The metadata-detected classes match imported objects by their metadata attributes using an MQL query.
The process of the data-detected classification tries to detect and match unique formats of imported objects. Some imported objects have not unique or detectable patterns about their data. In this case, you can try to identify similar imported objects using the metadata-detected classes.
For example, you can try to classify the date of birth columns by using their data type (DATE) and their name (DOB). A column can also have the DOB name and DATE data type but have nothing to do with the date of birth, such date of bankruptcy.