EUVIMEDEUVIMED entdecken
Crossref

Explainable Detection of Proxy Smart Contracts Using Ensemble Learning and SHAP Analysis

Iis Setiawan

Journal of Digital Market and Digital Currency · 02.11.2026 · 3 (4) · 334-353

DOI: 10.47738/jdmdc.v3i4.105 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Digital Market and Digital Currency, online erschienen am 02.11.2026. Verfasst von Iis Setiawan.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Proxy smart contracts have become a fundamental component of modern blockchain applications because they enable contract upgradeability without changing deployed contract addresses. However, identifying proxy contracts manually is challenging due to the complexity of Solidity source code and the diversity of proxy implementation standards. This study proposes an explainable ensemble learning framework for detecting proxy smart contracts using Solidity source code analysis. The proposed approach employs Term Frequency–Inverse Document Frequency (TF-IDF) to extract textual features from smart contract source code and evaluates three ensemble learning models, namely Random Forest, XGBoost, and LightGBM. Furthermore, SHapley Additive exPlanations (SHAP) is utilized to improve model interpretability and identify the most influential features contributing to proxy contract detection. Experiments were conducted on a dataset containing 691 Ethereum smart contracts, including 23 proxy contracts and 668 non-proxy contracts. The results show that Random Forest achieved the best overall performance with an accuracy of 98.56%, an F1-score of 0.800, and a ROC-AUC of 0.999. SHAP analysis revealed that features related to upgradeable smart contract architectures, such as the beacon, ibeacon, erc1967proxy, and implementation, were the most significant indicators of proxy contract behavior. The findings demonstrate that the proposed framework can effectively detect proxy smart contracts while providing interpretable insights into the architectural patterns underlying upgradeable blockchain applications.

Zitieren

APA

Setiawan, I. (2026). Explainable Detection of Proxy Smart Contracts Using Ensemble Learning and SHAP Analysis. Journal of Digital Market and Digital Currency, 3(4), 334-353. https://doi.org/10.47738/jdmdc.v3i4.105

Vancouver

Setiawan I. Explainable Detection of Proxy Smart Contracts Using Ensemble Learning and SHAP Analysis. Journal of Digital Market and Digital Currency. 2026;3(4):334-353. doi:10.47738/jdmdc.v3i4.105

BibTeX
@article{Setiawan2026,
  author = {Setiawan, Iis},
  title = {Explainable Detection of Proxy Smart Contracts Using Ensemble Learning and SHAP Analysis},
  journal = {Journal of Digital Market and Digital Currency},
  year = {2026},
  volume = {3},
  number = {4},
  pages = {334--353},
  doi = {10.47738/jdmdc.v3i4.105}
}
Kostenlos starten →