A simple example would be the color of your product. Perhaps your company sells T-shirts and your want to find out which colors are the most popular with your customers. However, looking into the data, you discover that aside from plain one-colored shirts, you also have some with patterns and prints in various colors on them. Aside from the column
COLOR, which you thought would give you all the information you need, you discover that there's another column called
PATTERN_COLOR that contains a comma-separated list of all the most notable colors in your shirt's pattern or print. This means that a T-shirt may qualify as
'blue' not only when
COLOR = 'blue', but also when 'blue' is contained as one of the entries in
PATTERN_COLOR. This also means that a white-and-blue striped shirt can belong to the set of white shirts AND the set of blue shirts at the same time. The simple question has just gotten a lot more complex and you need to figure out how you want to deal with this level of complexity.