Ontology learning plays an important role in many fields especially in the Semantic Web. The success of the Semantic Web depends on the quality of its ontologies. There is a lot of research work interested in Ontology Learning for Arabic texts. Most of these works focused on three main issues: extracting the terms, extracting the semantic relations, and building the ontology from the Arabic text. In this paper, first, we present the Arabic challenges that were reasons for developing few Arabic Ontology Learning systems. Second, we make a research comparison based on the techniques used and their results. Third, we pointed out the limitations and comments of research works on Arabic Ontology Learning. Finally, we concluded the paper and outlined our future research direction in this area.