RIBEIRO, I. P.; http://lattes.cnpq.br/1437042517615961; RIBEIRO, Iara Pereira.
Resumo:
The recent increase in demand for resources, and the imminent potential shortage of these has created a new kind of societal concern which spawned an emphasis for more efficient methods on how to use these resources. One resource, in particular, is electricity and the glaring concern for how it is consumed; mainly due to the use of non-renewable way for generating electricity, E.G. thermal power using coal. Currently, in Brazil, 73.1% of the country’s energy is generated from renewable sources. Other factors such as climate change and extended periods of drought may impact the total amount of energy being generated, thus making the use of alternative methods for power generation a necessity – which in turn inflates the costs for the consumer. Within this context comes the need to develop tools and ideas which help to make the consumption of energy more efficient by reducing the production of electricity which will be beneficial to both the dealers and end consumers. One option to solve this problem would be to focus on the consumption in residential areas, as in Brazil, the residential sector is the third largest consumer of energy, consuming on average 24.78% of the total power generated in the country. This paper proposes a solution which uses mapping between energy efficient concepts and concepts of recommender systems to help promote the reduction of electrical consumption. The proposed algorithms combined with Collaborative Filtering and Content has used the processed data from behavioral surveys among volunteers, data government and software to stimulate the residential electricity consumption. From the results, we can conclude that with this relatively new ambit of discovery comes many concepts yet to be explored in the use of data analysis techniques for energy efficiency, and the importance of the application to future work.