicator, where it is argued that incorrectly-labeled candidate sentences must be treated with a hard decision, rather than being dealt with soft attention weights.
Cross-relation Cross-bag Attention for Distantly-supervised Relation Extraction
TLDR
This paper proposes to conduct multi-instance learning with a novel Cross-relation Cross-bag Selective Attention (C2SA), which leads to noise-robust training for distant supervised relation extractor and employs the sentence-level selective attention to reduce the effect of noisy or mismatched sentences.
ARNOR: Attention Regularization based Noise Reduction for Distant Supervision Relation Classification
TLDR
This paper proposes ARNOR, a novel Attention Regularization based NOise Reduction framework for distant supervision relation classification that assumes that a trustable relation label should be explained by the neural attention model.
Denoising Distant Supervision for Relation Extraction via Instance-Level Adversarial Training
TLDR
This paper proposes a novel adversarial training mechanism over instances for relation extraction to alleviate the noise issue and develops a denoising method that can better discriminate those informative instances from noisy ones.
Fine-tuning Pre-Trained Transformer Language Models to Distantly Supervised Relation Extraction
TLDR
A pre-trained language model, the OpenAI Generative Pre-trained Transformer (GPT), is extended to the distantly supervised setting, and it is shown that it predicts a larger set of distinct relation types with high confidence.
A Soft-label Method for Noise-tolerant Distantly Supervised Relation Extraction
TLDR
This work introduces an entity-pair level denoise method which exploits semantic information from correctly labeled entity pairs to correct wrong labels dynamically during training, and proposes a joint score function which combines the relational scores based on the entity- Pair representation and the confidence of the hard label to obtain a new label.
...
...