A Machine Learning-Based Method for Automated Blockchain Transaction Signing Including Personalized Anomaly...

Area of Science:

• Computer Science
• Cryptography
• Machine Learning

Background:

• Blockchain technology relies on digitally signed transactions for data integrity on distributed ledgers.
• Current digital signing processes are time-consuming and hinder widespread blockchain adoption.

Purpose of the Study:

• To develop a machine learning-based method for automated blockchain transaction signing.
• To incorporate personalized identification of anomalous transactions within the signing process.

Main Methods:

• A novel machine learning algorithm was designed for automated transaction signing.
• The method was evaluated using data from the Ethereum public main network.
• Anomaly detection was integrated into the personalized signing workflow.

Main Results:

• The proposed method demonstrated promising results in automating blockchain transaction signing.
• The system effectively identified anomalous transactions, enhancing security.
• Experimental analysis on Ethereum data validated the method's efficacy.

Conclusions:

• The machine learning approach offers a user-friendly and efficient solution for blockchain transaction signing.
• This method can significantly improve the adoption and security of blockchain technology.
• Future integration into digital signing software is a viable next step.