Mining Social Media Data for Transport Policy: Approaches, Challenges, and Recommendations

Compared with surveys and interviews, social media data can yield a more sociological understanding of public perceptions toward transport policy in a time- and cost-effective manner. This paper offers a systematic review of the fundamental logic, methodologies, challenges, and some corresponding recommendations for using social media data in transport policy research. The paper summarizes two frameworks for social media-based policy analysis as well as the fundamental models. Five main challenges in social media-based policy research consisting of sampling representativeness, noise removal, text pre-processing for Chinese and English, result interpretation, and cognitive bias are proposed here. We conclude that employing manually double-checking, using multiple data sources, drawing portraits of target groups, and examining the existence of echo chambers can benefit the policy-side application. Furthermore, we provide some practical examples and case studies of transport policy to give in-depth explanations. This paper highlights the roles and directions of using social media to deliver transport policy goals in the new era of Information and Communication Technologies (ICTs).