Automating Data Classification for the 21st Century
blog.sdf.com
Introduction Imagine a warehouse with hundreds, thousands, or millions of tables. How do you know what purpose each table has, or what each column represents? In most SQL-based DBs, the best you can do is use meaningful names, types, and comments. But that is not enough to control the complexity that comes with scale. Unlike modern programming languages with advanced type, object, and procedural abstractions, SQL’s types and procedural abstractions are fairly primitive. In fact in most warehouses more than 50% of columns are just of type VARCHAR and most of the rest are other basic types. So how can you augment existing queries and pipelines with additional meaning? With data classification!
Automating Data Classification for the 21st Century
Automating Data Classification for the 21st…
Automating Data Classification for the 21st Century
Introduction Imagine a warehouse with hundreds, thousands, or millions of tables. How do you know what purpose each table has, or what each column represents? In most SQL-based DBs, the best you can do is use meaningful names, types, and comments. But that is not enough to control the complexity that comes with scale. Unlike modern programming languages with advanced type, object, and procedural abstractions, SQL’s types and procedural abstractions are fairly primitive. In fact in most warehouses more than 50% of columns are just of type VARCHAR and most of the rest are other basic types. So how can you augment existing queries and pipelines with additional meaning? With data classification!