A novel transit rider satisfaction metric: Rider sentiments measured from online social media data

Abstract

The goal of this paper is to demonstrate the use of an innovative social media-based data source, Twitter, to evaluate transit rider satisfaction. Transit authorities have access to vast amounts of performance metrics that measure ridership, timeliness, efficiency, safety, cleanliness, and service, to name a few. These performance metrics, however, are generally one-sided; they represent the interests of the business and are not customer-based. This paper recognizes the limitations of standard performance metrics and attempts to gauge transit rider sentiments by measuring Twitter feeds. Sentiment analysis is used to classify a population of rider sentiments over a period of time. Conclusions are drawn from totals of positive and negative sentiments, normalized average sentiments, and the total number of Tweets collected over a time period.