面向生物文本挖掘的语义标注研究  

Semantic Relation Annotation for Biomedical Text Mining Based on Recursive Directed Graph

作　　者：陈波[1,2] 吕晨[2] 魏小梅[2] 

机构地区：[1]湖北文理学院文学院,湖北襄阳441053 [2]武汉大学计算机学院,湖北武汉430072

出　　处：《湖北文理学院学报》2015年第7期29-32,共4页Journal of Hubei University of Arts and Science

基　　金：国家自然科学基金项目"汉语关联结构的资源建设及自动分析模型研究"(61202193);中国博士后科学基金(2013M540593;2014T70722);湖北省重点学科建设立项学科成果

摘　　要：文章提出了一个新颖的模型——"基于特征结构的递归有向图",将其用于描述英文生物文本中定语后置的语义关系。后置定语的用法是复杂多变的,主要有三类情况:现在分词充当后置定语,过去分词充当后置定语,介词短语充当后置定语,这为自动分析带来很多难题。我们总结和标注了这三类后置定语的语义信息。与依存结构相比,特征结构可以形式化为可递归的有向图,标注结果表明递归有向图更适合与生物文本挖掘中的复杂语义关系抽取。In this paper we propose a novel model ＂recursive directed graph＂ based on feature structure, and apply it to represent the semantic relations of postpositive attributive structures in biomedical texts. The usages of postpositive attributive are complex and variable, especially three categories： present participle phrase, past participle phrase, and preposition phrase as postpositive attributive, which always bring the difficulties of automatic parsing. We summarize these categories and annotate the semantic information. Compared with dependency structure, feature structure, being recursive directed graph, enhances semantic information extraction in biomedical field. The annotation results show that recursive directed graph is more suitable to extract complex semantic relations for biomedical text mining.

关 键 词：生物文本挖掘 语义标注 递归有向图 后置定语 

分 类 号：TP301[自动化与计算机技术—计算机系统结构] H085[自动化与计算机技术—计算机科学与技术]