Investigating the potential of social network data for transport demand models

Abstract

Location-based social network data offers the promise of collecting the data from a large base of users over a longer span of time at negligible costs. While several studies have applied social network data to activity and mobility analysis, a comparison with travel diaries and general statistics is lacking. In this paper we analyze geo-referenced Twitter activities from a large number of users in Singapore and its neighbouring countries. By combining this data, population statistics and travel diaries, and applying clustering techniques, we address questions regarding the detection of activity locations, the spatial separation between these locations and the transitions between these locations. Despite a large number of Twitter users present in the data set which we collected over a period of 8 months, only an amount comparable to a travel survey turned out to be useful for further analysis due to the scattered nature of the data. Kernel density estimation performs best to detect activity locations; more activity locations are detected per user than reported in the travel survey. Descriptive analysis shows that determining home locations is more difficult than detecting work locations for most planning zones. The spatial separation between detected activity locations from Twitter data and as reported in a travel survey and captured by public transport smart card data are at large similarly distributed, but also show relevant differences for very short and very long distances. This equally holds for the transitions between zones. Whether the differences between Twitter data and other data sources stem from differences in the population sub-sample, the clustering methodology or whether social networks are being used significantly more at certain locations is to be determined by further research. Despite these shortcomings, location-based social network data offers a promising data source for insights in activity locations and mobility patterns, especially for regions where travel survey data is not be readily available.