两阶段问答范式的生物医学事件触发词检测

doi:10.3778/j.issn.1002-8331.2301-0152

摘要/Abstract

摘要： 现有的生物医学事件触发词检测存在以下缺陷：保留了与触发词无关的冗余信息；忽略了实体与事件之间的潜在关联性；传统方法容易受到数据稀缺性的影响。针对上述问题，提出了一种两阶段问答范式的生物医学事件触发词检测方法。在事件类型识别阶段，采用基于句法距离的注意力捕获更有意义的上下文特征，排除无关信息的干扰；为了有效利用实体中的潜在特征，采用全局统计的单词-实体-事件共现特征，指导事件类型感知注意力挖掘词与事件之间的强关联性。在触发词定位阶段，根据识别出的事件类型，制定问题回答该事件对应的触发词索引，从而利用丰富的问答数据库实现数据增强。在MLEE语料库上的结果表明，两阶段问答范式、句法距离和事件类型感知注意力都有效地提升了模型性能，所提出的模型取得了81.39%的F1分数，并在多个事件类型上的详细结果均优于其他基线模型。

关键词: 生物医学事件, 触发词检测, 句法距离, 单词-实体-事件共现特征, 两阶段问答范式

Abstract: The existing biomedical event trigger detection methods have the following defects: Redundant information unrelated to triggers are retained; potential correlations between entities and events are ignored; traditional methods are vulnerable to data scarcity. A biomedical event trigger detection based on two-stage question answering paradigm is proposed to address the above problems. In the event type identification phase, in order to exclude the interference of irrelevant information, the attention based on syntactic distance is allowed to capture more meaningful contextual features. In order to effectively utilize the potential features in the entities, the word-entity-event co-occurrence feature based on global statistics is used to guide event type aware attention to explore the strong relationship between words and events. In the trigger localization phase, the trigger index of the event in the sentence is answered according to the identified event type questions, thus leveraging the rich question answering database to achieve data enhancement. The results on the MLEE corpus show that the two-stage question answering paradigm, syntactic distance attention, and event type aware attention effectively improve the performance of the model, and the proposed model achieves 81.39% F1-score, outperforming other baseline models in terms of detailed results for multiple event types.

Key words: biomedical events, trigger detection, syntactic distance, word-entity-event co-occurrence feature, two-stage question answering paradigm

行帅, 熊玉洁, 苏前敏, 黄继汉. 两阶段问答范式的生物医学事件触发词检测[J]. 计算机工程与应用, 2024, 60(10): 121-131.

XING Shuai, XIONG Yujie, SU Qianmin, HUANG Jihan. Biomedical Event Trigger Detection Based on Two-Stage Question Answering Paradigm[J]. Computer Engineering and Applications, 2024, 60(10): 121-131.

