Social Sensing on Singapore MRT
As more and more public MRT announcements, comments, and feedbacks are shared through social media platform such as Twitter, analysis of MRT-related social media content sheds light on MRT events and commuting experience. MRT Sense dashboard visualizes analysis of labelled MRT-related Tweets from January 2015 to July 2017.
We manually label the tweets into 11 categories: train operation anomaly, modification, design, normal, incidents, crowd, staff, temperature, etiquette, security, and announcement. Click category info on the navbar menu for detailed description on each category.
Tweets Over Time
Train Operation Anomaly!
Majority of the feedback classified as train operation anomaly.
Quiet in 2016
We sense less feedbacks in 2016, with the highest feedbacks of the year are about Pasir Ris accident in March.
More feedbacks in 2017
We sense more feedbacks in 2017, especially in June 2017.
Label Proportion Over Time
Main proportion comes from Train anomaly and temperature related feedbacks. We also observe a reasonable amount of feedbacks on normal trains.
2015
2016
2017
Label Correlation
We compute Pearson Correlation Coefficient based on event occurences within one hour span from 2015 to 2017.