Financial sentiment analysis (FSA) is an integral technique and is growing at a steadfast rate in order to help in market evaluations by determining the stock prices and allow investors to make informed financial decisions. Large language models (LLM)were developed, such as FinBERT, to ameliorate the performance in sentiment analysis. This study analyses the performance of LLMs and transformer models such as FinBERT, and their performance being used in order to predict stock prices by the scraped financial headlines. This system is built to help dynamic financial headlines help understand the relevance of investing in stocks and allows the model to adapt to diff linguistic patterns and financial contexts. Numerous factors contribute to the nature of the market, including economic factors, performance of organizations and legal factors, sentimental factors such as emotions in financial news and social media contribute largely to the stock prices as well as the overall performance as observed. To deeply comprehend and harness these, the study involves interrogating the influence of sentiment signals extracted from financial headlines using transformer-based models, evaluating their impact on stock price prediction and comparing them to traditional methods
Mariam Ghani (Mon,) studied this question.