Power-consumption analysis through web-based visual data exploration

Abstract

In recent years, the increasing capabilities of current technologies have made the acquisition and storage of large datasets an easy task. However, most times their unmanageable size, along with their complexity makes it a challenge to handle and analyze them. The emerging field of visual analytics relies on visualization principles and interactive interfaces to provide ways for amplifying human cognition, so that the information processing is improved and a better understanding of these datasets is achieved. In addition, current technologies to develop visual interfaces allow to develop powerful interaction mechanisms that enable the user to be an active part in the process. In this paper, data visualization foundations and web-based methods are considered for user-driven supervision tasks in decision-support systems. A real case consisting in the analysis of electric power demand in university buildings is presented via a web application developed using the recent data visualization library called D3. Time-series visualization and similarity maps are represented along with interactive techniques, allowing a dynamic data exploration.