Sentiment Analysis in Moroccan Dialect via Arabic Transcoding: Evaluating Different Machine Learning Strategies

This research aims to improve sentiment analysis for the Moroccan dialect (Darija) by converting it into Modern Standard Arabic (MSA). Darija poses unique challenges for natural language processing (NLP) due to the lack of extensive computational resources. To address this, we utilized an existing Arabic dictionary and developed a supplementary dictionary for specific Darija words significantly different from MSA. This approach allows us to leverage the robust tools and resources available for MSA, optimizing time and enhancing sentiment analysis effectiveness. Our study involves a comparative analysis of three classical machine learning models: Support Vector Machine (SVM), K-Nearest Neighbors (KNN), and Naive Bayes (NB). Testing these models on the MAC Dataset, both before and after applying our conversion method, revealed significant improvements in accuracy, demonstrating the potential of our approach.