Energy consumption prediction by using an integrated multidimensional modeling approach and data mining techniques with Big Data

Abstract

During the past decades the resources have been used of an irresponsible and negligent manner. This has led to an increasing necessity of adopting more intelligent ways to manage the existing resources, specially the ones related to energy. In this regard, one of the main aims of this paper is to explore the opportunities of using ICT (Information and Communication Technologies) as an enabling technology to reduce energy use in cities. This paper presents a study in which we propose a multidimensional hybrid architecture that makes use of current energy data and external information to improve knowledge acquisition and allow managers to make better decisions. Our main goal is to make predictions about energy consumption based on energy data mining and supported by external knowledge. This external knowledge is represented by a torrent of information that, in many cases, is hidden across heterogeneous and unstructured data sources, which is recuperated by an Information Extraction system. This paper is complemented with a real case study that shows promising partial results.