Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Detection and Analysis of Sensitive and Illegal Content on the Ethereum Blockchain Using Machine Learning Techniques

Xingyu Feng

TL;DR

<3-5 sentence high-level summary> The Ethereum blockchain enables embedding of arbitrary data in the input field, creating privacy and security risks when illegal or sensitive content is stored on-chain. The authors develop a data identification and restoration framework that decodes, reconstructs, and analyzes embedded data from ~3.4 billion transactions, applying UTF-8 decoding, file-signature-based restoration, and multi-transaction synthesis; text sentiment is assessed with FastText and images are screened with NSFWJS. They recover 175 common file types, 296 images, and 91,206 texts, uncovering pornographic content, sensitive personal data, and hate speech, and they construct an Information Embedding Network to characterize embedding patterns. The work provides empirical evidence of on-chain content diversity and harm, proposes preventive measures, and offers methodological foundations for blockchain privacy and regulatory considerations.

Abstract

Blockchain technology, lauded for its transparent and immutable nature, introduces a novel trust model. However, its decentralized structure raises concerns about potential inclusion of malicious or illegal content. This study focuses on Ethereum, presenting a data identification and restoration algorithm. Successfully recovering 175 common files, 296 images, and 91,206 texts, we employed the FastText algorithm for sentiment analysis, achieving a 0.9 accuracy after parameter tuning. Classification revealed 70,189 neutral, 5,208 positive, and 15,810 negative texts, aiding in identifying sensitive or illicit information. Leveraging the NSFWJS library, we detected seven indecent images with 100% accuracy. Our findings expose the coexistence of benign and harmful content on the Ethereum blockchain, including personal data, explicit images, divisive language, and racial discrimination. Notably, sensitive information targeted Chinese government officials. Proposing preventative measures, our study offers valuable insights for public comprehension of blockchain technology and regulatory agency guidance. The algorithms employed present innovative solutions to address blockchain data privacy and security concerns.

Detection and Analysis of Sensitive and Illegal Content on the Ethereum Blockchain Using Machine Learning Techniques

TL;DR

<3-5 sentence high-level summary> The Ethereum blockchain enables embedding of arbitrary data in the input field, creating privacy and security risks when illegal or sensitive content is stored on-chain. The authors develop a data identification and restoration framework that decodes, reconstructs, and analyzes embedded data from ~3.4 billion transactions, applying UTF-8 decoding, file-signature-based restoration, and multi-transaction synthesis; text sentiment is assessed with FastText and images are screened with NSFWJS. They recover 175 common file types, 296 images, and 91,206 texts, uncovering pornographic content, sensitive personal data, and hate speech, and they construct an Information Embedding Network to characterize embedding patterns. The work provides empirical evidence of on-chain content diversity and harm, proposes preventive measures, and offers methodological foundations for blockchain privacy and regulatory considerations.

Abstract

Blockchain technology, lauded for its transparent and immutable nature, introduces a novel trust model. However, its decentralized structure raises concerns about potential inclusion of malicious or illegal content. This study focuses on Ethereum, presenting a data identification and restoration algorithm. Successfully recovering 175 common files, 296 images, and 91,206 texts, we employed the FastText algorithm for sentiment analysis, achieving a 0.9 accuracy after parameter tuning. Classification revealed 70,189 neutral, 5,208 positive, and 15,810 negative texts, aiding in identifying sensitive or illicit information. Leveraging the NSFWJS library, we detected seven indecent images with 100% accuracy. Our findings expose the coexistence of benign and harmful content on the Ethereum blockchain, including personal data, explicit images, divisive language, and racial discrimination. Notably, sensitive information targeted Chinese government officials. Proposing preventative measures, our study offers valuable insights for public comprehension of blockchain technology and regulatory agency guidance. The algorithms employed present innovative solutions to address blockchain data privacy and security concerns.

Paper Structure

This paper contains 27 sections, 1 equation, 9 figures, 6 tables, 3 algorithms.

Table of Contents

  1. Introduction
  2. Background
  3. Ethereum and smart contracts
  4. Ethereum blockchain transaction fields
  5. Data collection
  6. Data restoration algorithm
  7. File restoration algorithm
  8. Multi-transaction-based file segmentation and embedding restoration algorithm
  9. Data embedding analysis
  10. File data analysis
  11. Image file data analysis
  12. Other types of file analysis
  13. File segmentation embedding analysis
  14. Text data analysis
  15. Link type text analysis
  16. ...and 12 more sections

Figures (9)

  • Figure 1: A framework for embedding restoration and analysis of data in the Ethereum blockchain network.
  • Figure 2: The file types and quantities are restored
  • Figure 3: Number of embedded files per quarter (2015-2021).
  • Figure 4: Incomplete image restoration using file segmentation embedded restore algorithm (Algorithm\ref{['alg:file-Split-restore']}), (a) incomplete image file found, and (b) image file restored using Algorithm\ref{['alg:file-Split-restore']}. The image file is split 5 times for embedding, and the top 5 bits of the hash value of each transaction are marked in the figure.
  • Figure 5: Number of embedded texts per 1 million blocks in the Ethernet blockchain.
  • ...and 4 more figures