Structured Approaches To Data Selection For Speaker Recognition