Hootenanny has the capability to derive type tag information automatically for elements that do not possess it based soley on the contents of their name tags. This capability can be used simply to add better tags to poorly typed data or to add tags with the goal of improving conflation matching between datasets.
This feature should not be used unless the addition of some false type tags to the input data is acceptable. There are no metrics yet for how many false tags this technique will generate on average, although the number is typically small and should become smaller over time as the feature is refined.
The default implicit feature tagger is ImplicitPoiPolygonTypeTagger, which will tag POIs and
building/POI related polygons. You can use any tagger class that inherits from
ImplicitTypeTaggerBase, however, it should be an implementation that works with the class used as
the element filter specified by the implicit.tagging.element.criterion when creating the implicit
tag rules database.
To use the feature from the command line with your input data:
hoot convert -D convert.ops=ImplicitPoiPolygonTypeTagger input.osm output.osm
To use the feature from the command line during conflation:
hoot conflate -D conflate.pre.ops="BuildingOutlineRemoveOp;MapCleaner;ImplicitPoiPolygonTypeTagger" \ input1.osm input2.osm output.osm
|
Note
|
In the above command, ImplicitPoiPolygonTypeTagger was appended to the default operations
for conflate.pre.ops configuration option. Check your version of configuration options to find the
default value for this option.
|
Features modified by the implicit feature tagger are marked with the tag
hoot:implicitTags:tagsAdded to let you know what tags were added to the data and which name tags
triggered the additions.