Recent Progress of Sentiment Analysis

The swift advancement of the Internet has led to a proliferation of individuals' opinions and commentary regarding daily life on social media and various other platforms. Concurrently, sentiment analysis (SA) has experienced significant growth, focusing on the emotional tone embedded in individuals' remarks. The advent of deep learning technologies, including multimodal and large language models, has facilitated the emergence of numerous novel sentiment analysis frameworks. This article provides a comprehensive overview of the definition, research significance, and contextual background of sentiment analysis. Then, according to the different methodology, we divide sentiment analysis into: Lexicon-Based Methods SA, Machine Learning Methods SA and Deep Learning Methods SA. Several corresponding state-of-the-art models are collected and discussed in this part. Finally, various prominent application domains such as business analysis, social media, and public opinion analysis are suggested.