Corpora: log likelihood statistic

Daniel Ridings (ridings@svenska.gu.se)
Mon, 21 Jul 1997 21:31:25 +0200 (METDST)

Ted Dunning writes about this in "Accurate Methods for the Statistics of
Surprise and Coincidence" in Computational Linguistics Vol. 19, 1993.

Could some kind soul walk me through the formula at the very end of the
article, using the first bigram in Table 2 for illustration? The copy I
have in front of me is a reprint in "Using Large Corpora" and there
appears to be a misprint, so I won't repeat the formula here. (The
misprint is in no way responsible for my dim wit).

Daniel Ridings
Gothenburg, Sweden