Barriers and drivers for biking: What can policymakers learn from social media analytics?

Abstract

Broader acceptance of biking as a means of urban transportation is essential for reducing the externalities of motorized transport and contributing to individual health. This paper explored the barriers and drivers for biking from citizens’ perspectives and aimed to provide policy suggestions to promote cycling. Social media analytics was applied supported by topic modeling, machine learning, and sentiment analysis. Topic modeling was used to assign social media messages to the relevant dimension, while sentiment analysis was employed to measure citizens' satisfaction levels. The study was conducted in Turkey and covers more than 600,000 tweets posted between 2016 and 2021. The results revealed that social media analytics generated compatible results with surveys and interviews and successfully defined the factors that affect cycling. Moreover, it captured the temporal changes and provided a dynamic view of policymaking. Findings pointed to the importance of economic conditions, physical infrastructure, and safety&security among barriers and elaborated health, entertainment, and socialization among drivers. The proposed new approach was compared with the traditional methods and its advantages and disadvantages were discussed. Policy implications were derived.