Skip to main content
. 2013 Dec 26;4:991. doi: 10.3389/fpsyg.2013.00991

Table 1.

HiDEx parameter set for computing co-occurrence measures.

Corpus UseNet Corpus
Corpus size 12,714,502,395 words
Context size 10000 words
Window length behind 5 words
Window length ahead 5 words
Weighting scheme Inverse Ramp
Normalization method PPMI
Similarity metric Cosine

For more details, see Shaoul and Westbury, 2010a,b,c, 2011.