Big Data Stream Analytics for Correlated Stock Price Movement Prediction

While much research work has been devoted to big data analytics in the past few years, very few studies about big data stream analytics are conducted and reported in existing Big Data literature. Empowered by recent success on automated mining of dynamic business networks, this paper illustrates the design of a novel framework named BIDSTA that leverages big data stream analytics to predict correlated stock price movements. In particular, a business network-based model, named Energy Cascading Model is designed to predict stock price movements by taking into account the evolving sentiments of firms mined from continuous data stream generated by online social media. The business implication of our research is that business managers can apply our design artifacts to more effectively analyze and predict the potential business performance of targeted firms based on the business intelligence extracted from the big data readily available on the Web.