Contrasting Knowledge Organization Systems for the Description of Research Products: the Case of Overlapping in the Agricultural Domain

The use of Knowledge Organization Systems (KOS) as ontologies or terminologies for the description of scholarly contents requires a careful consideration of the domain and the KOS available. KOS in the same domain may differ in several dimensions including purpose, level of formality, structure and language. In consequence, curators of scientific data face the problem of selecting the relevant KOS, developing mappings when appropriate and deciding on their usage for annotating resources. In domains in which more than a KOS is available, curators need tools to help them in the decision making process. Due to the available heterogeneity of KOS, exploratory tools are required for an initial assessment of overlapping and differences. This paper reports on a practical experience using simple mapping analysis and mapping visualizations in the domain of agriculture. These techniques represent promising directions for the development of decision tools based on the contrast of different KOS metrics.