The Great All-Knowing Google is upon us

Almost. This nifty paper, Automatic Meaning Discovery Using Google, describes a method for extracting meaning from Google searches. The authors compute probability functions based on the number of hits for each term. They take two terms, measure the number of hits for each term separately, and then measure the number of hits with both terms together. These probablities are combined to form a distance measure between the two terms. This distance measure can be fed into various knowledge classification and learning algorithms.