TY - CONF

T1 - Smart Water Management for Cities

AU - Kenda, Klemen

AU - Rizou, Stamatia

AU - Mellios, Nikos

AU - Kofinas, Dimitris

AU - Ritsos, Panagiotis D.

AU - Senozetnik, Matej

AU - Laspidou, Chrisy

PY - 2018/8

Y1 - 2018/8

N2 - The deployment of real-world water monitoring and analytics tools is still far behind the growing needs of cities, which are facing constant urbanisation and overgrowth of the population. This paper presents a full-stack data-mining infrastructure for smart water management for cities being developed within Water4Cities project. The stack is tested in two use cases - Greek island of Skiathos and Slovenian capital Ljubljana, each facing its own challenges related to groundwater. Bottom layer of the platform provides data gathering and provision infrastructure based on IoT standards. The layer is enriched with a dedicated missing data imputation infrastructure, which supports coherent analysis of long-term impacts of urbanisation and population growth on groundwater reserves. Data-driven approach to groundwater levels analysis, which is important for decision support in flood and groundwater management, has shown promising results and could replace or complement traditional process-driven models. Data visualization capabilities of the platform expose powerful synergies with data mining and contribute significantly to the design of future decision support systems in water management for cities.

AB - The deployment of real-world water monitoring and analytics tools is still far behind the growing needs of cities, which are facing constant urbanisation and overgrowth of the population. This paper presents a full-stack data-mining infrastructure for smart water management for cities being developed within Water4Cities project. The stack is tested in two use cases - Greek island of Skiathos and Slovenian capital Ljubljana, each facing its own challenges related to groundwater. Bottom layer of the platform provides data gathering and provision infrastructure based on IoT standards. The layer is enriched with a dedicated missing data imputation infrastructure, which supports coherent analysis of long-term impacts of urbanisation and population growth on groundwater reserves. Data-driven approach to groundwater levels analysis, which is important for decision support in flood and groundwater management, has shown promising results and could replace or complement traditional process-driven models. Data visualization capabilities of the platform expose powerful synergies with data mining and contribute significantly to the design of future decision support systems in water management for cities.

UR - https://ai4good.org/kdd-2018-workshop/

M3 - Other

T2 - ACM SIGKDD Conference on Knowledge Discovery and Data Mining

Y2 - 20 August 2018

ER -