Data parallelism is characterized by the parallel execution of the same operation on different parts of a large data set.