近年来,基于数据驱动的命名实体识别方法在新闻、生物医疗等领域上取得了很大的成功,然而许多领域缺少标签,且人工标注成本高昂。为了降低标注成本,该文尝试使用含有噪声的部分标签数据进行命名实体识别,提出了一种基于部分标签数据和经验分布的方法。首先介绍基于部分标签数据的建模方法,然后引入标签经验分布的假设,通过将经验分布加入模型,有效降低了数据中的噪声。最后分别在植物病虫害数据集和优酷视频数据集上进行测试,结果表明,该方法优于其他方法。
Abstract
In recent years, data-driven named entity recognition(NER) methods have achieved great success in many fields such as news, biomedical and so on. In order to reduce the cost of labeling for a new domain, a NER method based on partially labelled data and empirical distribution is proposed. We describe the modeling method based on partially labelled data, and then introduce the hypothesis of label empirical distribution. By adding the empirical distribution to the model, the noise in the data is effectively reduced. Tested on the datasets of plant diseases &insect pests and Youku video, and the results show that the proposed method is better than other methods.
关键词
部分标签数据 /
经验分布 /
命名实体识别
{{custom_keyword}} /
Key words
partially labelled data /
empirical distribution /
named entity recognition
{{custom_keyword}} /
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] Lample G, Ballesteros M, Subramanian S, et al. Neural architectures for named entity recognition[J]. arXiv preprint arXiv:1603.01360, 2016.
[2] Wang Q, Zhou Y, Ruan T, et al. Incorporating dictionaries into deep neural networks for the Chinese clinical named entity recognition[J]. Journal of Biomedical Informatics, 2019, 92: 103-133.
[3] Li L, He H, Liu S, et al. Research of word representations on biomedical named entity recognition[J]. Journal of Chinese Computer Systems, 2016, 37(2): 302-307.
[4] Lafferty J, Mccallum A, Pereira F C N. Conditional random fields: Probabilistic models for segmenting and labeling sequence data[C]//Proceedings of the 18th Intemational Conference on Machine Learning. San Francisco,CA,USA:Morgan Kaufmann Publishers Inc.,2001:282-289.
[5] Tsochantaridis I, Joachims T, Hofmann T, et al. Large margin methods for structured and interdependent output variables[J]. Journal of machine learning research. 2005, 6(9): 1453-1484.
[6] Tsuboi Y, Kashima H, Mori S, et al. Training conditional random fields using incomplete annotations[C]// Proceedings of the 22nd International Conference on Computational Linguistics. Manchester, UK: CoNLL,2008:897-904.
[7] Hovy D, Hovy E. Exploiting partial annotations with EM training[C]//Proceedings of the NAACL-HLT Workshop on the Induction of Linguistic Structure,2012:31-38.
[8] Flannery D, Miyao Y, Neubig G, et al. A pointwise approach to training dependency parsers from partially annotated corpora[J]. Information and Media Technologies, 2012, 7(4): 1489-1513.
[9] Mori S, Nakata Y, Neubig G, et al. Pointwise Prediction and Sequence-Based Reranking for Adaptable Part-of-Speech Tagging[C]//Proceedings of the Conference of the Pacific Association for Computational Linguistics. Springer, 2015:3-17.
[10] Jie Z, Xie P, Lu W, et al. Better modeling of incomplete annotations for named entity recognition[C]//Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. 2019:729-734.
[11] Bellare K, Mccallum A. Learning extractors from unlabeled text using relevant databases[C]//Proceedings of the 6th international workshop on information integration on the web. 2007.
[12] Greenberg N, Bansal T, Verga P, et al. Marginal likelihood training of bilstm-crf for biomedical named entity recognition from disjoint label sets[C]//Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing. 2018: 2824-2829.
[13] 庄福振,罗平,何清,等. 迁移学习研究进展[J]. 软件学报, 2015, 26(1): 26-39.
[14] Wang M, Manning C D. Cross-lingual projected expectation regularization for weakly supervised learning[J]. Transactions of the Association for Computational Linguistics, 2014, (2): 55-66.
[15] Zirikly A, Hagiwara M. Cross-lingual transfer of named entity recognizers without parallel corpora[C]//Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing. 2015:390-396.
[16] Lin B Y, Lu W. Neural adaptation layers for cross-domain named entity recognition[J]. arXiv preprint arXiv:1810.06368, 2018.
[17] Lee J Y, Dernoncourt F, Szolovits P. Transfer learning for named-entity recognition with neural networks[J]. arXiv preprint arXiv:1705.06273, 2017.
[18] Yang Z, Salakhutdinov R, Cohen W W. Transfer learning for sequence tagging with hierarchical recurrent networks[J]. arXiv preprint arXiv:1703.06345, 2017.
[19] 李伟康,李炜,吴云芳.深度学习中汉语字向量和词向量结合方式探究[J].中文信息学报,2017,31(06):140-146.
[20] Napoles C, Gormley M, Van Durme B. Annotated gi-gaword[C]//Proceedings of the Joint Workshop on Automatic Knowledge Base Construction and Web-scale Knowledge Extraction (AKBC-WEKEX), 2012: 95-100.
[21] Hedderich M A, Klakow D. Training a neural network in a low-resource setting on automatically annotated noisy data[J]. arXiv preprint arXiv:1807.00745, 2018.
{{custom_fnGroup.title_cn}}
脚注
{{custom_fn.content}}
基金
自然保护地生态监测系统(2018HXKFXX018)
{{custom_fund}}
PDF(2627 KB)
