Real-time Twitter data sentiment analysis to predict the recession in the UK using Graph Neural Networks.

Abstract

The global economy is rapidly contracting, and as more countries experience recessions, it is crucial to have reliable tools for predicting them and figuring out the variables that influence how well an economy is doing. Twitter and other social media platforms have become important informational resources for forecasting market movements and identifying potential future threats. In this study, sentiment analysis is performed on real-time Twitter data to forecast a recession for the UK economy. Twitter feeds on recession are fed into Azure to pre-processes the data, and to produce insights on the recession's contributing factors. A model to accurately forecast the recession using a Graph Neural Network (GNN) is created on the processed data. The main contribution of the research is the creation of a framework for forecasting the UK recession using GNN on realtime Twitter data. The purpose of the study is to give insights into the variables influencing the UK's economic health and to pinpoint reliable recession forecasting techniques. Policymakers, economists, and companies wishing to track the UK economy in real time may find the research findings interesting. Overall, the work has important ramifications for forecasting recessions and keeping tabs on economic circumstances utilising Twitter, GNN, and Azure Databricks.