ˇM & W examined the
frequency of 100 function words
ˇSmoothed
these frequencies using negative binomial (not Poisson) distribution
ˇ
ˇ
ˇ
ˇ
ˇUsed Bayes’ theorem and
linear regression to find weights
to fit for observed data
ˇ
ˇSample words:
ˇ as do has is no or than this
ˇ at down have it
not our that to
ˇ be even her its now shall the up