Journal of E - Technology Tharindu Ranasinghe, Hadeel Saadany, Alistair Plum, Salim Mandhari, Emad Mohamed, Constantin Orasan, Ruslan Mitkov 11 3 2020 https://doi.org/10.6025/jet/2020/11/3/83-90 http://www.dline.info/jet/fulltext/v11n3/jetv11n3_1.pdf Using specific deep learning models, we have introduced irony detection in Arabic language with the help of the IDAT 2019 Shared Task. We have tested a few available models and understand how the document content cleaning and pre-processing work. In the trials we have conducted we found that a higher F1 score is achieved and the RGGL ranks in a top level. We finally found that the introduced system can able to get competitive results.