Fig. 5: Word coverage of the PDS lexicon.
From: A decomposition of book structure through ousiometric fluctuations in cumulative word-time
The blue histograms correspond to the unique word coverage, while the orange histograms correspond to the word coverage allowing for word repetition. We show the histograms for all the books downloaded in our dataset (empty bars) as well as those we analyzed (shaded bars), and the dashed lines show the median word coverage of the books analyzed. The books included in our analysis do not include duplicate titles and must have (1) a unique word coverage of at least 60%, (2) at least one lexicon word in every window, and (3) intrinsic mode functions (IMFs) that can be successfully computed.
