When cleansing party master data it is often necessary to typify the records in order to settle if it is a business entity, a private consumer, a department (or project) in a business, an employee at a business, a household or some kind of dirt, test, comic name or other illegible name and address.
Once I made such a cleansing job for a client in the farming sector. When I browsed the result looking for false positives in the illegible group this name showed up:
- The Slurry Project (in Danish: Gylleprojektet)
So, normally it could be that someone called a really shitty project a bad name or provided dirty data for whatever reason. But in the context of the farming sector it makes a good name for a project dealing with better exploitation of slurry in growing crops.
A good example of the need for having the capability to adjust the bad word lists according to the context when cleansing data.
Liliendahl on Data Quality 