There are two meanings of "words": types and tokens.
To determine a key word list's
type-coverage you'd use the number of (positive) kws identified and compute that as a percentage of the total types in the study word list.
For
token-coverage you'd need to sum the frequencies and compute that as a percentage of the token count of the study word list. It'd be very easy to compute. And is not computed in WS1.0 to 8.0 so far. I will do so in the new 8.0 64bit version, and insert the two results in the Notes created as one makes a kw list.
Hope that helps you Abdullah,
Cheers -- Mike