Discovering Functional Zones in a City Using Human Movements and Points of Interest

Abstract

The development of a city gradually fosters different functional zones, such as educational areas and business districts. Typically, a city is segmented into disjointed regions by major roads, such as highways and urban expressways. In this chapter, we propose a framework for discovering functional zones in a city through the analysis of human mobility among regions and points of interest (POIs) within regions. Specifically, we infer the functions of each region from the results of a topic-modeling-based approach, which regards a region as a document, a function as a topic, categories of POIs (e.g., restaurants and shopping malls) as metadata (such as authors, affiliations, and keywords), and human mobility patterns (origins/destinations and arrival/departure times) as words. As a result, a region can be represented by a distribution of functions. This type of representation enables functional zones to be identified, which are comprised of clusters of regions with similar distributions of functions. We then further identify the intensity of each functional zone type occurring in different locations. We evaluated our method using large-scale, real-world datasets, consisting of two POI datasets of Beijing (in 2010 and 2011) and two three-month GPS trajectory datasets (representing human mobility); these trajectories were generated by over 12,000 taxicabs in Beijing in 2010 and 2011 respectively. The results demonstrate the advantages of our approach over baseline methods which solely use POIs or human mobility.