Matching results between landmark in different sources and landmark in a referenced dataset (BDTOPO)

The four datasets represent the results of a two sequentials processus. The first processus consists on a automatic matching between landmark in different sources and landmark in a referenced dataset (french national topographic data: BDTOPO). Then the links 1:1 are manually validated by experts in the second processus. The four different datasets and the BDTOPO dataset are archived here. The data matching algorithm is described in this paper. Each file represents the result matching for features belonging to a data source with: - the name of file depends on the data source - id_source corresponds to the identify of the landmark in data source - types_of_matching_results describes the type of result matching : « 1:0 » means that a landmark from a data source (e.g. Camptocamp) has no homologue landmark in BDTOPO « 1:1 validated » means that a homologous feature exist in BDTOPO and the link was validated « 1:1 non validated » means that the matching link was not validated « without candidates » represents the non-matched landmarks because there are no candidates in BDTOPO or because the landmark in data source is far away from its homologous in BDTOPO « uncertain » : uncertainty cases are complex cases where any decision is taken by the data matching algorithm - id_bdtopo corresponds to the identify of the landmark in BDTOPO if and only if there is a validated matching link <br>