amalfi

The AMALFI project investigates graph learning algorithms for analyzing cryptoasset transaction graphs across Bitcoin, Ethereum, and Tron.

Objectives

• Collaborative Ground-Truth Attribution: Collect and organize attribution tags into a cryptoasset knowledge graph with semantic concepts and entity disambiguation.
• Graph Learning Methods: Develop approaches for DeFi service categorization, bot detection, and cross-chain entity discovery.
• Multi-Layer Graph Framework: Implement a framework treating each token type as a distinct layer for intra-layer and inter-layer analysis.