Challenges and Approaches in Arabic Sentiment Analysis: A Review

A wide range of social media platforms have been established lately and have become an integral aspect of modern life in the digital age. Enormous volume of user-generated data through numerous social networking sites provide new perspectives to corporations and governments. However, it has become more challenging to properly glean relevant information from the huge volume of data. In order to solve this issue, sentiment analysis techniques have been used to extract and assess the emotion, opinion, and sentiment polarity in written communication. Numerous researches have been undertaken in the field of sentiment analysis, particularly on text written in English, whereas other languages such as Arabic have received limited attention. This paper aims to review sentiment analysis approaches in the Arabic language and discusses the challenges, processing pipeline, evaluation metrics, and levels of Arabic sentiment analysis (ASA). It also provides a summary of various cutting-edge techniques, studies, and datasets available for Arabic sentiment analysis in different domains.