Abstract:
This paper implements a semantic role labeling in Chinese language, uses the convolution tree kernel of Support Vector Machine (SVM). It focuses on how to properly express the structural representation between predicates and arguments on dependency tree and let the input tree contain less noise information. It explores two methods to prune the dependency tree: Shortest Path Tree(SPT) and Minimum Tree(MT). Experimental results on the transferred corpuses from Chinese PropBank and Chinese NomBank show the system achieves 83.66 in labeled F1 on verbal predicates and 76.87 in labeled F1 on nominal predicates
Key words:
Semantic Role Labeling(SRL),
tree kernel,
dependency relationship,
nominal predicate,
verbal predicate
摘要: 使用SVM提供的卷积树核函数构造一个中文语义角色标注系统,将依存关系作为标注单元进行中文语义角色标注。通过不同的裁剪方法获得依存树的结构化信息,裁剪后的依存树分别为最短路径树和最小树。在中文PropBank和NomBank的转换语料上进行实验,结果表明,该系统在动词性谓词和名词性谓词语料上的F1值分别为83.66和76.87。
关键词:
语义角色标注,
树核,
依存关系,
名词性谓词,
动词性谓词
CLC Number:
WANG Bu-Kang, WANG Gong-Ling, ZHOU Guo-Dong. Semantic Role Labeling in Chinese Language Based on Tree Kernel Function[J]. Computer Engineering, 2011, 37(22): 128-130.
王步康, 王红玲, 周国栋. 基于树核函数的中文语义角色标注[J]. 计算机工程, 2011, 37(22): 128-130.