(29-04-2024, 02:54 PM)pfeaster wrote: You are not allowed to view links. Register or Login to view.Interesting! See also Marco's study of textual patterning around image intrusions here:
You are not allowed to view links. Register or Login to view.
Thank you for pointing out Marco's work!
I wish I had found it prior to writing our paper so I could have included a reference to it.
(I had searched and inquired about anyone trying to analyze the feature, but had only found that of Julian Bunn.)
I should point out, as I did in the paper, there is a distinct difference in prevalence vs propensity. Looking at prevalence (i.e. counts of tokens) is a useful first pass analysis, but it can also be misleading.
I also wanted to explain that using a comparison to some results from previous work. Unfortunately, Julian's lists of tokens found in different positions happened to use a portion of the manuscript that was mutually exclusive to what we used.