Limiting the amount of fuzzy string comparisons by comparing by subgroup

Question

I have two datasets as follows:

DT1 <- structure(list(Province = c(1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 
2, 2, 2, 2, 2, 2, 2, 3, 3, 3, 3, 3, 3, 3, 3, 3), Year