New article published in “IEEE Access” journal

The Value of Geographic Locations Submitted by Internet Users


    • Dan Komosny
      • Department of Telecommunications, Brno University of Technology, Brno, Czech Republic
  • Miralem Mehic
      • Department of Telecommunications, University of Sarajevo, Sarajevo, Bosnia and Herzegovina
      • Department of Telecommunications, VSB-Technical University of Ostrava, Ostrava-Poruba, Czech Republic


Internet users may be asked to manually provide their contact details including city or full postal address. Examples include use of trial/free applications and services, filling out on-line surveys and petitions, and membership registration to loyalty programs. Many users may provide their correct location, whereas others may submit inaccurate data. The provided locations are used for geo-analytical purposes including the interpretation of survey results, on-line content personalization, and targeted marketing. In this paper we analyse differences in user behaviour when they provide their location. We work with two datasets of user-submitted locations. These datasets differ in how the locations were submitted: voluntary by purpose-aware users without being asked to or by requesting it from common Internet users. The locations from the purpose-aware users were about 2.5× more accurate than from the other users. We also present data for selected countries. The best result was found for the users in the USA with a median error of 44 km (the difference between the correct and user-submitted location). The results also show different user behaviour depends on the place from which they provide their location. The locations submitted from home were 1.5× more accurate than from the office.


User Location,Internet Geography,IP Address,RIPE Atlas


D. Komosny and M. Mehic, “The Value of Geographic Locations Submitted by Internet Users,” in IEEE Access, vol. PP, no. 99, pp. 1-1. doi: 10.1109/ACCESS.2018.2876973

IEEE Access

Journal Title

IEEE Access

Electronic ISSN




Impact Factor

3.557 (2017)