Abstract: Aiming at the bottlenecks of syntactic tree-based Semantic Role Labeling(SRL), this paper explores dependency relationship-based semantic role labeling. By properly integrating dependency parsing, predicate identification, feature extraction, semantic role identification and semantic role classification, this system achieves the F1 measure of 80.94% on the WSJ portion of the CoNLL2008 SRL Shared Task, using automatic dependency parsing. Experimental results show that it is better than the F1 measure of syntactic tree-based semantic role labeling apparently.

Key words: Semantic Role Labeling(SRL), dependency parsing, dependency relationship

摘要： 针对以句法成分为基本标注单元语义角色标注的瓶颈问题，描述一个以依存关系为标注单元的语义角色标注系统，经过依存关系分析、谓词标识、特征抽取、角色识别和角色分类，最终在CoNLL2008 SRL Shared Task自动依存分析的WSJ测试集取得了较好的结果，F1值达到了80.94%，结果证明其性能明显好于基于句法分析的SRL。

关键词: 语义角色标注, 依存分析, 依存关系

CLC Number: 

• TP311