Document-Level Relation Extraction with Adaptive Thresholding and Localized Context Pooling

By Wenxuan Zhou, Kevin Huang, Tengyu Ma, Jing Huang. AAAI Conference on Artificial Intelligence, 2020

379 citations recorded by the source corpus.

Document-level relation extraction (RE) poses new challenges compared to its sentence-level counterpart. One document commonly contains multiple entity pairs, and one entity pair occurs multiple times in the document associated with multiple possible relations. In this paper, we propose two novel techniques, adaptive thresholding and localized context pooling, to solve the multi-label and multi-entity problems. The adaptive thresholding replaces the global threshold for multi-label classification in the prior work with a learnable entities-dependent threshold. The localized context pooling directly transfers attention from pre-trained language models to locate relevant context that is useful to decide the relation. We experiment on three document-level RE benchmark datasets: DocRED, a recently released large-scale RE dataset, and two datasets CDRand GDA in the biomedical domain. Our ATLOP (Adaptive Thresholding and Localized cOntext Pooling) model achieves an F1 score of 63.4, and also significantly outperforms existing models on both CDR and GDA. We have released our code at https://github.com/wzhouad/ATLOP.

Subfields: Artificial Intelligence, Molecular Biology

View Paper: https://www.semanticscholar.org/paper/3ac08d8c8210252cef12787539a81b0c8280ee3d

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Linhao Luo
2. Yang You
3. Xindong Wu
4. Chen Chen
5. Fuzhao Xue
6. Jiapu Wang
7. Shirui Pan
8. Yufei Wang

In Molecular Biology

1. Xindong Wu
2. Shirui Pan

Citations in the corpus

• Unifying Large Language Models and Knowledge Graphs: A Roadmap (2023, methodology): “ATLOP [232] focuses on the multi-label problems in DocRE, which could be handled with two techniques,”
• Sequence Parallelism: Long Sequence Training from System Perspective (2021, background): “Transformer-based language models [1, 2, 12] have achieved impressive performance on various natural language understanding and generation tasks ( e.g., Q&A [11, 23], relation extraction [21, 22, 27] and dialogue system [10]).”