Text mining for the evaluation of public services: the case of a public bike-sharing system

This study conducted text mining analysis of the review text data (13,615 accounts) posted on SNS by users of the public bike-sharing service in South Korea. A total of 11,954 reviews were processed with SKT KoBERT and classified them either positive or negative. Subsequently, various text mining techniques were used to determine the factors affecting the users’ polarity. The study results revealed that the identification of the positive and negative factors affecting service quality through an analysis of reviews by text mining contributes to the improvement of the public bike-sharing system.