融合上下文依赖和句子语义的事件线索检测研究
|
摘要
事件线索检测旨在从自由文本中自动抽取触发事件的词或短语。现有的英文事件线索检测方法依赖于特征提取工具,这样会造成错误传递,而且忽略了待测词与上下文的依赖关系和句子的语义信息,这些信息对事件线索检测是很有帮助的。提出一种神经网络方法,利用双向长短时记忆网络(bidirectional long short-term memory,Bi-LSTM)抓取待测词在句子中的上下文依赖,同时使用门控循环神经网络(gated recurrent neural network,GRNN)学习句子的语义表示,融合这两种信息来提高事件线索词的识别能力。在KBP 2015评测语料上的实验结果显示,该方法是有效的,并且性能比baseline方法有显著提高。
Event nugget detection aims to extract words or phrases which trigger events from the free text automatically. Existing methods about English event nugget detection rely on feature extraction tools, which will cause error propagation. Besides, they ignore the dependency between the candidate word and the context and the semantic information of the sentence, which are helpful to detect event nugget. This paper proposes a neural network method,which uses bidirectional long short-term memory neural network(Bi-LSTM) to crawl the dependency of sentence context while using the gated recurrent neural network(GRNN) to learn sentence semantic representation. Then this paper combines these information to improve the ability of identifying event nugget. The experimental results on the corpus of KBP 2015 evaluation show that the proposed method is effective and significantly higher than baseline methods.
Event nugget detection aims to extract words or phrases which trigger events from the free text automatically. Existing methods about English event nugget detection rely on feature extraction tools, which will cause error propagation. Besides, they ignore the dependency between the candidate word and the context and the semantic information of the sentence, which are helpful to detect event nugget. This paper proposes a neural network method,which uses bidirectional long short-term memory neural network(Bi-LSTM) to crawl the dependency of sentence context while using the gated recurrent neural network(GRNN) to learn sentence semantic representation. Then this paper combines these information to improve the ability of identifying event nugget. The experimental results on the corpus of KBP 2015 evaluation show that the proposed method is effective and significantly higher than baseline methods.
引文
[1]Li Qi,Ji Heng,Huang Liang.Joint event extraction via structured prediction with global features[C]//Proceedings of the51st Annual Meeting of the Association for Computational Linguistics,Sofia,Aug 4-9,2013.Stroudsburg:ACL,2013:73-82.
[2]Liao Shasha,Grishman R.Using document level cross-event inferent to improve event extraction[C]//Proceedings of the48th Annual Meeting of the Association for Computational Linguistics,Uppsala,Jul 11-16,2010.Stroudsburg:ACL,2010:789-797.
[3]Ji Heng,Grishman R.Refining event extraction through crossdocument inferent[C]//Proceedings of the 46th Annual Meeting of the Association for Computational Linguistics,Columbus,Jun 15-20,2008.Stroudsburg:ACL,2008:254-262.
[4]Hong Yu,Zhang Jianfeng,Ma Bin,et al.Using cross-entity inference to improve event extraction[C]//Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics:Human Language Technologies,Portland,Jun 19-24,2011.Stroudsburg:ACL,2011:1127-1136.
[5]Chen Yubo,Xu Liheng,Liu Kang,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,Beijing,Jul 26-31,2015.Stroudsburg:ACL,2015:167-176.
[6]Nguyen T H,Grishman R.Event detection and domain adaptation with 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,Beijing,Jul 27-29,2015.Stroudsburg:ACL,2015:365-371.
[7]Le Q,Mikolov T.Distributed representations of sentences and documents[C]//Proceedings of the 31st International Conference on Machine Learning,Beijing,Jun 21-26,2014:1188-1196.
[8]Hochreiter S,Schmidhuber J.Long short-term memory[J].Neural Computation,1997,9(8):1735-1780.
[9]Chen Xinchi,Qiu Xipenng,Zhu Chenxi,et al.Long shortterm memory neural networks for Chinese word segmentation[C]//Proceeding of the 2015 Conference on Empirical Methods in Natural Language Processing,Lisbon,Sep 17-21,2015.Stroudsburg:ACL,2015:1197-1206.
[10]Sutskever I,Vinyals O,Le Q.Sequence to sequence learning with neural networks[C]//Proceedings of the 27th International Conference on Neural Information Processing Systems,Montreal,Dec 8-13,2014.Cambridge:MIT Press,2014:3104-3112.
