Participatory Socio-Environmental Systems Modeling over Knowledge Graphs

By considering the dynamicity and complexity that is present in modern Socio-Environmental Systems along with the misalignment in the usage of terms that are used by scientists in different disciplines, there is a need to support new ways of modelling and analysis based on the establishment of synergies and the collaboration among scientists. To address this challenge, we present a novel paradigm for modeling of Socio-Environmental Systems that aims to enable interdisciplinary scientists to realise participatory, reproducible and easily extensible modelling and analysis, across different temporal and spatial scales. To conceptualize the overall paradigm, emerging technologies for knowledge management and analysis are exploited, such as Knowledge Graphs and Machine Learning techniques. Knowledge Graphs are used as a variant of a semantic network, where constraints, structural elements and characteristics of nodes and links are continuously evolving. Machine Learning techniques are used for populating the Knowledge Graph with different types of data, as well as supporting analysis and inference processes over the produced knowledge. A proof of concept scenario has been implemented, focusing on the tracking of indicators specified in the Sustainable Development Goals.