A few days ago a new search engine DarkSearch for Tor launched, adding to the mix of other existing search engines out there like Ahmia, Torch, Not Evil, and Haystack – it’s time for a feature comparison!
No search engine can cover 100% of the pages due to the nature of Tor. There is no central .onion repository so the first challenge is to find the .onion links. Other challenges when running a search engine include data size (and associated storage and processing power), data formats, and many smaller challenges like depth of crawling (i.e. how many sub-pages, how to behave when there are infinite sup-pages).
The following graph shows our index of Tor (dark blue) and I2P (light blue). As of April 2019, we have 10,197,379 items indexed for Tor and 1,557,915 items for I2P. An item can be any supported file format – including HTML, text, PDF, office documents (Word, Excel, and PowerPoint files), and since yesterday, even eBooks.
We have 2,250,020 .onion addresses in our index, although only a small fraction is actually active. For I2P our index has 3,565 .i2p domains listed.
We added support for US Social Security Numbers (SSNs)! You can immediately search for them here: https://intelx.io/?s=086-38-5955 Whis this is important Searching at Intelligence X works based on selectors (strong search terms). When you search for something, the system automatically detects suitable selectors and performs a search. Below is the list of supported selectors: Email
We have just turned on email verification for all accounts on intelx.io. Existing accounts receive a verification link via email upon login.
We just added native support to Intelligence X for the following data formats: Excel Sheets (.XLS, .XLSX, .ODT) PowerPoint Presentations (.PPTX) Native support means end-to-end support. This ranges from indexing and crawling files of various data sources, to processing them internally and presenting them to the end-user on the frontend intelx.io. Indexing is the process