Google’s First Great Idea: PageRank

Before Google, search engines didn’t have a great way to rank their results in order of importance. Instead most of them just found every single website containing your search terms and listed them in an arbitrary order. As you can imagine, this was not very user friendly. Search results were jammed full of insignificant and unreliable websites, making it hard to find anything useful. The innovation that shot Google to the top of the industry was their method of ranking websites in order of their importance. That method is called PageRank and it is still used today as one of the many signals Google considers when returning search results.

Videos:

• A practical discussion of the many elements of a seach engine and how PageRank fits in:
  • https://www.youtube.com/watch?v=v7n7wZhHJj8
• An introduction to the theory of Markov chains as it applies to PageRank:
  • https://www.youtube.com/watch?v=JGQe4kiPnrU

Additional Resources:

• Wikipedia’s page on the PageRank algorithm:
  • https://en.wikipedia.org/wiki/PageRank
• Google’s own webpage about how they rank search results now:
  • https://www.google.com/search/howsearchworks/how-search-works/ranking-results/

Technical Resources:

• The original paper describing PageRank and other aspects of Google’s operation (note that there is a typo in the PageRank formula in section 2.1.1. The (1-d) term should be divided by the number of websites that they are ranking):
  • Brin, S. & Page, L. The anatomy of a large-scale hypertextual Web search engine. Computer Networks and ISDN Systems 30, 107–117 (1998).
• Chapter about Markov chains on LibreTexts:
  • https://math.libretexts.org/Bookshelves/Applied_Mathematics/Applied_Finite_Mathematics_(Sekhon_and_Bloom)/10%3A_Markov_Chains/10.01%3A_Introduction_to_Markov_Chains