Social media data assisted inference with application to stock prediction

The access to the massive amount of social media data provides a unique opportunity to the signal processing community for extracting information that can be used to infer about unfolding events. It is desirable to investigate the convergence of sensor networks and social media in facilitating the data-to- decision making process and study how the two systems can complement each other for enhanced situational awareness. In this paper, we propose a copula-based joint characterization of multiple dependent time series from sensors and social media. As a proof-of-concept, this model is applied to the fusion of Google Trends (GT) data and stock price data of Apple Inc. for prediction, where the stock data serves as a surrogate for sensor data. Superior prediction performance is demonstrated, by taking the non-linear dependence among social media data and sensor data into consideration.