Utilizing Spatial Big Data platform in evaluating correlations between rental housing car sharing and public transportation

Abstract

Car sharing service for public rental housing in Korea, Happy Car, has increased the mobility for the residents and used as transportation welfare in the era of sharing economy. However, the car sharing service operator for rental housing have needed to improve the services more efficiently and access easily to public transportation but have difficulties in geocoding, analyzing and geovisualizing the large volume of disaggregate data in the services. Therefore, it needs a correlation analysis between rental housing car sharing usage patterns and public transportation accessibility by processing this big volume of data. In this study, we used the sharing car’s GPS data in the 45 rental housing districts in Seoul Metropolitan Area and the transportation card transaction data and analyzed the correlation between car sharing service and the supply level of public transportation. To handle the large volume of spatial location data and the card data during the analysis, we used Spatial Big Data platform to process this spatially referenced Big Data in a parallel distributed way. From the spatial analysis, the level of public transportation supply doesn’t affect the use pattern of sharing car in the rental housing and low income rental housing district households showed the longest moving distance. In terms of analytical functions of the Spatial Big Data platform, spatial computation through the platform such as the buffering analysis used for the accessibility analysis of the public transportation was found to be efficient in repetitive spatial operations of the massive amount of data.