Few-Shot Joint Extraction of Entity-Relation Triples in the Domain of Ancient Chinese Medical Texts

Chaofan Wang; Jun Pan

doi:10.63313/EBM.9056

Authors

Chaofan Wang School of Science, Zhejiang University of Science and Technology, HangZhou, 310023, P.R. China. Author
Jun Pan School of Science, Zhejiang University of Science and Technology, HangZhou, 310023, P.R. China. Author

DOI:

https://doi.org/10.63313/EBM.9056

Keywords:

Joint entity and relation extraction, Few-shot relation extraction, Prototype network, Ancient Chinese medical texts

Abstract

Entity–relation extraction aims to identify structured triples from unstructured text. However, most existing models are evaluated on large, densely annotated corpora; when applied to vertical domains with scarce labeled data, fully super-vised models tend to generalize poorly. Existing few-shot learning approaches often simplify the triple extraction task into relation classification with known entities, or overlook the common real-world scenario where a sentence con-tains multiple triples. To address these challenges, this paper focuses on ancient Chinese medicine texts and proposes a multi-granularity fused prototype net-work. The model fully integrates semantic information between relations and entities during prototype construction to enhance the discriminability among different prototypes. A pointer network is further employed to enable the ex-traction of multiple triples from a single text. Experiments on a self-constructed dataset of ancient Chinese medicine texts demonstrate the effectiveness of the proposed approach.

References

[1] Wei, Z., Su, J., Wang, Y., et al. (2020). A Novel Cascade Binary Tagging Framework for Rela-tional Triple Extraction. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics (pp. 1476 - 1488).

[2] Zheng, H., Wen, R., Chen, X., et al. (2021). PRGC: Potential Relation and Global Correspond-ence Based Joint Relational Triple Extraction. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers) (pp. 6225 - 6235).

[3] Yuan, Y., Zhou, X., Pan, S., et al. (2021). A relation-specific attention network for joint entity and relation extraction. In International joint conference on artificial intelligence. Interna-tional Joint Conference on Artificial Intelligence.

[4] Wang, Y., Yu, B., Zhang, Y., et al. (2020). TPLinker: Single-stage Joint Extraction of Entities and Relations Through Token Pair Linking. In Proceedings of the 28th International Con-ference on Computational Linguistics (pp. 1572 - 1582).

[5] Shang, Y. M., Huang, H., & Mao, X. (2022). Onerel: Joint entity and relation extraction with one module in one step. In Proceedings of the AAAI conference on artificial intelli-gence (Vol. 36, No. 10, pp. 11285 - 11293).

[6] Han, X., Zhu, H., Yu, P., et al. (2018). FewRel: A Large-Scale Supervised Few-Shot Relation Classification Dataset with State-of-the-Art Evaluation. In Proceedings of the 2018 Con-ference on Empirical Methods in Natural Language Processing (pp. 4803 - 4809).

[7] Snell, J., Swersky, K., & Zemel, R. (2017). Prototypical networks for few-shot learning. Ad-vances in neural information processing systems, 30.

[8] Gao, T., Han, X., Liu, Z., et al. (2019). Hybrid attention-based prototypical networks for noisy few-shot relation classification. In Proceedings of the AAAI conference on artificial intelligence (Vol. 33, No. 01, pp. 6407 - 6414).

[9] Han, J., Cheng, B., & Lu, W. (2021). Exploring Task Difficulty for Few-Shot Relation Extrac-tion. In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing (pp. 2605 - 2616).

[10] Cong, X., Sheng, J., Cui, S., et al. (2022). Relation-guided few-shot relational triple extraction. In Proceedings of the 45th International ACM SIGIR Conference on Research and Devel-opment in Information Retrieval (pp. 2206 - 2213).

[11] Zhang, N., Chen, M., Bi, Z., et al. (2022). CBLUE: A Chinese Biomedical Language Under-standing Evaluation Benchmark. In Proceedings of the 60th Annual Meeting of the Associ-ation for Computational Linguistics (Volume 1: Long Papers) (pp. 7888 - 7915).