We call Normalized Web Distance a generalization, to any search engine, of the
Cilibrasi and Vitanyi's Normalized Google Distance NGD (visit
"Google's search for meaning"
for further information about Google distance).
This application computes the Normalized Web Distance and its transformation into
a Semantic Relatedness measure between any two words,
calculated in base of frequencies given by a
web search engine (click
here to obtain more information about this Semantic Relatedness measure).
It is supposed that two closely related terms have a shorter distance
(and a greater relatedness) than two loosely coupled terms (try, for example, 'metre' and 'kilometre'
and compare it to 'red' and 'kilometre'). You can choose different web search engines as source of web frequencies.
Normalized Web Distance measure ranges between 0 and infinity, while relatedness comes between
0 and 1.