基于分层输出神经网络的汉语语义角色标注

摘要
参考文献
相关文章 (15)

全文: PDF (1574 KB) HTML (1 KB)
输出: BibTeX | EndNote (RIS)

摘要语义角色标注是自然语言处理中的一项重要任务。当下针对中文语义角色标注的主流做法是通过基于特征的统计机器学习实现的。然而,统计机器学习的方法需要引入经验性的人工特征,这在一定程度上增加了工作量。深度学习在自然语言处理领域的应用使得特征的自动学习成为可能。文章尝试了一种适用于语义角色标注的深层神经网络架构,该模型能自然地推广到其他标注任务。实验表明,深度学习算法能够有效地用于语义角色标注任务,但是我们仍然发现,模型对语义层面知识的学习是相当有限的,基于深度学习的方法还不能取代基于人工特征的统计机器学习算法。

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	王臻
	常宝宝
	穗志方

关键词 ：语义角色标注, 深度学习, 特征向量

Abstract：Semantic role labeling is an important task in Chinese natural language processing. Using feature based statistical machine learning to perform semantic role labeling is the mainstream method nowadays, denpeding heavily on manually designed features. This paper investigates semantic role labeling based on deep neural nets, which can learn features automatically. Experimental results show that our algorithm is promising. However, it cannot reach conventional machine learning methods with manually designed features yet.

Key words： semantic role labeling deep learning feature vectors

收稿日期: 2014-06-20

中图分类号:

TP391

基金资助:国家重点基础研究发展计划(2014CB340504);国家自然科学基金(61273318,61375074)

作者简介: 王臻(1990—),通讯作者,硕士研究生,主要研究领域为自然语言处理。常宝宝(1971—),副教授,主要研究领域为自然语言处理。穗志方(1970—),教授,主要研究领域为自然语言处理。

链接本文:

http://jcip.cipsc.org.cn/CN/Y2014/V28/I6/56

[1] Nianwen Xue. Building a Large-Scale Annotated Chinese Corpus[C]//Proceedings of the 19th international conference on Computational linguistics.2002: 1-8.
[2] Nianwen Xue, Martha Palmer. Annotating the propositions in the Penn Chinese Treebank[C]//Proceedings of the second SIGHAN workshop on Chinese language processing.2003: 47-54.
[3] Honglin Sun, Daniel Jurafsky. Shallow semantic parsing of Chinese[C]//Proceedings of NAACL-HLT.2004.
[4] Nianwen Xue. Labeling Chinese predicates with semantic roles[J]. Computational Linguistics, 2008, 34(2): 225-255.
[5] Weiwei Sun, Zhifang Sui. Chinese function tag labeling[C]//Proceedings of the 23^rd Pacific Asia Conference on Language, Information and Computation.2009.
[6] Weiwei Sun, Zhifang Sui, Meng Wang and Xin Wang. Chinese semantic role labeling with shallow parsing[C]//Proceedings of the 2009 Conference on Empirical Methods in Natural Language Processing.2009: 1475-1483.
[7] Collobert Ronan, Weston Jason. A unified architecture for natural language processing: Deep neural networks with multitask learning[C]//Proceedings of the 25^th international conference on machine learning.2008: 160-167.
[8] Nianwen Xue, Martha Palmer. Automatic semantic role labeling for Chinese verbs[C]//Proceedings of the 19^th International Joint Conference on Artificial Intelligence.2005.
[9] Weiwei Ding, Baobao Chang. Improving Chinese semantic role classification with hierarchical feature selection strategy[C]//Proceedings of the 2008 Conference on Empirical Methods in Natural Language Processing.2008.
[10] Yaodong Chen, Ting Wang, Huowang Chen, and Xishan Xu. Semantic role labeling of Chinese using transductive svm and semantic heuristics[C]//Proceedings of the Third International Joint Conference on Natural Language Processing.2008.
[11] Weiwei Sun. Improving Chinese semantic role labeling with rich syntactic features[C]//Proceedings of the ACL 2010 Conference Short Papers.2010: 168-172.
[12] Weiwei Sun. Semantics-driven shallow parsing for Chinese semantic role labeling[C]//Association for Computational Linguistics (ACL).2010.
[13] Bengio, Y. Practical recommendations for gradient-based training of deep architectures. In NN: Tricks of the Trade[M]. 2012: 437-478.
[14] Weiwei Ding, Baobao Chang. Fast semantic role labeling for Chinese based on semantic chunking[C]//Proceedings for of the 22^nd International Conference on Computer Processing of Oriental Languages.2009: 79-90.