信息抽取 | 72篇论文梳理：涉及NER、复杂关系、小样本、文档级、多模态、开放域抽取...

每天给你送来NLP技术干货！

---

文：JayJay

来自：高能AI

JayJay注：本文从EMNLP2021主会论文中收集，信息抽取方向共72篇，如有遗漏，请在留言区补充。

地址：https://2021.emnlp.org/papers#main-long

一、实体抽取（共19篇）

实体抽取主要涉及小样本NER、降噪NER、跨领域NER、细粒度实体分类、实体集拓展、NER鲁棒性、实体链接和消歧等

小样本&低资源&降噪&跨领域NER

1. Few-Shot Named Entity Recognition: An Empirical Baseline Study

2. Learning from Noisy Labels for Entity-Centric Information Extraction

3. Data Augmentation for Cross-Domain Named Entity Recognition

4. Unsupervised Paraphrasing Consistency Training for Low Resource Named Entity Recognition

细粒度实体分类&实体集拓展

1. Fine-grained Entity Typing without Knowledge

2. An Empirical Study on Multiple Information Sources for Zero-Shot Fine-Grained Entity Typing

3. Fine-grained Entity Typing via Label Reasoning

4. Progressive Adversarial Learning for Bootstrapping: A Case Study on Entity Set Expansion

NER鲁棒性

1. RockNER: A Simple Method to Create Adversarial Examples for Evaluating the Robustness of Named Entity Recognition Models

2. SeqAttack: On Adversarial Attacks for Named Entity Recognition

实体链接&消歧

1. Low-Rank Subspaces for Unsupervised Entity Linking

2. ActiveEA: Active Learning for Neural Entity Alignment

3. Highly Parallel Autoregressive Entity Linking with Discriminative Correction

4. End-to-End Entity Resolution and Question Answering Using Differentiable Knowledge

5. From Alignment to Assignment: Frustratingly Simple Unsupervised Entity Alignment

6. Robustness Evaluation of Entity Disambiguation Using Prior Probes: the Case of Entity Overshadowing

其他

1. TEBNER: Domain Specific Named Entity Recognition with Type Expanded Boundary-aware Network

2. Document-level Entity-based Extraction as Template Generation

3. Injecting Entity Types into Entity-Guided Text Generation

4. Enhanced Language Representation with Label Knowledge for Span Extraction

5. ChemNER: Fine-Grained Chemistry Named Entity Recognition with Ontology-Guided Distant Supervision

二、关系抽取（共19篇）

关系抽取主要涉及等远程监督抽取、联合抽取、文档级别抽取、小样本抽取等。

远程监督关系抽取

1. Distantly Supervised Relation Extraction using Multi-Layer Revision Network and Confidence-based Multi-Instance Learning

2. Knowing False Negatives: An Adversarial Training Method for Distantly Supervised Relation Extraction

实体关系联合抽取

1. TDEER: An Efficient Translating Decoding Schema for Joint Extraction of Entities and Relations

2. A Partition Filter Network for Joint Entity and Relation Extraction

3. A Novel Global Feature-Oriented Relational Triple Extraction Model based on Table Filling

4. Synchronous Dual Network with Cross-Type Attention for Joint Entity and Relation Extraction

文档级别关系抽取

1. Modular Self-Supervision for Document-Level Relation Extraction

2. CodRED: A Cross-Document Relation Extraction Dataset for Acquiring Knowledge in the Wild

3. Learning Logic Rules for Document-Level Relation Extraction

小样本&低资源关系抽取

1. Exploring Task Difficulty for Few-Shot Relation Extraction

2. MapRE: An Effective Semantic Mapping Approach for Low-resource Relation Extraction

3. Label Verbalization and Entailment for Effective Zero and Few-Shot Relation Extraction

4. Unsupervised Relation Extraction: A Variational Autoencoder Approach

5. Gradient Imitation Reinforcement Learning for Low Resource Relation Extraction

6. Towards Realistic Few-Shot Relation Extraction

对话关系抽取

1. Speaker-Oriented Latent Structures for Dialogue-Based Relation Extraction

其他

1. Relation Extraction with Word Graphs from N-grams

2. Separating Retention from Extraction in the Evaluation of End-to-end Relation Extraction

3. Incorporating medical knowledge in BERT for clinical relation extraction

三、事件抽取（共7篇）

1. Machine Reading Comprehension as Data Augmentation: A Case Study on Implicit Event Argument Extraction

2. Modeling Document-Level Context for Event Detection via Important Context Selection

3. Extracting Event Temporal Relations via Hyperbolic Geometry

4. Utilizing Relative Event Time to Enhance Event-Event Temporal Relation Extraction

5. Corpus-based Open-Domain Event Type Induction

6. Honey or Poison? Solving the Trigger Curse in Few-shot Event Detection via Causal Intervention

7. Learning Prototype Representations Across Few-Shot Tasks for Event Detection

四、信息抽取（综合，共9篇）

综合类的信息抽取主要涉及开放抽取、多模抽取、跨语言抽取、长尾抽取。

开放信息抽取

1. Maximal Clique Based Non-Autoregressive Open Information Extraction

2. A Relation-Oriented Clustering Method for Open Relation Extraction

3. Zero-Shot Information Extraction as a Unified Text-to-Triple Translation

多模信息抽取

1. Entity Relation Extraction as Dependency Parsing in Visually Rich Documents

2. Cost-effective End-to-end Information Extraction for Semi-structured Document Images

跨语言信息抽取

1. Everything Is All It Takes: A Multipronged Strategy for Zero-Shot Cross-Lingual Information Extraction

2. Crosslingual Transfer Learning for Relation and Event Extraction via Word Category and Class Alignments

3. XLEnt: Mining a Large Cross-lingual Entity Dataset with Lexical-Semantic-Phonetic Word Alignment

长尾信息抽取

1. Uncovering Main Causalities for Long-tailed Information Extraction

五、关键词&短语抽取（共4篇）

1. Unsupervised Keyphrase Extraction by Jointly Modeling Local and Global Context

2. Back to the Basics: A Quantitative Analysis of Statistical and Graph-Based Term Weighting Schemes for Keyword Extraction

3. AttentionRank: Unsupervised Keyphrase Extraction using Self and Cross Attentions

4. Importance Estimation from Multiple Perspectives for Keyphrase Extraction

六、属性抽取&观点抽取（共4篇）

1. DILBERT: Customized Pre-Training for Domain Adaptation with Category Shift, with an Application to Aspect Extraction

2. Progressive Self-Training with Discriminator for Aspect Term Extraction

3. PASTE: A Tagging-Free Decoding Framework Using Pointer Networks for Aspect Sentiment Triplet Extraction

4. An Empirical Study on Leveraging Position Embeddings for Target-oriented Opinion Words Extraction

---

投稿或交流学习，备注：昵称-学校（公司）-方向，进入DL&NLP交流群。

方向有很多：机器学习、深度学习，python，情感分析、意见挖掘、句法分析、机器翻译、人机对话、知识图谱、语音识别等。

记得备注呦

整理不易，还望给个在看！