Retrieving Individual Attributes from Aggregate Dataset for Urban Micro-simulation:A Preliminary Exploration

LONG Ying,SHEN Zhenjiang,MAO Qizhi,Anrong Dang

DOI: https://doi.org/10.1109/GEOINFORMATICS.2010.5567738

2011-01-01

GEOINFORMATICS

Abstract:As the traditional top-bottom based macro-simulation models can not properly adapt to the present research of spatiotemporal dynamic urban system,the bottom-up micro-simulation models using individual data have gradually become a novel perspective for investigating urban systems recently.However,one of the factors restraining the wide application of micro-simulation models is the limited individual data due to the difficulties of data fetching and processing.Such a situation is especially serious in China,where individual dataset is not available from the official census and can only be obtained via various surveys in small scale conducted by separate units.This paper proposes a new solution to retrieve individual dataset from aggregate dataset,e.g.statistical data,for urban micro-simulation under the current sparse-data environment.Based on the existing multi-resource official statistical data,non-official surveys,as well as general relationships among individual attributes,our approach can disaggregate the individual attributions and location obeying the input aggregation observations data.This approach has proved to be significantly accurate at the statistical level,and can be conveniently adopted for urban micro-simulation under the present statistical institutions of China.