Real time road traffic event detection using Twitter and spark

Because of its real-time nature Twitter is a powerful source of information for detecting real time event happening in the surrounding. People make tweets on whatever events they are seen in their day to day life. In this paper, we take advantage of this for detecting road traffic by using tweets related to traffic and accidents. The goal of this system is to assign an appropriate class label by using classification techniques. For classification of tweets, we employed logistic regression and SVM classifier. We also used some text mining, machine learning, statistics and natural language processing techniques to extract meaningful information from raw tweets. As there are 6,000 tweets are twitted per second on twitter. So Apache spark is used to handle such large amount of data.