General Framework for Event Extraction Based on Task Transformation

doi:10.3778/j.issn.1002-8331.2304-0387

Abstract

Abstract: Arguments scattering, multiple events and argument overlapping are the long-term challenges of event extraction, and the absence of trigger-word labels or position labels widely exists in many event extraction tasks. To address above problems, a general framework for event extraction based on task transformation is proposed, which mainly includes three modules：task transformation, relationship extraction, and event prediction. The labeled events with and without trigger-words are transformed into different forms of relational triplets, respectively. These triplets, along with the original text, serve as the training data for the relationship extraction model. During event prediction, relational triples are extracted from the input text, and then they are combined into target events. In addition, an argument localization algorithm based on shortest distance is designed to address the situation of missing positional information in the training corpus. The framework has achieved an average F1-score of 81.6% on the ChFinAnn dataset and an F1-score of 72.04% on the DuEE-Fin dataset (online submission), both reaching the current SOTA levels. The experimental results show that this framework not only significantly improves the performance of event extraction, but also has a wide range of adaptability.

Key words: event extraction, general framework, task transformation, argument positioning

摘要： 论元分散、多重事件和论元重叠是事件抽取长期面临的挑战，许多任务还存在触发词或位置标注缺失的情况。对此，提出一种基于任务转化的事件抽取通用框架，包括任务转化、关系抽取和事件预测3个模块。有触发词和无触发词的标注事件分别被转化为不同形式的关系三元组；这些三元组将与原始文本一起作为关系抽取模型的训练数据；事件预测时先从输入文本中抽取三元组，再将它们还原为目标事件。对于训练语料中位置标注缺失的情况，设计了基于最短距离的论元定位算法。该框架在ChFinAnn数据集上取得81.6%的平均F1值，在DuEE-Fin数据集上的F1值为72.04%（在线提交结果），均达到目前的SOTA水平。实验结果表明，该框架不仅可以显著提高事件抽取效果，而且具有广泛的适应能力。

关键词: 事件抽取, 通用框架, 任务转化, 论元定位

LI Jian, HU Ruijuan, ZHANG Keliang, LIU Haiyan. General Framework for Event Extraction Based on Task Transformation[J]. Computer Engineering and Applications, 2024, 60(15): 133-142.

李健, 胡瑞娟, 张克亮, 刘海砚. 基于任务转化的事件抽取通用框架[J]. 计算机工程与应用, 2024, 60(15): 133-142.

