Publication Alert: New Benchmark Dataset for Greenhouse Gas Emissions

New Benchmark Dataset for Greenhouse Gas Emissions: Addressing Data Gaps in Sustainability Reporting

Large companies in the EU are legally required to report their greenhouse gas (GHG) emissions. But extracting this information from lengthy PDF sustainability reports is a time-consuming challenge for researchers.
Our new paper published in Scientific Data introduces a gold standard benchmark dataset for GHG emission extraction. The dataset is based on 139 sustainability reports from companies in the MSCI World Small Cap index and the German DAX.

Key insights include:

• About half of the reports contained no usable GHG data, highlighting major reporting gaps.

• Where emissions were reported, Scope 1 (direct) and Scope 2 (indirect from energy use) were most common. Scope 3 emissions (supply chain, travel, etc.) were rarely complete.

• Even expert annotators disagreed frequently, underlining the complexity of extracting reliable GHG metrics

This dataset provides researchers with a transparent and rigorously curated foundation to evaluate automated approaches for sustainability reporting.

Link to the paper: https://rdcu.be/eCEyr