Making sense of spatial relationships trough local knowledge discovery in social media

Abstract

Over time, theories, models and methods have used scientific knowledge and structured data to model the relationships occurring across agents and places in a variety of domains. Recently, the diffusion of Social Networks is opening new research scenarios. Huge flows of data are being made available from Social Networks which contribute to the diffusion of local knowledge and to unravel several complex system dynamics. The present work aims at advancing the understanding of how spatial relationships embedded in natural language communications may help harness local knowledge in planning- and decision-making processes. To that purpose, we developed a pilot study on disaster response in the Metropolitan Area of Bari (Italy), by administering an on-line survey focussed on social media use in emergency situations. Main results suggest that the correct interpretation of local knowledge-laden natural language becomes a challenging problem when the role of tacit or implicit knowledge is taken into account. We argue in favour of acknowledging the importance of local knowledge for a full understanding of spatial relationships and calls for implementing spatial data science tools in such a way that local and tacit knowledge (including vernacular forms) may be adequately understood.