Behind The Scenes At Google
Very interesting lecture given by Jeff Dean of Google at the University of Washington. Jeff talks about the Google infrastructure and how they handle and serve queries.
Interesting points covered:
While PageRank is query independent, Jeff mentions it is used in calculating results: "High page rank is better than low page rank". Hmmm. He doesn't mention the weighting of ranking factors, however.
He also notes that PageRank is used to decide which shards should have more replications (i.e. because if they have high PageRank, they're more important documents).
10% of queries contain mis-spellings.
Jeff also goes into stemming and explains how Google uses clusters and probability to determine best serps for a given query. Certainly a good example of why good seo is more about semantics and synonyms than exact match keyword phrases.
(Hat Tip: WMW)
Interesting points covered:
While PageRank is query independent, Jeff mentions it is used in calculating results: "High page rank is better than low page rank". Hmmm. He doesn't mention the weighting of ranking factors, however.
He also notes that PageRank is used to decide which shards should have more replications (i.e. because if they have high PageRank, they're more important documents).
10% of queries contain mis-spellings.
Jeff also goes into stemming and explains how Google uses clusters and probability to determine best serps for a given query. Certainly a good example of why good seo is more about semantics and synonyms than exact match keyword phrases.
(Hat Tip: WMW)
