基于形式概念分析的面向药物适应症的领域本体构建方法
发布时间：2018-07-24 18:25
【摘要】：[目的/意义]使用本体对医药信息进行组织和表示,不仅能够在语义的层面上有效地管理和利用医药知识,而且能够实现领域知识的融合共享,推动医药领域的智能应用。当前对药物本体的构建主要借助规范化的医学术语系统,忽略了药物适应症自然文本中隐含的药物语义关联。因此提出一种面向药物适应症的领域本体构建方法。[方法/过程]以"抗肿瘤药物"为例,对265条药物信息进行处理;将适应症文本进行语义补全,使用二次分词和合并的方法抽取特征词;利用形式概念分析,以"药物—适应症"的关联构建形式背景并生成概念格,根据概念格的结构对药物关系进行描述;使用OWL语言对概念格转化而来的本体进行描述。[结果/结论]对构建的本体进行一致性检验,并以可视化方式展示和分析。实验表明该构建方法能够表示药物与适应症的语义关联。
[Abstract]:[objective / significance] Organization and representation of pharmaceutical information by ontology can not only effectively manage and utilize pharmaceutical knowledge at the semantic level, but also realize the fusion and sharing of domain knowledge and promote the intelligent application in the field of medicine. At present, the construction of drug ontology mainly relies on the standardized medical terminology system, ignoring the drug semantic association implied in the natural text of drug indications. Therefore, a domain ontology method for drug indication is proposed. [method / process] taking "antitumor drugs" as an example, 265 drug information were processed; the indication text was semantically completed, and the feature words were extracted using the method of secondary participle and combination. The formal background and concept lattice are constructed with the association of "drug indications", and the drug relations are described according to the structure of concept lattice, and the ontology transformed by concept lattice is described by OWL language. Results / conclusions: the conformance of the constructed ontology is verified, and visualized and analyzed. The experimental results show that this method can represent the semantic association between drugs and indications.
【作者单位】： 武汉大学信息系统研究中心;武汉大学信息管理学院;
【基金】：国家自然科学基金项目“基于文本和Web语义分析的智能咨询服务研究”(编号:71673209)研究成果之一
【分类号】：R-05;TP391.1
，
本文编号：2142263
[Abstract]:[objective / significance] Organization and representation of pharmaceutical information by ontology can not only effectively manage and utilize pharmaceutical knowledge at the semantic level, but also realize the fusion and sharing of domain knowledge and promote the intelligent application in the field of medicine. At present, the construction of drug ontology mainly relies on the standardized medical terminology system, ignoring the drug semantic association implied in the natural text of drug indications. Therefore, a domain ontology method for drug indication is proposed. [method / process] taking "antitumor drugs" as an example, 265 drug information were processed; the indication text was semantically completed, and the feature words were extracted using the method of secondary participle and combination. The formal background and concept lattice are constructed with the association of "drug indications", and the drug relations are described according to the structure of concept lattice, and the ontology transformed by concept lattice is described by OWL language. Results / conclusions: the conformance of the constructed ontology is verified, and visualized and analyzed. The experimental results show that this method can represent the semantic association between drugs and indications.
【作者单位】： 武汉大学信息系统研究中心;武汉大学信息管理学院;
【基金】：国家自然科学基金项目“基于文本和Web语义分析的智能咨询服务研究”(编号:71673209)研究成果之一
【分类号】：R-05;TP391.1
，
本文编号：2142263
本文链接：https://www.wllwen.com/kejilunwen/ruanjiangongchenglunwen/2142263.html
