Capturing the Relationship between Evolving Biomedical Concepts via Background Knowledge
Capturing the Relationship between Evolving Biomedical Concepts via Background Knowledge
复制标题
通过背景知识捕捉不断发展的生物医学概念之间的关系
DOI：
--
复制
发表时间：
2016
期刊：
影响因子：
--
通讯作者：
Marcos Da Silveira
中科院分区：
文献类型：
--
作者：
C. Pruski;Júlio Cesar dos Reis;Marcos Da Silveira
语义Web应用程序和基于知识的系统严重依赖于最新本体的使用。为了支持它们充分的进化，方法必须检测到概念中含义随时间的变化。本文建议利用特定领域的外部知识源来描述动态本体中概念的演变。我们的原始技术分析了概念属性中价值的演变。该方法使用来自在线存储库的本体属性和本体之间的映射来推断概念及其后续版本之间关系的本质。通过使用各种连续版本的生物医学本体比较不同的配置，对所提出的算法进行了实验评估。研究结果揭示了在本体演化过程中考虑外部知识来源对识别正确的语义关系的好处。
Semantic Web applications and knowledge-based systems heavily rely on the use of up-to-date ontologies. To support their adequate evolution, methods must detect the changes of meanings in concepts over time. This article proposes exploiting domain-specific external source of knowledge to characterize the evolution of concepts in dynamic ontologies. Our original technique analyses the evolution of values in concept attributes. The approach uses ontological properties and mappings between ontologies from online repositories to deduce the nature of the relationship between a concept and its successive version. The proposed algorithm is experimentally evaluated comparing different configurations by using various successive versions of biomedical ontologies. The obtained results reveal the benefits of considering external sources of knowledge to identify the correct semantic relation in ontology evolution.