This paper presents a systematic review of natural language processing (NLP) and text mining techniques applied to the financial system, covering literature from 2018 to 2023. The authors examine applications across asset pricing, corporate finance, derivatives, risk management, and public finance. They find that most studies combine probabilistic and vector-space models with both text and numerical data, with information classification being the most common technique. Long Short-Term Memory (LSTM) networks and bidirectional encoder models like BERT are the most frequently used algorithms. The review also notes a trend toward developing new, specialized algorithms and a predominant focus on asset pricing. The paper proposes an engineering-oriented pathway for researchers to analyze financial text and highlights critical challenges such as data quality, context adaptation, and model interpretability that must be addressed to fully integrate advanced NLP models into financial analysis and prediction.