Work

Enhancing Dark Web Classification: A Dynamic Crawler and Robust Classification Framework

Year: 2023
Type: article
Abstract: The dark web presents significant challenges for law enforcement agencies due to its anonymous and constantly evolving nature, making it difficult to trace and monitor illegal activities. To combat this issue, a proposed system collects and cleans dark web pages, focusing on studying the market that specializes in selling illegal and harmful products through three key modules. The Crawler module accesses the market through the Tor Network, gathers data, and extracts crucial information about the products, sellers, and prices. The Pre-Processing module cleans and organizes the extracted data, ensuring its integrity and transforming it into a mineable format. The Data Mining module extracts insights and knowledge from the processed data, using techniques like clustering, classification, and association rule mining to identify patterns and trends. These modules provide valuable insights to law enforcement agencies and security researchers to combat illicit activities on the dark web market. The system's efficiency is evaluated through metrics like throughput and speedup, demonstrating its capability to handle large datasets and improve performance through parallel processing. Additionally, the proposed Support Vector Machine (SVM) with Neural Network (NN) outperforms other methodologies, highlighting its accuracy in predicting dark web links and establishing a robust classification framework. The research contributes to a comprehensive understanding of the dark web landscape and fosters advancements in cybersecurity and law enforcement practices. By integrating these solutions, this study aims to enhance the accuracy, adaptability, and effectiveness of dark web classification. The research contributes to a comprehensive understanding of the dark web landscape and fosters advancements in cybersecurity and law enforcement practices.
Language: English
Subfield: Law
Open Access status: diamond
OpenAlex ID: W7218233318
FWCI: 1.205
Cites: 17
Cited by: 2
Related to: 9