Product classification mappings to BONSAI climate footprint database

Zenodo (CERN European Organization for Nuclear Research)

Abstract

This dataset is curated and developed by bonsai.uno, a not-for-profit association providing open, transparent sustainability data to support research, business and policy settings. It provides mappings between product classifications used in external statistical and life cycle inventory data sources and the product classification used by BONSAI database. The dataset is intended to support the integration of heterogeneous datasets into BONSAI, and provide training and evaluation data for automated classification-linking methods. It contains more than 10,000 manually developed or reviewed correspondence records linking source codes and names to BONSAI product categories. Covered source classifications include HS2007, CPA, CPA 2.1, FAOSTAT, EXIOBASE, IEA, JRC-IDEES, UN Energy Statistics and several domain-specific classifications. Each record identifies the source classification (optional), the source code or label, and the corresponding BONSAI product name. Where available, provenance information such as the source reference and creation date is also included. Potential applications include: training machine-learning or language-model-based classification matching systems; evaluating automated product-mapping approaches; linking statistical, trade, energy, agricultural and life cycle inventory datasets to BONSAI; The mappings may be many-to-one and one-to-many: a single BONSAI product can correspond to multiple source products, and a single source product can also be mapped to multiple BONSAI products. This reflects the fact that the BONSAI classification is designed to harmonize information from sources with different levels of detail and different classification principles. The dataset should therefore be interpreted as a curated correspondence table rather than a strict equivalence or concordance between classifications. By publishing these mappings openly, the aim is to make the BONSAI data integration process more transparent and to enable external researchers and developers to build tools that can connect new datasets to BONSAI.

Authors 2

1. Fan Yang
   Affiliation as printed

   2-0 LCA

2. Stefano Merciai Aachen

   Leiden University

   Affiliation as printed

   Leiden University

   2-0 LCA