Opinion Analysis from Arabic Texts: A Case Study of the Algerian Dialect Comments

Abstract

Opinion mining or sentiment analysis has emerged as a dynamic field within natural language processing, focusing on the computational analysis of sentiments and emotions expressed in written human languages. In recent years, sentiment analysis has gained significant traction across various domains including politics, production, services, marketing and others. This interest stems from the understanding that opinions hold substantial impact and contribute to decision-making processes across these domains. In this study, our main objective is to develop a system that performs sentiment analysis on Arabic texts, specifically focusing on the Algerian dialect. The system aims to analyze the extracted sentiments and classify them into positive, negative, or neutral classes using machine learning techniques and deep learning algorithms