id	author	title	date	pages	extension	mime	words	sentences	flesch	summary	cache	txt
work_6x6zapa3jvhznbmyfn6kraekqm	Levi King	Word-level language identification inThe Chymistry of Isaac Newton	2014.0	6	.pdf	application/pdf	1740	127	60	Word-level Language Identification in "The Chymistry of Isaac Newton" languages within a sentence, as in the "Chymistry of Isaac Newton" (Walsh and Hooper given that all these additional Newton-era and modern texts are monolingual, and that the All previous work in language identification assumes that each text to be identified is The simplest methods use the presence of language-specific characters in a text grams rather than words and reach an accuracy of 99.8% given texts with at least 400 nFor both English and Latin, texts of approximately 70,000 words were used as Word-Based Language Identification: The Newton Corpus the languages used in the corpus: English, Latin, and French. Since there is no training data for French, we use only English and Latin for Word-Based Language Identification: Using Other Corpora for Training either training texts from Newton's era, or modern corpora. available, we used the Newton-era Latin texts.	./cache/work_6x6zapa3jvhznbmyfn6kraekqm.pdf	./txt/work_6x6zapa3jvhznbmyfn6kraekqm.txt