The analysis of word frequency count data can be very useful in authorship attribution problems. Zero-truncated generalized inverse Gaussian-Poisson mixture models are very helpful in the analysis of these kinds of data because their model-mixing density estimates can be used as estimates of the...