A Tax-Map-Based Address Point Data Model for Geocoding Thai addresses

摘要

With a foreseen of significant geospatial data usage, several providers offer an online geocoding service that allows Internet users to convert a text address to coordinates with ease and no cost. However, the quality of their geocoding results is still uncertain, especially in developing countries like Thailand where the complexity of city plans and address formats are still opened issues. This paper proposes a tax-map based address point data model (TAPDM) which aims to improve geocoding quality of Thai addresses. This study developed a sample reference dataset of the TAPDM from tax map databases acquired from two municipalities in Phitsanulok Province, a study area. An address locator created from the prototype was used to perform geocoding in ArcGIS Desktop. Urban and rural addresses were sampled from the study area with one hundred each. The results were compared with those returned from four geocoding service providers: ArcGIS Online, Google, MapQuest, and Bing. The results revealed that the geocoding results of the TAPDM have the highest match rates in both urban and rural areas with 92% and 100%, respectively, and produced the highest positional accuracy in both urban and rural with the median of distance errors 2.65 meters and 11.68 meters, respectively.