Initial design process of the sustainability science ontology for knowledge-sharing to support co-deliberation

Implementation of the sustainability science (SS) approach is often difficult because of poor communication between experts from different academic fields. We focused on ontology engineering as a method of knowledge structuring that supports the co-deliberation process. However, SS is too broad for a few experts to construct an ontology because SS targets and covers almost all existing research fields from the viewpoint of problem-solving. The N-iteration process is required for completing an SS ontology. In the present paper, we discuss the initial design process for constructing an ontology on SS from the aspect of a knowledge-sharing tool to support co-deliberation. First, we identified the SS ontology by referring to the existing literature. Second, we traced the structuring process of the SS ontology, which is independent of the existing research domain. Third, we compared the SS ontology with existing ontologies or concept structures on SS. Fourth, we assessed the SS ontology produced in the initial process in terms of relevance and coverage and addressed areas for improvement in order to facilitate co-deliberation among researchers from different domains. As a result of developing the SS ontology and applying it to the mapping tool that we developed based on the ontology, we found the following three points: the SS ontology enables us to define concepts relevant to SS without overlapping by distinguishing part-of and attribute-of relationships at the upper level of the ontology; the SS-based mapping tool successfully represents the potential countermeasures required by the targeted problem for all scientific fields except experimental engineering; however, the SS ontology requires further improvement in order to represent the conceptual linkage arising from compound and secondary problems and the fulfillment of classes at the lower hierarchy of Shortage problem, and requires slots for the entire hierarchy. In addition, based on the discussion of the areas for improvement, we found that missing slots and classes should be added in the process in which we use or improve tools corresponding to a variety of requirements for supporting co-deliberation. In this way, we are able to propose an incremental process for constructing the SS ontology from the aspect of a knowledge-sharing tool to support co-deliberation.