joined:Jan 9, 2002
Researchers at the NEC Research Institute at Princeton have developed a novel search engine algorithm
[nature.com]. Like Google, it relies on links to provide relevance, but unlike Google, it pays little attention to text. Instead, it looks for pages that refer back to each other. This strikes me as something that would fall apart the minute a SE spammer got to it, but they claim they've found interesting sites missed by Google and Yahoo.