ANÁLISE E PREVISÃO DE TENDÊNCIAS NO MERCADO FINANCEIRO ATRAVÉS DO PROCESSAMENTO DE LINGUAGEM NATURAL

Abstract

The contemporary financial market is characterized by high complexity, volatility, and the influence of unstructured qualitative factors such as news, reports, and social media. In response to the growing demand for more sensitive and effective predictive tools, this study aims to assess the potential of Natural Language Processing (NLP) in the analysis and forecasting of financial trends. Based on a theoretical and exploratory approach, the research critically reviews traditional forecasting methods,such as fundamental, technical, and econometric analyses, highlighting their limitations in the face of the explosion of textual data. The study explores the use of NLP techniques such as sentiment analysis, text classification, and deep neural network models (LSTM, BERT), examining their application in extracting linguistic patterns that anticipate market movements. Findings from the literature indicate that these techniques enhance forecasting capabilities, especially in highly unpredictable contexts. Additionally, the study discusses ethical and technical challenges related to algorithm use, such as structural biases and lack of transparency in models. It concludes that the integration of NLP and predictive models represents a promising innovation in the financial sector, supporting more informed and contextualized decisions, while requiring continuous technical refinement and critical reflection on its practical applications.