MAG contains duplicated author and possibly institution names. Microsoft has done an incredible job on entity disambiguation, however, it's not a solved problem.
Given time constraints and in the spirit of being more agile, I would recommend working on this at a later stage and only if we notice that it greatly distorts the results.
MAG contains duplicated author and possibly institution names. Microsoft has done an incredible job on entity disambiguation, however, it's not a solved problem.
Given time constraints and in the spirit of being more agile, I would recommend working on this at a later stage and only if we notice that it greatly distorts the results.
This covers Munging and shaping:[Disambiguate people and institutions] from Roman's trello board.