Extraction of traffic information from social media interactions: Methods and experiments

With the rapid development of social media, User Generated Content (UGC) has spawned huge amount of information in the society today. In the age of Big Data, we can provide important information for peoples' transportation needs through exploring and making full use of traffic data in social media. This paper introduces techniques like Natural Language Processing, cloud and open platform, mobile Internet, and human computer interacions to extract traffic information from text-based data buried in social media. We utilized Sina Weibo(weibo.com, a Twitter equivalent in China) as our main source of data, and developed a prototype system that published and captured traffic status through an Android based app (application). Experiments showed that the prototype system ran well in real time.