No hints available.
select a node in the graph

Choose input

Match mappings from preloaded named graph against concepts in source and target vocabularies
Generate new candidates based on exact matching labels of source and target concepts.
Generate new candidates based on similar labels. The matcher is based on the 'isub' metric introduced in 'A string metric for ontology alignment' by Giorgos Stoilos, 2005.
A label matcher matching after compound splitting the label(s) of the source concepts. Source labels are splitted using rdf_tokenize_literal/2. Warning: source label tokens are matched against complete target labels!
Generate new candidates based on label matching after (snowball) stemming.
Generate new candidates based on similar numeric literals. The score is the absolute difference of the two numeric literals compared, or infinite if these are not numbers.
Generate new candidates by looking for 1 or more ancestors that have already been mapped.
Generate new candidates by looking for 1 or more descendents that have already been mapped.
Generate new candidates by looking for 1 or more related concepts that have already been mapped.
  • partition a set according to some criterium
  • typically resulting in selected and discarded subsets
Select correspondences with a unique source, target or both, discard others
Select correspondences that have the best score considering some numerical ranking, discard others.
Select mappings with corresponding mappings in the preloaded mapping, discard others with the same source/target.
Select mappings with the most matching labels, discard others for the same source/target. If type=all, all candidates with matching labels are selected. All matching is done after compound splitting the label(s) of the source concepts. Source labels are splitted using rdf_tokenize_literal/2. Warning: source label tokens are matched against complete target labels!
Select mappings with the most matching labels, discard others for the same source/target. If type=all, all candidates with matching labels are selected.
Select mappings with the most similar labels, discard others for the same source/target. If type=all, all candidates with sufficiently similar labels are selected. The matcher is based on the 'isub' metric introduced in 'A string metric for ontology alignment' by Giorgos Stoilos, 2005.
Select mappings with the most matching labels after (snowball) stemming, discard others for the same source/target. If type=all, all candidates with matching labels are selected.
Select the most generic concepts among alternatives for the same source/target, discard others.
Select correspondences that already have the most matching labels (according to their evidence data.)
Select correspondences that have been selected by the most other methods (according to their evidence data.)
Select mappings with similar numeric literals. If type=all, all candidates with sufficiently similar numbers are selected. The score is the absolute difference of the two numeric literals compared, or infinite if these are not numbers.
Partition a vocabulary based on the concepts having a specific property/value, or not.
Partition a vocabulary based on the concepts being in the sub-tree below (using BT/NT) a common parent concept, or not.
Partition a vocabulary based on the concepts being of a specific type, or not.
Partition a vocabulary based on the concepts having already been mapped, or not.
Partition a vocabulary based on the concepts being uniquely labeled, or not.
Component that randomly samples correspondences to create a new mapping.
Select the siblings (if any) among alternatives for the same source/target. This is used, for example, when a single source with two labels matches on two distinct but very similar targets and both mappings need to be preserved.
Select mappings with the most mapped ancestors, discard others for the same source/target. If type=all, all correspondences with one or more ancestors are selected.
Select mappings with the most mapped descendents, discard others for the same source/target. If type=all, all correspondences with one or more descendents are selected.
Select mappings with the most mapped related concepts, discard others for the same source/target. If type=all, all correspondences with one or more related concepts are selected.