Our KYC customer table has 100,000 records. In your data quality check you find 3,000 records with no date of birth, 1,500 where the date of birth is later than the account open date, and 2,000 tax IDs that each appear on two different customer records. How would you measure each issue, and which one would you fix first?

KYC 客户表有 10 万条记录，你做数据质量检查时发现：3,000 条没有出生日期，1,500 条出生日期比开户日期还晚，还有 2,000 个税号各自出现在两条客户记录里。这几个问题分别怎么衡量，你会先修哪个？