This can be as simple as storing all text that passes through it in a spare file and recalculating with the corpus + that file. Could also be on the fly, I think manipulation of counts and NW\* values is O(1)
This can be as simple as storing all text that passes through it in a spare file and recalculating with the corpus + that file.
Could also be on the fly, I think manipulation of counts and NW* values is O(1)