Visualising high-resolution energy maps through the exploratory analysis of energy performance certificates

Abstract

This paper presents a new data mining engine, named EXTREMA (EXploitation of Turin high Resolution Energy MAps), to automatically visualise high-resolution energy maps exploring interesting and human-readable knowledge items from large collections of EPCs. EXTREMA, developed in Python, generates geo-located maps to summarise the main relationships among variables affecting the energy efficiency of buildings at different spatial granularity levels. The visualised knowledge is discovered through a two-level data analytics methodology based on exploratory and unsupervised algorithms. First an unsupervised algorithm divides EPCs into homogeneous groups of buildings with similar thermo-physical characteristics. Each group is then locally characterised through interesting patterns to concisely represent each group. The experimental evaluation, performed on a real dataset collected in a major Italian city in North-West Italy, demonstrates the effectiveness of EXTREMA in extracting and graphically display on geo-located multivariate energy maps a manageable set of human-readable knowledge items.