Are a key component in data set building by matching common fields in tables that are related. To ensure data accuracy and limit redundancy, data is split into subject-based tables so that each fact is represented only once; table relationships are then defined and common fields mapped to paint a complete picture.