Matching of people's names is too loose
I often get a list of people who's names are very different to the one I typed. It should probably use a Levenstein distance to reduce the number of false positives. Also consider eliminating people with no work more than 40 years before or after the production that is being edited.
1
vote