Question:

If an Admin enables multilingual document handling (To do so, navigate to Admin --> Settings --> Language & Region Settings ) for the Vault, each document in a multilingual Vault includes the Language field. For PDF and text-based files like HTML or CSV, Vault attempts to assign a language automatically upon import based on the document’s language. 

If a PDF file or a text-based file contains two languages, how does Vault decide the language?

Answer:

The first 1024 characters of the file are used to detect the language. If there is a mix of Japanese and English in those first 1024 characters, the one with the higher probability score is selected. 

Related Documentation:

Vault Documentation: Vault Help