Linking Cryptoasset Attribution Tags to Knowledge Graph Entities: An LLM-based Approach

2025-02-12

AI Summary

To address inconsistent and misleading attribution labels in cryptocurrency asset tracking, this paper proposes the first unsupervised label–knowledge graph entity alignment method leveraging large language models (LLMs). The method requires no labeled data and integrates concept filtering with blocking strategies to generate high-precision candidate alignments and mappings. Through prompt engineering and cost-aware LLM selection, it supports flexible deployment across local or remote LLMs. Evaluated on three public datasets, the approach achieves up to a 37.4% improvement in F1-score and a 93% top-5 recall rate. Local LLM inference attains 90% F1 (vs. 94% for remote), while optimal configuration reduces computational cost by 90% with only a 1% performance degradation—demonstrating the feasibility and practicality of lightweight LLMs for on-chain forensic analysis.

Technology Category

• Data Mining & Knowledge Management: Linked Open Data, Knowledge Graphs & KB Completion
• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Fact-Checking / Misinformation Detection (NLP Focus)

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Economics, Online Markets and Human Computation: Humans versus LLMs for data annotation and labeling

Abstract

Attribution tags form the foundation of modern cryptoasset forensics. However, inconsistent or incorrect tags can mislead investigations and even result in false accusations. To address this issue, we propose a novel computational method based on Large Language Models (LLMs) to link attribution tags with well-defined knowledge graph concepts. We implemented this method in an end-to-end pipeline and conducted experiments showing that our approach outperforms baseline methods by up to 37.4% in F1-score across three publicly available attribution tag datasets. By integrating concept filtering and blocking procedures, we generate candidate sets containing five knowledge graph entities, achieving a recall of 93% without the need for labeled data. Additionally, we demonstrate that local LLM models can achieve F1-scores of 90%, comparable to remote models which achieve 94%. We also analyze the cost-performance trade-offs of various LLMs and prompt templates, showing that selecting the most cost-effective configuration can reduce costs by 90%, with only a 1% decrease in performance. Our method not only enhances attribution tag quality but also serves as a blueprint for fostering more reliable forensic evidence.

Problem

• Linking cryptoasset attribution tags to knowledge graph entities.
• Improving accuracy and consistency in cryptoasset forensic investigations.
• Developing a cost-effective LLM-based method for forensic evidence enhancement.

Innovation

• LLMs link crypto tags to knowledge graphs
• End-to-end pipeline improves F1-score by 37.4%
• Local LLMs match remote models' performance