参考文献

[1] ANANIADOU S, THOMPSON P, NAWAZ R, et al. Event-based text mining for biology and functional genomics[J]. Briefings in Functional Genomics, 2015, 14(3): 213-230.
[2] ROTMENSCH M, HALPERN Y, TLIMAT A, et al. Learning a health knowledge graph from electronic medical records[J]. Scientific Reports, 2017, 7(1): 1-11.
[3] SPANGHER A, PENG N, MAY J, et al. Enabling low-resource transfer learning across COVID-19 corpora by combining event-extraction and co-training[C]//Proceedings of the 1st Workshop on NLP for COVID-19 at ACL 2020, 2020.
[4] WANG J, ZHANG J, AN Y, et al. Biomedical event trigger detection by dependency-based word embedding[J]. BMC Medical Genomics, 2016, 9(2): 123-133.
[5] ZHOU D, ZHONG D, HE Y. Event trigger identification for biomedical events extraction using domain knowledge[J]. Bioinformatics, 2014, 30(11): 1587-1594.
[6] CUI S, YU B, LIU T, et al. Edge-enhanced graph convolution networks for event detection with syntactic relation[J]. arXiv:2002.10757, 2020.
[7] FEI H, REN Y, JI D. A tree-based neural network model for biomedical event trigger detection[J]. Information Sciences, 2020, 512: 175-185.
[8] LI D, HUANG L, JI H, et al. Biomedical event extraction based on knowledge-driven tree-LSTM[C]//Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1, Long and Short Papers)， 2019: 1421-1430.
[9] ZHAO W, ZHANG J, YANG J, et al. A novel joint biomedical event extraction framework via two-level modeling of documents[J]. Information Sciences, 2021, 550: 27-40.
[10] AHMAD W U, PENG N, CHANG K W. GATE: graph attention transformer encoder for cross-lingual relation and event extraction[C]//Proceedings of the 35th AAAI Conference on Artificial Intelligence, 2021: 12462-12470.
[11] LI L, LIU Y. Exploiting argument information to improve biomedical event trigger identification via recurrent neural networks and supervised attention mechanisms[C]//Proceedings of the 2017 IEEE International Conference on Bioinformatics and Biomedicine, 2017: 565-568.
[12] LI F, PENG W, CHEN Y, et al. Event extraction as multi-turn question answering[C]//Findings of the Association for Computational Linguistics: EMNLP 2020, 2020: 829-838.
[13] RONG W, NIE Y, SHEN Y, et al. Biological event trigger identification with noise contrastive estimation[J]. IEEE/ACM Transactions on Computational Biology and Bioinformatics, 2017, 15(5): 1549-1559.
[14] TEES15. BJ?RNE J, SALAKOSKI T. TEES 2.2: biomedical event extraction for diverse corpora[J]. BMC Bioinformatics, 2015, 16(16): 1-20.
[15] PYYSALO S, OHTA T, MIWA M, et al. Event extraction across multiple levels of biological organization[J]. Bioinformatics, 2012, 28(18): 575-581.
[16] HE X, LI L, LIU Y, et al. A two-stage biomedical event trigger detection method integrating feature selection and word embeddings[J]. IEEE/ACM Transactions on Computational Biology and Bioinformatics, 2017, 15(4): 1325-1332.
[17] RAHUL P V S S, SAHU S K, ANAND A. Biomedical event trigger identification using bidirectional recurrent neural network based models[J]. arXiv:1705.09516, 2017.
[18] WANG Y, WANG J, LIN H, et al. Biomedical event trigger detection based on bidirectional LSTM and CRF[C]//Proceedings of the 2017 IEEE International Conference on Bioinformatics and Biomedicine, 2017: 445-450.
[19] WANG Y, WANG J, LIN H, et al. Bidirectional long short-term memory with CRF for detecting biomedical event trigger in FastText semantic space[J]. BMC Bioinformatics, 2018, 19(20): 59-66.
[20] SHEN C, LIN H, FAN X, et al. Biomedical event trigger detection with convolutional highway neural network and extreme learning machine[J]. Applied Soft Computing, 2019, 84: 105661.
[21] WEI H, ZHOU A, ZHANG Y, et al. Biomedical event trigger extraction based on multi-layer residual BiLSTM and contextualized word representations[J]. International Journal of Machine Learning and Cybernetics, 2022, 13(3): 721-733.
[22] HE X, REN Y, TAI P, et al. A two-stage biomedical event trigger detection method based on hybrid neural network and sentence embeddings[J]. IEEE Access, 2021, 9: 81926-81935.
[23] HE X, LI L, WAN J, et al. Biomedical event trigger detection based on BiLSTM integrating attention mechanism and sentence vector[C]//Proceedings of the 2018 IEEE International Conference on Bioinformatics and Biomedicine, 2018: 651-654.
[24] LI L, HUANG M, LIU Y, et al. Contextual label sensitive gated network for biomedical event trigger extraction[J]. Journal of Biomedical Informatics, 2019, 95: 103221.
[25] HUANG K H, YANG M, PENG N. Biomedical event extraction with hierarchical knowledge graphs[J]. arXiv:2009.09335, 2020.
[26] DEVLIN J, CHANG M W, LEE K, et al. BERT: pre-training of deep bidirectional transformers for language understanding[J]. arXiv:1810.04805, 2018.
[27] MCCANN B, KESKAR N S, XIONG C, et al. The natural language decathlon: multitask learning as question answering[J]. arXiv:1806.08730, 2018.
[28] WANG X D, WEBER L, LESER U. Biomedical event extraction as multi-turn question answering[C]//Proceedings of the 11th International Workshop on Health Text Mining and Information Analysis, 2020: 88-96.
[29] GU Y, TINN R, CHENG H, et al. Domain-specific language model pretraining for biomedical natural language processing[J]. ACM Transactions on Computing for Healthcare, 2021, 3(1): 1-23.
[30] WADDEN D, WENNBERG U, LUAN Y, et al. Entity, relation, and event extraction with contextualized span representations[J]. arXiv:1909.03546, 2019.
[31] KLEIN D, MANNING C D. Accurate unlexicalized parsing[C]//Proceedings of the 41st Annual Meeting of the Association for Computational Linguistics, 2003: 423-430.
[32] DIAO Y, LIN H, YANG L, et al. FBSN: a hybrid fine-grained neural network for biomedical event trigger identification[J]. Neurocomputing, 2020, 381: 105-112.
[33] HE X, TAI P, LU H, et al. A biomedical event extraction method based on fine-grained and attention mechanism[J]. BMC Bioinformatics, 2022, 23(1): 1-17.