Every year business forecast impacts the economy in the financial industries. The most intense infeasible forecast model can cause extensive bankruptcy and damage of trust. To mitigate the loss of business and damage of trust, early forecast predictions based on social media clients' opinions are vital. Topic analysis is one of the major critical roles in the world of financial technology to determine clients' trends and patterns for forecast analysis. A comparative analysis of different banks based on their official social media sites gives a contradictory perspective of client needs around the world. Clients' tweets were processed using data science strategies such as deep learning for text classification TensorFlowKeras, latent Dirichlet allocation, Natural Language Toolkit--Natural Language Processing, and long short-term memory using Python/R analyze the history of tweets and identify the key business factors involved. This quantitative study examined the impact of three key factors, security breaches, innovation, and stock exchange, which were tweeted by clients. This automated system helps to analyze all impacts of a single centralized system using deep learning techniques for the forecast model. This research analyzed the seven-day market capitalization forecast provided from clients' tweets over a period of 10 years. The study revealed the significant difference between two banks and filled the three gaps-data breach, innovation, and stock exchange-in the existing model. This new information provides benefits to both banks and clients to predict market capitalization for the successful forecast.
展开▼
