The combination of data driven and keyword driven framework is called the hybrid. Here the operations/instructions/keywords in a separate repository (.csv/.xls/.json/DB) and data is in separate (.csv/.xls/.json/db from data provider) and the tests/driver would read both and perform the actual tests automatically. In this design, we get the best of both methodologies, and it is kind of practical in most of the automation cases.
Hope, you got the answer.