基于图的半监督关系抽取-软件学报.pdfVIP

ISSN 1000-9825, CODEN RUXUEW E-mail: jos@ Journal of Software , Vol.19, No.11, November 2008, pp.2843−2852 DOI: 10.3724/SP.J.1001.2008.02843 Tel/Fax: +86-10 © 2008 by Journal of Software . All rights reserved. ∗ 基于图的半监督关系抽取 1+ 2 陈锦秀 , 姬东鸿 1(厦门大学 智能科学与技术系,福建 厦门 361005) 2(武汉大学计算机系,湖北 武汉 430072) Graph-Based Semi-Supervised Relation Extraction CHEN Jin-Xiu1+, JI Dong-Hong2 1(Department of Cognitive Science, Xiamen University, Xiamen 361005, China) 2(Department of Computer Science, Wuhan University, Wuhan 430072, China) + Corresponding author: E-mail: cjx@ Chen JX, Ji DH. Graph-Based semi-supervised relation extraction. Journal of Software, 2008,19(11): 2843−2852. /1000-9825/19/2843.htm Abstract : This paper investigates a graph-based semi-supervised learning algorithm, that is, label propagation algorithm for relation extraction. Labeled and unlabeled examples are represented as the nodes, and their distances as the weights of edges in the graph. The relation extraction tries to obtain a labeling function on this graph to satisfy the global consistency assumption. Experimental results on the ACE (automatic content extraction) corpus showed that this method achieves a better performance than SVM (support vector machine) when only very few labeled examples are available, and it also performs better than bootstrapping for the relation extraction task. Key words: relation extraction; graph-based; semi-supervised learning; label propagation 摘 要: 提出利用基于图的半监督学习算法,即标注传递算法,指导计算机从非结构化的文本中自动识别出实体 之间的关系.该方法首先利用图策略来建立关系抽取的模型.在这个图模型中,各个有标签和未标签的样本被表示成 图上的各个节点,而样本间的距离则作为图上各边的权重.然后,关系抽取的任务就转化成在这个图上估计出一个满 足全局一致性假设的标注