Review of Researches on Arabic Social Media Text Mining
Manal Othman Hamad Alothman, Muhammad Badruddin Khan, Mozaherul Hoque Abul Hasanat · Journal of Intelligent Systems and Computing · 2021
Social media sites and applications have allowed people to share their comments, opinions, and point of views in different languages on mass scale. Arabic language is one of the languages that has seen huge surge in production of its digital textual content. The Arabic content and its metadata are a goldmine of useful information for a wide variety of applications. A large number of researchers are working on Arabic data in various domains of research such as natural language processing, sentiment analysis, event detection, named entity recognition, etc. This article presents a review of number of such studies conducted between 2014 and 2019 using their data sources from social media websites. We found that Twitter was the most used source to contribute data for dataset construction for Arabic text mining researchers. Our study also found that the Sport Vector Machine (SVM) and Naïve Bayesian (NB) classifiers were the most used classifiers in the previous researches. Moreover, the results of the previous studies indicate that SVM classifier provided the best performance compared to other classifiers.