DOI: https://doi.org/10.36347/sjet.2025.v13i01.007

Pages: 56-65

In an era defined by economic ups and downs, as well as the general loss in purchasing power parity, many are utilizing additional passive income streams to protect from inflation. Due to the peak of digitization and an increase in online services than offline, investing in the stock market is now preferred by people as a passive income. But investing in the stock market has risks that these can result in great financial losses. One of the risks posed by this volatility is that driven by investor sentiment, a single piece of news can go viral completely rightly - or misleadingly wrong, and its impact on an individual’s investment decisions would be all too significant to influence in which direction trends may shift within financial markets. This paper addresses the evolving field of sentiment analysis for understanding stock market dynamics using machine learning methods. It provides a comprehensive review of the current state-of-the-art literature in terms of research approaches, datasets and results with respect to works published from 2018 till early 2023. It describes a varity of machine learning models, sentiment lexicons and data sources needed to analyze sentiments and how they affect us when it comes to predictions in stock markets. Furthermore, this study serves to explain the workings of sentiment and its role in market movements or so it appears when a narrative is told about stock trends from such perspective that we compile here.