汉语框架网络本体时间语义属性的构建与信息抽取研究  

Construction of Time Semantic Properties in the Chinese FrameNet Ontology and Information Extraction

作　　者：由丽萍[1] 

机构地区：[1]山西大学管理学院,山西太原030006

出　　处：《情报科学》2013年第8期84-89,共6页Information Science

基　　金：教育部人文社会科学研究规划基金项目(12YJA740095)

摘　　要：首先对自主研发的汉语框架网络本体中时间语义信息进行了补充,大大细化了时间子类范畴,并用OWL本体标记语言进行了形式化表示;相应地构建了时间类语义属性标注规则,为计算机自动语义分析提供规则库;最后设计了基于本体的信息抽取模型,并利用构建的时间语义标注规则,在大规模语料数据的支持下进行了实验,取得了较理想的实验结果。This paper firstly amended and supplemented the semantic information of time in the existing self-developed Chinese FrameNet Ontology,significantly refining the subclass categories and represent ing them in the OWL ontology markup language;secondly built a set of rules for the identification of time properties,used as a rule base for computer automatic semantic analysis;finally designed an ontolo gy-based information extraction model,made an experiment with the improved ontology in support of large-scale corpus data and experimental results and got an ideal experimental result.

关 键 词：本体 框架网络 信息抽取 

分 类 号：G254[文化科学—图书馆学]