unstructured data is simply data that has not been processed to yield information.
examples of both types would include:
An invoice. If one were to take an invoice and simply scan it into a graphic, it would be unstructured data. In contrast, if it were processed and put into a database (subsequently becomming structured data), employees could eventually find the monthly averages, amount owed, etc. from various invoices.
While both are prevelant, I would think semistructed data would be the most common in a typical business. Some data is stored but not processed (unstructured data such as memos), and some others are stored in databases (such as invoices) but most data are only processed to a certain extent that is displayed in a prearranged format but not able to yield all of the information contained within.