Dialect Normalization using Large Language Models and Morphological Rules