TPLinker: Single-stage Joint Extraction of Entities and Relations Through Token Pair Linking

By Yucheng Wang, Yu Bowen, Yueyang Zhang, Tingwen Liu, Hongsong Zhu, Limin Sun. International Conference on Computational Linguistics, 2020

447 citations recorded by the source corpus.

Extracting entities and relations from unstructured text has attracted increasing attention in recent years but remains challenging, due to the intrinsic difficulty in identifying overlapping relations with shared entities. Prior works show that joint learning can result in a noticeable performance gain. However, they usually involve sequential interrelated steps and suffer from the problem of exposure bias. At training time, they predict with the ground truth conditions while at inference it has to make extraction from scratch. This discrepancy leads to error accumulation. To mitigate the issue, we propose in this paper a one-stage joint extraction model, namely, TPLinker, which is capable of discovering overlapping relations sharing one or both entities while being immune from the exposure bias. TPLinker formulates joint extraction as a token pair linking problem and introduces a novel handshaking tagging scheme that aligns the boundary tokens of entity pairs under each relation type. Experiment results show that TPLinker performs significantly better on overlapping and multiple relation extraction, and achieves state-of-the-art performance on two public datasets.

Subfields: Artificial Intelligence

View Paper: https://www.semanticscholar.org/paper/4d4a00970f2751c3cc8a036c9ac1cb68eacb45d3

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. Tao Qin
2. Tie-Yan Liu
3. Yingce Xia
4. Philip S. Yu
5. Shirui Pan
6. Renqian Luo
7. Xu Tan
8. Erik Cambria

Citations in the corpus

• A Comprehensive Survey on Automatic Knowledge Graph Construction (2023, background): “[223] developed a hand-shaking scheme to alleviate exposure bias within overlapping entities.”
• A Survey on Knowledge Graphs: Representation, Acquisition, and Applications (2020, background): “Recently, Wang et al. [172] proposed a one-stage joint extraction framework by transforming joint entity and relation extraction into a token pair linking task to mitigate error propagation and exposure bias.”
• BioGPT: Generative Pre-trained Transformer for Biomedical Text Generation and Mining (2022, methodology): “Table filling methods formulate the task as a table constituted by the Cartesian product of itself and predicts the relations between the token pairs [39, 40, 41].”