Sentiment analysis has emerged as a powerful method to extract polarization of opinion from text. A vast number of prominent sentiment analysis methods can be categorized into rule-based method, lexicon-based method, machine learning method, and deep learning, and ontology-based method. An increasing interest toward sentiment analysis is motivated, among others, by some successful use cases of sentiment analysis in several domains and availability of enormous data as the result of big data explosion in the past ten years. The objective of this study is to identify some latest study reports that represent various trends of sentiment analysis. We implemented sentiment analysis for Arabic tweets, we applied SVM and Naive Bayes algorithms.