Google is based on the concept of the web as a gigantic directed graph. Edges in this graph are determined by directional hyperlinks between web pages. How many of us understand how the usage of rel and rev attributes on HTML hyperlinks affects this notion of the web as a graph? These attributes annotate the link (edge) between two pages (nodes) with a semantic value (a human relationship). Will this catch on? How will this change the way we think about internet topology?
Google, while not seeking to codify popular behavior or semantics, is bound to the primacy of authorship in the sense that whatever forms of content authors put on web pages, Google must parse and "understand". If we attach "human valued" attributes to relationships in web page markup, will this affect the value of the statistical information that Google currently generates?