References

[1] HOGENBOOM F, FRASINCAR F, KAYMAK U, et al. A survey of event extraction methods from text for decision support systems[J]. Decision Support Systems, 2016, 85: 12-22.
[2] SHIBUYA T, HOVY E H. Nested named entity recognition via second-best sequence learning and decoding[J]. Transactions of the Association for Computational Linguistics, 2020, 8: 605-620.
[3] LI X, YIN F, SUN Z, et al. Entity-relation extraction as multi-turn question answering[C]//Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics: Association for Computational Linguistics, 2019: 1340-1350.
[4] DODDINGTON G R, MITCHELL A, PRZYBOCKI M A, et al. The automatic content extraction (ACE) program-tasks, data, and evaluation[C]//Proceedings of the Fourth International Conference on Language Resources and Evaluation (LREC’04), 2004.
[5] SU J. GPlinker: entity-relation joint extraction based on GlobalPointer[EB/OL]. (2022). https://kexue.fm/archives/8373.
[6] WEI Z, SU J, WANG Y, et al. A novel cascade binary tagging framework for relational triple extraction[C]//Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, 2020: 1476-1488.
[7] WANG Y, YU B, ZHANG Y, et al. TPLinker: single stage joint extraction of entities and relations through token pair linking[C]//Proceedings of the 28th International Conference on Computational Linguistics, 2020: 1572-1582.
[8] DEVLIN J, CHANG M W, LEE K, et al. BERT: pre-training of deep bidirectional transformers for language understanding[C]//Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics, 2019: 4171-4186.
[9] SU J, LU Y, PAN S, et al. Roformer: enhanced Transformer with rotary position embedding[J]. arXiv:2104.09864, 2021.
[10] LI X, LI F, PAN L, et al. DuEE: a large-scale dataset for Chinese event extraction in real-world scenarios[C]//CCF International Conference on Natural Language Processing and Chinese Computing, 2020: 534-545.
[11] LI X. DuEE-Fin: a document-level event extraction dataset in the financial domain released by Baidu[EB/OL]. (2021). https://aistudio.baidu.com/aistudio/competition/detail/46.
[12] ZHENG S, CAO W, XU W, et al. Doc2EDAG: an end-to-end document-level framework for Chinese financial event extraction[C]//Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing. [S.l.]: Association for Computational Linguistics, 2019: 337-346.
[13] YANG H, CHEN Y, LIU K, et al. DCFEE: a document level Chinese financial event extraction system based on automatically labeled training data[C]//Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics-System Demonstrations, Melbourne, Australia, July 15-20, 2018: 50-55.
[14] YANG H, SUI D, CHEN Y, et al. Document-level event extraction via parallel prediction networks[C]//Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics, 2021: 6298-6308.
[15] XU R, LIU T, LI L, et al. Document-level event extraction via heterogeneous graph-based interaction model with a tracker[C]//Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics, 2021: 3533-3546.
[16] RILOFF E. Automatically constructing a dictionary for information extraction tasks[C]//Proceedings of the AAAI Conference on Artificial Intelligence, 1993: 811-816.
[17] KIM J, MOLDOVAN D I. Acquisition of linguistic patterns for knowledge-based information extraction[J]. IEEE Transactions on Knowledge and Data Engineering, 1995, 7: 713-724.
[18] RILOFF E, SHOEN J. Automatically acquiring conceptual patterns without an annotated corpus[C]//Third Workshop on Very Large Corpora, 1995.
[19] JUNGERMANN F, MORIK K. Enhanced services for targeted information retrieval by event extraction and data mining[C]//International Conference on Application of Natural Language to Information Systems. Berlin Heidelberg: Springer, 2008: 335-336.
[20] CHIEU H L, NG H T. A maximum entropy approach to information extraction from semi-structured and free text[C]//Eighteenth National Conference on Artificial Intelligence. [S.l.]: American Association for Artificial Intelligence, 2002: 786-791.
[21] AHN D. The stages of event extraction[C]//Proceedings of the Workshop on Annotating and Reasoning About Time and Events. [S.l.]: Association for Computational Linguistics, 2006: 1-8.
[22] MAJUMDER A, EKBAL A, NASKAR S K. Feature selection and class-weight tuning using genetic algorithm for bio-molecular event extraction[C]//Natural Language Processing and Information Systems. [S.l.]: Springer International Publishing, 2017: 28-33.
[23] LI Q, JI H, HUANG L. Joint event extraction via structured prediction with global features[C]//Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), 2013: 73-82.
[24] LIAO S, GRISHMAN R. Can document selection help semi-supervised learning? a case study on event extraction[C]//Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies. [S.l.]: Association for Computational Linguistics, 2011: 260-265.
[25] CHEN Y, XU L, LIU K, et al. Event extraction via dynamic multi-pooling convolutional neural networks[C]//Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing (Volume 1: Long Papers). [S.l.]: Association for Computational Linguistics, 2015: 167-176.
[26] NGUYEN T H, CHO K, GRISHMAN R. Joint event extraction via recurrent neural networks[C]//Proceedings of the 2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. [S.l.]: Association for Computational Linguistics, 2016: 300-309.
[27] LIU J, CHEN Y, LIU K, et al. Event extraction as machine reading comprehension[C]//Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing. [S.l.]: Association for Computational Linguistics, 2020: 1641-1651.
[28] LYU Q, ZHANG H, SULEM E, et al. Zero-shot event extraction via transfer learning: challenges and insights[C]//Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 2: Short Papers). [S.l.]: Association for Computational Linguistics, 2021: 322-332.
[29] GAO J, ZHAO H, YU C, et al. Exploring the feasibility of ChatGPT for event extraction[J]. arXiv:2303.03836, 2023.