Online geocoding services: a benchmarking analysis to some European cities
Conference at Baltic Geodetic Congress 2017, Gdánsk University of Technology, Poland 22-25 June 2017, Paper: Online Geocoding Services: a benchmarking analysis to some European cities. Author: Gianfranco Di Pietro, Fabio Rinnone – meters Srls Italy
— Abstract no. 14 (p.135 of Book of Abstract) —
The geocoding problem:
Geocoding is the computational process that allows to transforming an address of a place into a location on the Earth’s surface using geographic coordinates. The quality of geocoding have a relevant importance in spatial analysis. Furthermore, in recent years we have seen an enormous spread of Web technologies that have involved a growing number of users. At the same time have been born many Online Geocoding Service that supports the increasingly webGIS consumer applications. It has been proposed efficient algorithms to dynamically determine which Geocoding Service is more accurate .
Gfgeocoder a tool for benchmark
For these reasons, we present Gfgecoder, a tool based on a methodology capable to efficiently benchmark Online Geocoding Services. Gfgeocoder analyzes geocoding result obtained by some Online Geocofing Services, such as Google Geocoder, MapQuest and Open RouteService. The tool is written in Java language and it allow to make parallel requests to the Online Geocoding Services, passing in input the files containing strings of addresses. We used as input results of geocoding strings of addresses, published by some European Municipalities in their official open data portals. Input files are formatted as couples of type (address; WGS84 coordinates). Gfgeocoder is a multiplatform tool executable from command line and it can be runned as a background daemon in Windows and Unix-like servers.
Results obtained by Gfgeocoder are stored in tabular mode, in CSV format. At the end of the phase of geocoding, we done appropriate analyzes based on geometric and geodetic comparison of the distances between the strings of the coordinates obtained from Online Geocoding Services and the coordinates provided by Municipalities. In particular, we propose a benchmarking analysis performed using a parameter named “GA10”, defined as “percentage of addresses that are less than 10 me ters from the coordinates provided by Municipalities”
 National Criminal Justice Reference Service, “Spatially enabling the data: What is geocoding?” [Online]. Available: https://www.ncjrs.gov/html/nij/mapping/ch4_3.html.
 P. A. Zandbergen (2009). “Geocoding quality and implications for spatial analysis”, Geography Compass, 3(2), 647-680, DOI: 10.1111/j.1749 – 8198.2008.00205.x.
 H. A. Karimi, M. H. Sharker, D. Roongpiboonsop it (2011), “Geocoding recommender: an algorithm to recommend optimal online geocoding services for applications”. Transactions in GIS, 15(6), 869-886.
— 978-1-5090-6040-5/17 ©2017 European Union —