Mining Complaints for Traffic-Jam Estimation: A Social Sensor Application

Physical events in the real world are known to trigger reactions and then discussions in online social media. Mining these reactions through online social sensors offers a fast and low cost way to understand what is happening in the physical world. In some cases, however, further study of the affected population's emotional state can improve this understanding. In our study we analyzed how car commuters react on Twitter while stuck in heavy traffic. We discovered that the online social footprint does not necessarily follow a strict linear correlation with the volume of a traffic jam. Through our analysis we offer a potential explanation: people's mood could be an additional factor, apart from traffic severity itself, that leads in fluctuations of the observed reaction in social media. This finding can be important for social sensing applications where external factors, like sentiment, also contribute on how humans react. We propose a novel traffic-congestion estimation model that utilizes the volume of messages and complaints in online social media, based on when they happen. We show through experimental evaluation that the proposed model can estimate, with higher accuracy, traffic jam severity and compare the results with several baselines. The model achieves at least 38% improvement of absolute error and more than 45% improvement of relative error, when compared with a baseline that assumes linear correlation between traffic and social volume. To support our findings we combined data from the California Department of Transportation (CALTRANS) and Twitter, for a total of 6 months, and focused on a major traffic-heavy freeway in Los Angeles, California.