A blockchain-based trust-free building sustainability performance assessment system using machine learning models trained on expert knowledge

Date

2026-6-24

Author

Sayın, Melis

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

52
views

0
downloads

Cite This

Considering the construction industry's significant environmental impact, the sustainability performance of construction projects and green buildings has become increasingly important for reducing the environmental burden of the built environment and supporting sustainable development. However, conventional green building certification processes require significant time, cost, and reliance on certification institutions or accredited professionals as third parties. Although the assessment criteria used in these processes are publicly available, they require high-level technical knowledge, mathematical formulations, and expert-level domain knowledge, which limits their applicability, especially for smaller-scale projects. Moreover, since green certification systems are used as marketing tools and their calculation processes are publicly available, project design decisions can be biased toward obtaining specific certification levels rather than achieving the highest possible sustainability performance. To address these limitations, this thesis develops a blockchain-based trust-free building sustainability performance assessment system that enables stakeholders to evaluate their projects on an objective, transparent, and immutable platform without prior domain knowledge or the need for a third party. For the Sustainability Index Assessment System, detailed sustainability performance indicators are constructed by taking the LEED certification system as a basis. A dataset of 32 real-world projects is constructed, and expert knowledge is incorporated through collected expert opinions. Different machine learning models are trained using collected expert evaluations, and the final model is selected based on comparative prediction performance. Finally, the selected model is integrated into a verified decentralized smart contract and implemented through a web-based user interface. The resulting system provides transparent, trust-free, and verifiable sustainability performance assessments.

Subject Keywords

Blockchain technologies, Smart contracts, Sustainability assessment, Machine learning, Green building certification

URI

https://hdl.handle.net/11511/119774

Collections

Graduate School of Natural and Applied Sciences, Thesis