Online Spatial Evaluation of Residential Livability Based on POI Data Mining and LMBP Algorithm

摘要

For the cost-effective, real-time, and accurate evaluation of residential livability, this study obtains and examines 20 factors from relevant dimensions-the medical, educational, traffic-related, economic, and ecological environments-as a set to choose the location for livable residence through the spatial data mining of points of interest (POI) by using a kernel density estimation method. An online platform for evaluating residential livability is designed by using the ArcGIS Server. The Levenberg-Marquardt backpropagation algorithm (LMBP) is also designed by using the Gauss-Newton method to spatially assess residential livability. Such functions as real estate information queries, POI analysis, nuclear density analysis, statistical analysis, and the evaluation of residential livability are subsequently implemented. Gray relational analysis and the fuzzy analytic hierarchy process were used to verify the performance of the LMBP algorithm in terms of assessing residential livability. The results show that the proposed method can carry out a cost-efficient, quick, and accurate online evaluation of residential livability to provide reasonable choices of residential locations to users. The results of this study can help with decision-making relating to the creation of suitable spaces for dwelling, everyday life, and work.