Stock Market Prediction using Hadoop Map-Reduce Ecosystem

Hadoop is used as a solution to handle huge amount of data. Hadoop distributed file system (HDFS) is used to store the data on different nodes. Currently, academia and industry people work on huge amount of data usually in petabytes. They use the map reducing technique for analysing the data. This paper is Hadoop based Stock Market prediction using sentiment analysis and clustering algorithms. The prediction is done using the tweets on tweeter. Twitter API is used to fetch the tweets and Google finance API is used to fetch the cost of stocks. Naive Bayes Algorithm is used to generate the sentiment score. On the basis of the sentiment score the prediction is done. Hadoop Map Reduce is used for managing the large amount of da-ta. Our experimental results show the increased accuracy of the prediction as compared to the existing system. The results shows that this method has a better prediction effect on stock price and helps the investors by guiding them whether to buy or sell the shares.