Reinforcement Learning for Relation Classification from Noisy Data
arXiv:1808.08013
Abstract
Existing relation classification methods that rely on distant supervision assume that a bag of sentences mentioning an entity pair are all describing a relation for the entity pair. Such methods, performing classification at the bag level, cannot identify the mapping between a relation and a sentence, and largely suffers from the noisy labeling problem. In this paper, we propose a novel model for relation classification at the sentence level from noisy data. The model has two modules: an instance selector and a relation classifier. The instance selector chooses high-quality sentences with reinforcement learning and feeds the selected sentences into the relation classifier, and the relation classifier makes sentence level prediction and provides rewards to the instance selector. The two modules are trained jointly to optimize the instance selection and relation classification processes. Experiment results show that our model can deal with the noise of data effectively and obtains better performance for relation classification at the sentence level.
Cited by in corpus (58)
- A Survey on Knowledge Graphs: Representation, Acquisition and Applications
- More Data, More Relations, More Context and More Openness: A Review and Outlook for Relation Extraction
- Prototypical Representation Learning for Relation Extraction
- RLAS-BIABC: A Reinforcement Learning-Based Answer Selection Using the BERT Model Boosted by an Improved ABC Algorithm
- Learning to Selectively Transfer: Reinforced Transfer Learning for Deep Text Matching
- Improving Neural Relation Extraction with Implicit Mutual Relations
- Deep Reinforcement Learning for Imbalanced Classification
- Reinforced Neighborhood Selection Guided Multi-Relational Graph Neural Networks
- Leveraging 2-hop Distant Supervision from Table Entity Pairs for Relation Extraction
- Distant Supervision Relation Extraction with Intra-Bag and Inter-Bag Attentions
- Document-level Relation Extraction as Semantic Segmentation
- Double Graph Based Reasoning for Document-level Relation Extraction
- Weak Supervision for Fake News Detection via Reinforcement Learning
- Why People Skip Music? On Predicting Music Skips using Deep Reinforcement Learning
- OpenNRE: An Open and Extensible Toolkit for Neural Relation Extraction
- A Question-answering Based Framework for Relation Extraction Validation
- Long-tail Relation Extraction via Knowledge Graph Embeddings and Graph Convolution Networks
- A Hierarchical Reinforced Sequence Operation Method for Unsupervised Text Style Transfer
- Learning Relation Prototype from Unlabeled Texts for Long-tail Relation Extraction
- Domain-Level Explainability -- A Challenge for Creating Trust in Superhuman AI Strategies
- Doc2EDAG: An End-to-End Document-level Framework for Chinese Financial Event Extraction
- InSRL: A Multi-view Learning Framework Fusing Multiple Information Sources for Distantly-supervised Relation Extraction
- DIAG-NRE: A Neural Pattern Diagnosis Framework for Distantly Supervised Neural Relation Extraction
- Discriminative Reasoning for Document-level Relation Extraction
- Are Noisy Sentences Useless for Distant Supervised Relation Extraction?
- On Robustness and Bias Analysis of BERT-based Relation Extraction
- Search from History and Reason for Future: Two-stage Reasoning on Temporal Knowledge Graphs
- Named Entity Recognition Only from Word Embeddings
- Reinforcement Learning-based N-ary Cross-Sentence Relation Extraction
- Learning Classifiers on Positive and Unlabeled Data with Policy Gradient
- Looking Beyond Label Noise: Shifted Label Distribution Matters in Distantly Supervised Relation Extraction
- Reinforcement Learning based Collective Entity Alignment with Adaptive Features
- Neural relation extraction: a survey
- IPRE: a Dataset for Inter-Personal Relationship Extraction
- Improving Distant Supervised Relation Extraction by Dynamic Neural Network
- Denoising Relation Extraction from Document-level Distant Supervision
- Towards Accurate and Consistent Evaluation: A Dataset for Distantly-Supervised Relation Extraction
- Integrating Relation Constraints with Neural Relation Extractors
- Uncovering Relations for Marketing Knowledge Representation
- Mention-centered Graph Neural Network for Document-level Relation Extraction
- Manual Evaluation Matters: Reviewing Test Protocols of Distantly Supervised Relation Extraction
- Learning to Decouple Relations: Few-Shot Relation Classification with Entity-Guided Attention and Confusion-Aware Training
- CCKS 2019 Shared Task on Inter-Personal Relationship Extraction
- Active Testing: An Unbiased Evaluation Method for Distantly Supervised Relation Extraction
- Towards Robust Knowledge Graph Embedding via Multi-task Reinforcement Learning
- Revisiting the Negative Data of Distantly Supervised Relation Extraction
- SENT: Sentence-level Distant Relation Extraction via Negative Training
- Uncover the Ground-Truth Relations in Distant Supervision: A Neural Expectation-Maximization Framework
- Learning to Selectively Learn for Weakly-supervised Paraphrase Generation
- Distantly Supervised Relation Extraction in Federated Settings
- Personalizing Search Results Using Hierarchical RNN with Query-aware Attention
- From Bag of Sentences to Document: Distantly Supervised Relation Extraction via Machine Reading Comprehension
- Deep Ranking Based Cost-sensitive Multi-label Learning for Distant Supervision Relation Extraction
- H-FND: Hierarchical False-Negative Denoising for Distant Supervision Relation Extraction
- Regularized Attentive Capsule Network for Overlapped Relation Extraction
- A Sample-Based Training Method for Distantly Supervised Relation Extraction with Pre-Trained Transformers
- Relabel the Noise: Joint Extraction of Entities and Relations via Cooperative Multiagents
- Learning to Expand: Reinforced Pseudo-relevance Feedback Selection for Information-seeking Conversations