
114 ◾ Saravanan Muthaiyah and Larry Kerschberg
5.3.1.7 NSS Measure (NS)
Cilibrasi and Vitanyi (2007) introduced the Normalized Search Similarity (NSS)
adapted from Normalized Google Distance (NGD). ey measured similarities
between two concepts using probability of co-occurrences as demonstrated by the
following equation:
NGD (c1, c2)=
max {log (c1), log (c2)} – lf f oog (c1, c2)
logM – min {log (c1), log (c
f
f f
22)}
(5-7)
M is the number of searchable Google pages, and f(x) is the number of pages that
Google search returns for searching x. NGD is based on the Google search engine.
e equation may also be used with other text corpora such as ...