README.md

---
license: mit
---
The entire Western Canon as a dataset. Or, at least, a part of it.

Several translations of the Bible. Various ebooks. I took a large list of ebooks and filtered it to only books that contained one of the following keywords:

"the iliad, by homer", "the odyssey, by homer", "euclid", "Plato", "aristotle", "by Cicero", "by Saint Augustine", "Thomas Aquinas", "galileo", "copernicus", "kepler", "martin luther", "Descartes", "Spinoza", "Leibniz", "Isaac Newton", "david hume", "Kant", "fichte", "hegel", "kierkegaard", "Nietzsche", "by adam smith", "wittgenstein", "Einstein", "Bohr", "George Boole", "Gottlob Frege", "henri bergson"

About 500 ebooks total.

I then further refined the list to remove as much noise as I possibly could. Their license allowed me to remove their license, so I used various scripts to do so.

This is not a complete list by any means. There is also still a good bit of noise in the dataset, such as emails, phone numbers, etc from the contributors. Because I filtered this list by keywords in book titles, there are some books in here that are not part of the Western Canon, such as commentaries on the above authors. This book also contains a mix of text in both the author's native language and English translations. And there are some repeats across this dataset. And there might be some formatting errors here and there. As time goes on I hope to cultivate a better open source dataset of the entire Western Canon.