A Real-Time Spatio-Temporal Bigdata System for Instant Analysis of Twitter Data to Monitor of Advertising Campaigns; Case Study New York City

Abstract

Advertising monitoring is useful for detecting and determining user reaction, campaign performance and their challenges. Popular social networks especially Twitter are one of the best resources that could been analyzed for this purpose. Twitter's data analyzing needs different methods and tools because Twitter is one of the bigdata sources. In this paper, a system is developed dealing with Twitter data streams and can monitor advertising keyword spatio-temporal trends, real time. Also, it can predict the location of later tweets using MLP model. The prediction accuracy is 0.78.