Chen, B., Chen, H., Quan, Z. et al. (4 more authors) (2025) SemaNet: Bridging words and numbers for predicting missing environmental data in life cycle assessment. Environmental Science & Technology. ISSN: 0013-936X
Abstract
Life Cycle Assessment (LCA) is one of the most used methodologies for evaluating environmental impact, but its effective application is severely limited by missing data, an issue that existing methods for Life Cycle Inventory (LCI) data completion cannot address effectively. This paper proposes a paradigm shift: rather than depending exclusively on numerical correlations, we leverage the extensive contextual information inherent in process descriptions via pretrained language models, establishing a semantic bridge between qualitative descriptions and quantitative environmental flows. Our semantic-based neural network framework, SemaNet, achieves superior performance in predicting missing LCI values, surpassing existing state-of-the-art methods in various evaluation metrics. The results are significant: while existing approaches fail completely under high data sparsity, our method achieves high accuracy even with 100% missing numerical data while reducing computational requirements by 99% through the use of semantic filtering. This new method for LCI data completion significantly reduces the data collection efforts and time for LCA practitioners, making reliable and faster environmental impact assessment feasible, even when primary data does not exist, thus facilitating reliable sustainability assessment across industrial sectors.
Metadata
Item Type: | Article |
---|---|
Authors/Creators: |
|
Copyright, Publisher and Additional Information: | © 2025 American Chemical Society. |
Keywords: | data efficiency; life cycle assessment; life cycle inventory; neural network method; semantic learning |
Dates: |
|
Institution: | The University of Sheffield |
Academic Units: | The University of Sheffield > Faculty of Science (Sheffield) > School of Mathematical and Physical Sciences The University of Sheffield > University of Sheffield Research Centres and Institutes > AMRC with Boeing (Sheffield) The University of Sheffield > Advanced Manufacturing Institute (Sheffield) > AMRC with Boeing (Sheffield) |
Funding Information: | Funder Grant number MANCHESTER PRIZE UNSPECIFIED |
Date Deposited: | 06 Oct 2025 08:05 |
Last Modified: | 06 Oct 2025 08:05 |
Status: | Published online |
Publisher: | American Chemical Society (ACS) |
Refereed: | Yes |
Identification Number: | 10.1021/acs.est.5c07557 |
Related URLs: | |
Open Archives Initiative ID (OAI ID): | oai:eprints.whiterose.ac.uk:232552 |