The rapid dissemination of information on social media platforms during the COVID-19 pandemic underscores the need for real-time sentiment analysis to gauge public opinion and emotional responses. This paper presents a review of Twitter stream analysis techniques employed for monitoring sentiment surrounding COVID-19 discourse. Utilizing natural machine learning (ML) algorithms, researchers have developed various methodologies to analyze the sentiment of tweets related to the pandemic. Key aspects of these approaches include data preprocessing techniques, feature extraction methods, sentiment classification models, and evaluation metrics. Additionally, the review discusses challenges such as handling noisy and unstructured data, addressing linguistic nuances, and adapting to evolving discourse trends.