But, all the updates that I've observed over the last decade have actually just been improved implementations of the basic principles that I learned a decade ago.
|With Penguin 2.0, Google understands that Elton John is a singer|
1. Google tracks citations about things and enters them in the algorithm even if there is no link.
2. Semantics and intelligence. This has been an area of real progress for Google with several stages.
In the old days when I started, Google was a dumb word counting system. When I started, Google produced very different results when you searched for "1st grade" vs "first grade". By about five years ago, Google seemed to understand that these were exact synonyms and produced identical results.
Google also tried to infuse its results with appropriate information for the search. Search on "apple pie" and you would get recipes with the ingredients summarized right in the search engine. Search for a specific model of a piece of consumer electronics and the price was highlighted in the search engine. These were early implementations of Google's use of schema to provide more info right at the point of search. This has been enhanced with rich snippets.
Today, it sounds like Google's intelligence is jumping up another few notches as it develops an understanding of different sorts of entities and associations. I've got some research and thinking to do to try to understand how this might affect us.