We need to have a contributor normalization based on their real name (from their community profile. Ideally the matching alogorithm should be able to cope with missing middle name, and common shortenings for names (e.g. Pete -> Peter)