The energy signature is a widely adopted method for assessing the energy performance of buildings, traditionally relying on a linear correlation between energy consumption and outside air temperature. While useful, this model has significant limitations as it fails to account for other critical environmental and operational factors. This thesis presents a significant extension of the traditional energy signature by developing a multidimensional model that incorporates a broader range of variables, including humidity, wind speed, solar radiation, and building occupancy patterns. Leveraging datasets from seven different building locations across five Italian cities (Bologna, Florence, Genova, Milan, and Turin) over two winter seasons (2022-2023 and 2023-2024), this research employs a systematic, data-driven approach.