doi: 10.5281/zenodo.4682936
Pre-trained models to parse Arabic text using the STEPS dependency parser (https://github.com/boschresearch/steps-parser).