Copy the page URI to the clipboard
Markovic, Milan; Garijo, Daniel; Germano, Stefano and Naja, Iman
(2023).
DOI: https://doi.org/10.1007/978-3-031-47243-5_5
Abstract
Greenhouse gas emissions have become a common means for determining the carbon footprint of any commercial activity, ranging from booking a trip or manufacturing a product to training a machine learning model. However, calculating the amount of emissions associated with these activities can be a difficult task, involving estimations of energy used and considerations of location and time period. In this paper, we introduce the Transparent Emissions Calculation (TEC) toolkit, an open source effort aimed at addressing this challenge. Our contributions include two ontologies (ECFO and PECO) that represent emissions conversion factors and the provenance traces of carbon emissions calculations (respectively), a public knowledge graph with thousands of conversion factors (with their corresponding YARRRML and RML mappings) and a prototype carbon emissions calculator which uses our knowledge graph to produce a transparent emissions report.