Social Media-Based Traffic Situational Awareness under Extreme Weather

Traffic information under extreme weather cannot be collected in a timely and accurate manner by traditional traffic information collection methods, due to various restrictions. Although recent evidence suggests that social media, such as Twitter, is a powerful supplement to traditional traffic information collection methods, which plays an important role in addressing the issue of traffic situational awareness, few studies take Chinese social media data as the object of analysis. To fill this gap, based on the social media data scraped from Weibo, which is the largest Chinese social platform in China, this study provides new insights into comparing the metrics of different machine learning classification algorithms. The results indicate that the long short-term memory (LSTM) classifier outperformed others and achieved 93.8%-95.8% accuracy.