The Semantic Geoweb: a case around "crowd sourced" geodata

From Master Projects
Jump to: navigation, search


About The Semantic Geoweb: a case around "crowd sourced" geodata

  • This project has been fulfilled.
  • This project fits in the following Bachelor programs: {{#arraymap:|, |xXx|bachelorproject within::xXx|,}}
  • This project fits in the following masterareas: {{#arraymap:Knowledge Technology and Intelligent Internet Applications, Technical Artificial Intelligence, Information Sciences|, |xXx|project within::xXx|,}}


Description

The semantic geoweb: a case around "crowd sourced" geodata


Postal codes in the Netherlands are not freely available. Some companies sell datasets with boundaries of postal codes (e.g. Geodan or postcode.nl). However, with all volunteered efforts to collectively open up and free datasets (good examples are wikipedia.org, openstreetmap.org or even movies---www.bigbuckbunny.org), also an effort was started to 'crowd source' the postal codes in the Netherlands---see: kvdb.net/projects/6pp.

The aim of this research project is to combine the information from this crowd sourced postal code dataset together with other open geographic datasets (such as the streets from OpenstreetMap, names from GeoNames, neighbourhoods from the CBS) to improve the location information of the postal codes. The current locations are far from perfect: If you put the current dump of the free 6 postal code database in a GIS and visually inspect the locations, a lot of points are at the same location and also other errors can be observed, mostly because the matching did take place at the municipal level, using the municipality name against GeoNames, and not at the streetlevel, for which the 6 places postal code was developed.

The main aspect to be considered in the project is the semantic description and enrichment of the original datasets and use of semantic technologies and platforms (such as RDF, linkedgeodata.org and reasoning): How do these techniques help you to solve the problem of getting the location for the postal codes?

The final postal code data (with location) should be published on the web using a diversity of services (e.g. web map service, web feature service, semantically enabled web services) and/or fed back into the original data sources.

Questions to be answered during this project: How to describe the quality of the point locations for the postal codes (correctness of the geocoding result)? Which semantic and geo-standards can be used to disseminate the data over the web? Can the results be used in the semantic web and how? How to deal with updates of the location of the postal codes in this distributed environment?

Supervisors: Martijn Meijers (TU Delft), Marian de Vries (TU Delft), Willem Robert van Hage (VU)