Ya, it seems like trying to choose words that are hard to mistake for each other (high lev distance, and perhaps also considering letter similarity) would be a

Ya, it seems like trying to choose words that are hard to mistake for each other (high lev distance, and perhaps also considering letter similarity) would be a very good idea for selecting a dictionary like this. I wonder what the original process was?