Exploring Patterns in Water Consumption by Clustering. (2015)

• Record Type: Journal Article
• Authors: Laspidou, Chrysi; Papageorgiou, Elpiniki; Kokkinos, Konstantinos; Sahu, Sambit; Gupta, Arpit; Tassiulas, Leandros
• Is Part Of: Procedia engineering. Volume 119(2015)
• Page Start: 1439
• Page End: 1446
• Publication Date: 2015
• Subjects: Kohonen Self-Organized Maps (SOM) -- urban water demand -- clustering algorithms -- data mining -- water consumption analysis; Engineering -- Congresses; Engineering -- Periodicals; Engineering; Conference proceedings; Periodicals; 620.005
• DOI: 10.1016/j.proeng.2015.08.1004
• Languages: English
• ISSNs: 1877-7058

Abstract

Abstract: Water scarcity, high water demand due to increasing urbanization and the ongoing liberalization of the water and energy markets makes water utilities look into innovative ways to approach consumers, to offer attractive plans and educate them by raising their awareness of their resource use. We analyze water consumption data from a group of consumers at the Greek island of Skiathos, for which we have additional information about features concerning their water consumption patterns. These features are used as input vectors for the construction of Kohonen Self-Organized Maps that are used as classification methods to cluster consumers according to their water consumption. Results show that such analysis can be promising for the automatic classification of water consumers, based on urban water demand data, even if the data is not real-time, or even frequent, since consumptions from standard quarterly water bills are used.