DataNet: linking multi-faceted building energy datasets for effective performance analysis at scale

초록

Recently, several administrative registration datasets in the Republic of Korea have become public or partially accessible. These include building registries, monthly energy billing data, building footprint polygons, business registration records, datasets on facilities maintained by ministries, and weather information. However, from the perspective of building energy analysis, these datasets remain fragmented because they are by-products of administrative processes not originally intended for integration. To achieve carbon neutrality in the building sector, comprehensive understanding of energy consumption patterns is essential—requiring integration of diverse contextual factors from climate and building characteristics to operational patterns and socioeconomic conditions. This paper introduces the DataNet project, which aims to collect and link these multifaceted building energy datasets at a national scale to enable evidence-based identification of inefficient buildings and support targeted interventions across Korea's stock of 7.3 million buildings. In this project, the building registry serves as the reference dataset, and other datasets are linked via addresses and associated attributes, such as building or business names. This paper outlines the overall data framework used to identify, categorize, and integrate multiple datasets with a macro-level temporal resolution at a national scale. A summary and overview of 13 out of the 17 currently available datasets are presented. The overall matching rate of these 13 datasets, based on the building registry, is 95%. Finally, the paper discusses potential applications of this integrated data framework for evidence-based performance analysis to accelerate the decarbonization of the building stock. Copyright © 2026. Published by Elsevier B.V.

키워드