MUSTAFA MARZOUK AND MOHAMED ASHRAF

October University for Modern Sciences and Arts (MSA University) and the Faculty of Computer Science extend heartfelt congratulations to the remarkable students Mustafa Marzouk and Mohamed Ashraf.

Their collaborative paper, titled An enhanced Transformer-based Approach with Meta-Ensemble Learning for Arabic Sentiment Analysis” (Sep-2023)’ has been successfully published in the International Journal (IEEE) under the supervision of Professor Ammar Mohammed.

Their paper was published at IEEE at the 3rd International Mobile, Intelligent, and Ubiquitous Computing Conference (MIUCC-23). This research paper proposes an approach to address Arabic sentiment analysis by integrating ensemble learning with transformers. The fundamental goal of this work is to suggest an effective strategy for leveraging the potential of baseline transformer models, which are initially fine-tuned and rigorously evaluated for their performance.

Their approach achieved 95.2% accuracy, which is 2% higher than the best meta-ensemble deep learning model.

You can read the paper through the following link:

Congratulations for your outstanding work