Private Information Inference of Households from Electricity Consumption Data

Abstract

The spread of smart home technologies not only brings convenience but also creates various security and privacy concerns among users. Electricity consumption data collected by smart meters is one of the sources of these concerns. The electricity consumption of the appliances working at home made it possible to have information about the private life of the household. This study is aimed to reveal a classification model by using the electricity consumption data obtained as a result of the study conducted in Ireland and the results of the survey study conducted with the households. While the first method in the study aims to access information about private life directly with electricity consumption data, the second method uses the predictions of one private information to improve the results of the prediction of another related information. As a result, it has been concluded that electricity consumption data can be used in the process of obtaining information about private life, and that the use of relationship between two information leads to an improvement in model performance. This study shows one of the obstacles that may occur in the spread of smart houses and has prepared the environment for studies that can be done on the subject of solution.