id	author	title	date	pages	extension	mime	words	sentences	flesch	summary	cache	txt
work_hh4gchixdzfrnbivmmqly4iaza	Patrick Juola	A Prototype for Authorship Attribution Studies	2006.0	15	.pdf	application/pdf	6054	425	56	any technical judgment questionable, especially if the test involves subtle statistical properties such as "vocabulary size" or "distribution of function words," methods designed/tested for the most part for modern English on documents in Middle English, the size of these documents (very few letters, today or in centuries past, exceed 1000 words) makes statistical inference difficult. process of developing a new authorship attribution algorithm : if you can't get As an example of how this procedure works, we consider a method for identifying the language in which a document is written. The Burrows methods [Burrows, 1989, Burrows, 2003] for authorship attribution can be described in similar terms. For example, [Juola and Baayen, 2005] describes two techniques based on cross-entropy that differ only in their event models (words vs. 5Java Graphical Authorship Attribution Program; the authors invite suggestions for a better name for future versions. These methods apply authorship attribution techniques to	./cache/work_hh4gchixdzfrnbivmmqly4iaza.pdf	./txt/work_hh4gchixdzfrnbivmmqly4iaza.txt