Real-Time Sentiment Analysis of Twitter Streaming data for Stock Prediction

In this study, an attempt has been made for making financial decisions such as stock market prediction, to predict the potential prices of a company’s stock and to serve the need of this, Twitter data 1 2 has been considered for scoring the impression that is carried for a particular firm. Streaming data proves to be a perennial source of data analysis collected in real-time. Streaming data basically deals with the continuous flow of data which carries information from sources like websites, mobile phone applications, server logs, social websites, trading floors, etc. The major characteristics of such data being its accessibility and availability, help in proper analysis and prediction of user behavior in a ceaseless manner. The classifying model made out of historical data can be relentlessly honed to give even more accurate results since its outcome is always compared to the next tick of the clock. Spark streaming has been considered for the processing of humongous data and data ingestion tools like Twitter API and Apache Flume have been further implemented for analysis.