Discovering Urban Functional Regions with Call Detail Records and Points of Interest: A Case Study of Guangzhou City

Abstract

Urban spatial structure is becoming increasingly complex with the development of cities and there are usually regions of different functions in a city. Understanding the distribution of these functional regions is significant for urban management and planning. The rapid growth of mobile phone users and telecommunication data motivates us to utilize a topic-modeling-based method using call detail records (CDRs) and points of interest (POIs) for urban functional region discovering. In this paper, we describe a data-driven framework based on latent topic model. Specially, we propose a method to obtain the human mobility patterns between different areas by extracting their trajectories from CDRs, which is important temporal-spatial information to discover the functions of a region. We evaluated our method using large-scale and real-world datasets including a POI dataset of Guangzhou and a CDR dataset provided by a certain telecom operator. The results verify that the framework is effective and the proposed method can be a new tool for computational urban science.