We study the variation of word frequencies in Russian literary texts. Ourfindings indicate that the standard deviation of a word's frequency acrosstexts depends on its average frequency according to a power law with exponent$0.62,$ showing that the rarer words have a relatively larger degree offrequency volatility (i.e., "burstiness"). Several latent factors models have been estimated to investigate thestructure of the word frequency distribution. The dependence of a word'sfrequency volatility on its average frequency can be explained by the asymmetryin the distribution of latent factors.
展开▼