A company wants consistent metadata identifying sensitive columns across many schemas. Which capability helps?

1. Only longer object names
2. A larger warehouse cache
3. Object tags and governed classification practices
4. Only query-result ordering