Analyzing spatial land use interlinkages using Co-location pattern mining from POI data

Abstract

Cities are not random mix of land uses but rather a deliberate arrangement of uses and functions formed via market and government intervention. In order to minimise urban problems and promote better built environments a good understanding of how land uses interact with each other is vital. Along with this urbanization is at its peak with 50% of the country expected to be urban by 2050, as more and more people will flood the cities, they will become dense and compact and if not managed properly chaotic, there is a need for systematic development of cities and optimum utilization of land resource before it’s too late. Land use relates to how various uses are arranged in a spatial environment and that determines how efficiently city function. The aim is to club together uses that minimise the need of resources or which act in synergy for better urban functioning. The study of land use dynamics and the interactions between various uses will help planners plan for better cities. In Indian context we have land use planning in the form of development plans for a span of 20 years. It contains broad land uses and a set of ancillary uses allowed in it. The market plays its part and dictates what comes up as a result of need and regulations in place. But what uses have really come up and whether they are conforming or non conforming in nature and how they are spatially arranged at the lowest level is still quite unknown to us which is what the study tries to uncover. The focus is to study how land use linkages occur in the cities under study and quantify them. This will act as a base for urban planners while creating land use plans and associating various land uses which currently is intuition-based. Point of interest data from Google maps is mined to study land-use linkages at the lowest level possible. Once the POI data is mined co-location pattern mining algorithm is used to analyse how these land uses interact with one another in the cities under study. This will uncover detailed land use dynamic which is currently unknown and help decide optimal location for a facility. The results showcase the level of interaction between land uses and their sub-types based on market forces.