SDE083  |  Bus Analytics Using Twitter

In a rapidly growing city population, a commuter friendly public transport system is essential in coping with the increasing demands of a thriving city. We propose a crowdsensing framework to gather and analyse real-time commuter feedback from Twitter. We perform text mining tasks to identify feedback comments containing bus-related micro-events; extracting entities; and, predicting event and sentiment labels. The experiments show that our domain-sensitive text analysis models are effective. We further solicit structured feedback tweets from commuters. This technology is useful in gathering feedback for all service-oriented systems. 

Exhibitor : SMU Living Analytics Research Centre