I am reading an article where the authors mention something they call the expected frequency of a word in a corpus without explaining or giving a definition of what this means.
The article in question is A New Academic Vocabulary List by D. Gardner and M. Davies. You can find a version online: http://courses.washington.edu/englhtml/engl563/PDFs/Applied%20Linguistics-2013-Gardner-applin_amt015.pdf
Is there a standard way expected frequency is calculated?
The paper you linked to describes how the researchers compiled their Academic Vocabulary List (AVL). They first identified a corpus of academic texts, which they extracted from a larger and more general Corpus of Contemporary American English (COCA). They then came up with various measurable criteria by which to select words (or rather lemmas, to be precise) from their Academic Corpus. These criteria involve comparing word frequencies - how often a word appears in the Academic Corpus versus the general COCA. The frequency of the word in COCA (or the non-academic part of it) is the expected frequency.
The use of the word expected here is related to the notion of mathematical expectation.
You need to be able to divide samples into some kind of rows and columns, like "word class" vs. "register". You compute the percentage of examples in the data for each row and each column, and multiply. Read here for a quick summary.