[11]Liu Pengfei,Qiu Xipeng,Chen Xinchi,et al.Multi-timescale long short-term memory neural network for modelling sentences and documents[C]//Proceeding of the 2015Conference on Empirical Methods on Natural Language Processing,Lisbon,Sep 17-21,2015.Stroudsburg:ACL,2015:2326-2335.
[12]Johnson R,Zhang Tong.Effective use of word order for text categorization with convolutional neural networks[C]//Proceedings of the 2015 Conference of the North American Chapter of the Association for Computational Linguistics:Human Language Technologies,Denver,May 31-Jun 5,2015.Stroudsburg:ACL,2015:103-112.
[13]Kim Y.Convolutional neural networks for sentence classification[C]//Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing,Doha,Oct 25-29,2014.Stroudsburg:ACL,2014:1746-1751.
[14]Hinton G E,Srivastava N,Krizhevsky A,et al.Improving neural networks by preventing co-adaptation of feature detectors[J].Computer Science,2012,3(4):212-223.
[15]Zhang Rui,Gong Weiguo,Grzeda V,et al.An adaptive learning rate method for improving adaptability of background models[J].IEEE Signal Processing Letters,2013,20(12):1266-1269.
[2]Liao Shasha,Grishman R.Using document level cross-event inferent to improve event extraction[C]//Proceedings of the48th Annual Meeting of the Association for Computational Linguistics,Uppsala,Jul 11-16,2010.Stroudsburg:ACL,2010:789-797.
[3]Ji Heng,Grishman R.Refining event extraction through crossdocument inferent[C]//Proceedings of the 46th Annual Meeting of the Association for Computational Linguistics,Columbus,Jun 15-20,2008.Stroudsburg:ACL,2008:254-262.
[4]Hong Yu,Zhang Jianfeng,Ma Bin,et al.Using cross-entity inference to improve event extraction[C]//Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics:Human Language Technologies,Portland,Jun 19-24,2011.Stroudsburg:ACL,2011:1127-1136.
[5]Chen Yubo,Xu Liheng,Liu Kang,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,Beijing,Jul 26-31,2015.Stroudsburg:ACL,2015:167-176.
[6]Nguyen T H,Grishman R.Event detection and domain adaptation with 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,Beijing,Jul 27-29,2015.Stroudsburg:ACL,2015:365-371.
[7]Le Q,Mikolov T.Distributed representations of sentences and documents[C]//Proceedings of the 31st International Conference on Machine Learning,Beijing,Jun 21-26,2014:1188-1196.
[8]Hochreiter S,Schmidhuber J.Long short-term memory[J].Neural Computation,1997,9(8):1735-1780.
[9]Chen Xinchi,Qiu Xipenng,Zhu Chenxi,et al.Long shortterm memory neural networks for Chinese word segmentation[C]//Proceeding of the 2015 Conference on Empirical Methods in Natural Language Processing,Lisbon,Sep 17-21,2015.Stroudsburg:ACL,2015:1197-1206.
[10]Sutskever I,Vinyals O,Le Q.Sequence to sequence learning with neural networks[C]//Proceedings of the 27th International Conference on Neural Information Processing Systems,Montreal,Dec 8-13,2014.Cambridge:MIT Press,2014:3104-3112.
[11]Liu Pengfei,Qiu Xipeng,Chen Xinchi,et al.Multi-timescale long short-term memory neural network for modelling sentences and documents[C]//Proceeding of the 2015Conference on Empirical Methods on Natural Language Processing,Lisbon,Sep 17-21,2015.Stroudsburg:ACL,2015:2326-2335.
[12]Johnson R,Zhang Tong.Effective use of word order for text categorization with convolutional neural networks[C]//Proceedings of the 2015 Conference of the North American Chapter of the Association for Computational Linguistics:Human Language Technologies,Denver,May 31-Jun 5,2015.Stroudsburg:ACL,2015:103-112.
[13]Kim Y.Convolutional neural networks for sentence classification[C]//Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing,Doha,Oct 25-29,2014.Stroudsburg:ACL,2014:1746-1751.
[14]Hinton G E,Srivastava N,Krizhevsky A,et al.Improving neural networks by preventing co-adaptation of feature detectors[J].Computer Science,2012,3(4):212-223.
[15]Zhang Rui,Gong Weiguo,Grzeda V,et al.An adaptive learning rate method for improving adaptability of background models[J].IEEE Signal Processing Letters,2013,20(12):1266-1269.