We currently use the PR creators as the contributor information, but we need to find better way to extract the actual contributor information