Capturing the relationship between evolving biomedical concepts via background knowledge

Auteurs

C. Pruski, C. J. Dos Reis, and M. Da Silveira

Référence

in 9th International SWAT4LS Conference: Semantic web applications and tools for the life sciences, December 5-8th, 2016 Amsterdam, The Netherlands, 2016

Description

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 map-pings 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.