Travel origin-destination (OD) demand is of critical significance in decision making of urban transportation planning and infrastructure investment. The conventional travel survey methods suffer from several problems such as declining sample sizes; increasing non-response rates; non-representative samples; missing activities and trips and imprecise travel time. Moreover, these surveys are always costly, time-consuming, and labor-intensive and almost no organizations or institutes can afford them on an annual base. Therefore, it is extremely difficult to capture the up-to-date trend in travel demand patterns using the conventional methods, which prevents policy makers from assessing and adapting policies and management strategies in a timely manner.
On the other hand, recent development of social media tools (e.g. Facebook, Twitter, etc.) has made available the collection of passive datasets for traveler's real-time information. Passive data, compared to active solicitation (e.g. travel survey), is not intended to be collected for travel behavior analysis. However, such data has great potentials for research and practical transport planning applications due to some unique features such as massive, low cost and extensive spatial and temporal coverage.
Motivated by the combination of increasing challenges in administering household travel surveys and advances in emerging social media platforms, this project suggests a bottom-up travel behavior driven approach which obtains trends in individual travel behavior from longitudinal social media data first and use such information to enhance longitudinal origin-destination demand monitoring.