Preliminary Study for Impact of Social Media Networks on Traffic Prediction

Abstract

While smart cities have the required infrastructure for traffic prediction, underdeveloped cities lack the budget and technology to perform an accurate model. Current research uses data mining of tweets and specific posts to provide population trends, but there is no work done in social network analysis for the same end. This paper proposes an applied informatics application with social network usage to aid in the lack of data due to nonexistent traffic sensors. The Twitter API was used to download a network of users that follows traffic updates accounts and then, use a model of information diffusion (independent cascade model) to retrieve a variable that holds a metric of how the information regarding current traffic has traveled through the network. Finally, an updated traffic dataset with the new social network variable is used to train and validate an LSTM neural network to show if the new variable can be a predictor for traffic. Results show that a deterministic independent cascade model ran on a New York City-based 2-tier social network marginally improved the prediction by 0.4%. This proposal can be replicated in other information diffusion models.