Abstract
The rapid growth of anonymous communication technologies has led to the emergenceof the dark web as a significant component of the internet ecosystem. Unlike the surfaceweb, dark web content is hosted on hidden services accessible through anonymitynetworks such as The Onion Router (Tor). The dynamic and decentralized nature of darkweb resources presents substantial challenges for information retrieval systems,particularly search engines that rely on crawler programs for content discovery andindexing. This study presents a quantitative comparative analysis of crawler programsutilized by three prominent dark web search engines: Ahmia, Torch, and Haystak. Theresearch investigates crawler efficiency, indexing coverage, retrieval accuracy, latency,and usability through experimental benchmarking and document analysis. A sample of1,000 indexed .onion websites was examined to evaluate crawler performance under controlled conditions. Findings indicate significant differences among the search enginesin terms of indexing breadth, response speed, and retrieval precision. Haystakdemonstrated the highest indexing coverage, Torch exhibited superior crawler reachacross active hidden services, and Ahmia achieved the highest retrieval accuracy andusability scores. The study contributes to understanding dark web information retrievalmechanisms and provides insights for researchers, cybersecurity practitioners, anddevelopers seeking to improve dark web search technologies.Keywords: Dark Web, Tor Network, Web Crawlers, Search Engines, Ahmia, Torch,Haystak, Information Retrieval, Benchmarking
Showing the abstract — retrieve the full paper via the Exa API.