【作者】 吴楠； 王庆林； 刘禹；

【Author】 WU Nan~1,WANG Qinglin~1,LIU Yu~2 (1.School of Automation,Beijing Institute of Technology,Beijing 100081,China; 2.Institute of Automation,Chinese Academy of Sciences,Beijing 100190,China)

【机构】 北京理工大学自动化学院； 中国科学院 自动化研究所；

【摘要】 提出了一种基于百科词条的领域本体关系抽取方法,该方法首先把气候变化领域种子术语作为候选词,抽取领域相关词汇及其对应的百科名片词条解释,然后对所有百科名片词条解释进行分词、词性标注、词语过滤、频次统计,筛选出用于学习关系抽取模板的特征词,最后利用特征词得到关系抽取模板。实验表明,将该方法得到的模板应用于气候变化领域概念的上下位关系抽取,可得到较好的抽取效果。更多还原

【Abstract】 A method of domain ontology relation extraction was presented based on Encyclopedia entries.First,the field of climate change seed terminologies were used as the candidate words in order to extract the related vocabularies and its corresponding Encyclopedia card entries.Then,the process of word segmentation,POS tagging,words filtering,and word-frequency statistics,was employed to screen out the feature words for learning relation extraction templates.Finally, the relation extraction templates are obtained based on the feature words.The experiments verify that the extraction rate of the method is good.更多还原