Determining Geographic Representations for Arbitrary Concepts at Query Time. Henrich, A. & Lüdecke, V. In pages 17-24.
abstract   bibtex   
In typical concept@location-queries, the location is sometimes given by terms that cannot be found in gazetteers or geographic databases. Such terms range from names of vague geographical regions to more general terms such as mining or theme parks. In the present paper we describe our approach to deal with such vague location specifications. Roughly spoken, we derive a geographic representation for these location specifications from the top documents resulting from a query using the terms describing the location. These documents are parsed for toponyms and from these toponyms the region is derived. In the paper we describe an efficient process to derive the geographic representation for such situations at query time. Furthermore we present experiments depicting the performance of our approach as well as the result quality. Our approach allows for an efficient execution of queries such as camping ground near theme park or cycle path near brewery. It can also be used as a standalone-application giving a visual impression of the geographic footprint of arbitrary terms.
@inproceedings{ hen08a,
  crossref = {locweb2008},
  author = {Andreas Henrich and Volker Lüdecke},
  title = {Determining Geographic Representations for Arbitrary Concepts at Query Time},
  pages = {17-24},
  abstract = {In typical concept@location-queries, the location is sometimes given by terms that cannot be found in gazetteers or geographic databases. Such terms range from names of vague geographical regions to more general terms such as mining or theme parks. In the present paper we describe our approach to deal with such vague location specifications. Roughly spoken, we derive a geographic representation for these location specifications from the top documents resulting from a query using the terms describing the location. These documents are parsed for toponyms and from these toponyms the region is derived. In the paper we describe an efficient process to derive the geographic representation for such situations at query time. Furthermore we present experiments depicting the performance of our approach as well as the result quality. Our approach allows for an efficient execution of queries such as camping ground near theme park or cycle path near brewery. It can also be used as a standalone-application giving a visual impression of the geographic footprint of arbitrary terms.}
}
Downloads: 0