Arabic Sentiment Analysis: Reviews Of The Effective Used Algorithms

Authors : İnas CUMAOĞLU, Vedat TÜMEN, Yuksel CELIK

Pages : 24-30

View : 14 | Download : 6

Publication Date : 2022-06-30

Article Type : Review Paper

Abstract :With the large number of pioneers of social networking sites and the large number of web users in general, many texts are formed in an unstructured way, but they may be useful in several areas if they are structured and processed using Natural Language Processing insert ignore into journalissuearticles values(NLP); techniques. Through these texts, comments, tweets, or even product reviews or books, we can get to know the author’s thoughts and viewpoint on a specific matter. From this principle came the idea of sentiment analysis insert ignore into journalissuearticles values(SA);, which is an advanced and important science in artificial intelligence insert ignore into journalissuearticles values(AI); and machine learning insert ignore into journalissuearticles values(ML); and insert ignore into journalissuearticles values(NLP); that aims to know the aspirations and trends of people through their writings on websites in order to be used in improving a product, predicting the state of the stock market, knowing the public`s political opinions, and many more applications. However, it is still at the beginning of its development in the processing of Arabic texts compared to English texts, due to the complexity of the Arabic language grammatically and morphologically, as well as the lack of Arabic corpus, so in this study we shed light on the latest literary and scientific studies that focused on Arabic sentiment analysis insert ignore into journalissuearticles values(ASA); to identify On the most important algorithms that have proven their quality and effectiveness in this field, where we noted the researchers’ interest in the experience of using deep learning algorithms insert ignore into journalissuearticles values(DL);, which showed their efficiency in this field, in addition to the use of many text extraction techniques, which was the most prominent TF-IDF, CBOW and Skip-gram.
Keywords : Arabic sentiment analysis, data mining, deep learning, NLP, social media