Some have written to me to include a Google-like Page Ranking to the Content Classification Engine I am writing. Great suggestion! Imagine what it could to research and knowledge databases. A most cited paper automatically would become the most authoritative reference!
To explore this idea, I conducted a small experiment.
Donald Knuth's masterpiece
on computer programming, The Art of Computer Programming is the single most cited
reference in Computer Science. So any search for "fundamental algorithms"
should point to this work. Right? You will be amazed on what I discovered:
• Google - Top match. Dead on! Google, don't we just love you!
• EBSCO - Pointed me to a CAD Algorithm
• Questia - found nothing (and this is a great student resource?)
• Ingenta - found an article on industrial scheduling
• Northern Light - 2nd item (pretty good) in search results
Notice how I searched generic web search engines (meant for common man), as well as the so called scholarly sources. While you'd think a scholary source like a library or a research database gives you a better search result, it was Google and Northern Light that gave me most relevant pointers to further study.
This indicates that Google is a far superior research (or discovery as the industry calls it) tool than most people realize(I often use Google to search even MSDN), and the intelligence of the tool can be traced to its fundamental Page Ranking algorithm.
