“…Although there has been a significant improvement in the number of available NLP resources e.g. Hamshahri Corpus (Darrudi et al, 2004), Bijankhan Corpus (Bijankhan et al, 2011, FarsName (Hajitabar et al, 2017), ShEMO (Nezami et al, 2019), Persian Dependency Treebank (Rasooli et al, 2013), SentiPers (Hosseini et al, 2018), FarsNet (Shamsfard et al, 2010a) or PersBERT (Farahani et al, 2020) in recent years, Persian is still a heavily unresourced language compared to English or German. The second problem is related to the challenging character of Persian itself and inconsistencies in the writing system.…”