Changes between Version 4 and Version 5 of generator_HotSpot
- Timestamp:
- 09/27/06 14:11:40 (19 years ago)
Legend:
- Unmodified
- Added
- Removed
- Modified
-
generator_HotSpot
v4 v5 9 9 This '''experimental''' AlignmentGenerator tries to tackle large scale ontology alignment, for which other approaches fail due to high memory and/or performance requirements. The general idea is a "divide and conquer" approach, continously aligning a number of parts of the ontologies and joining them it to a single, large alignment. 10 10 11 A basic concept, that is used within this approach, is the ''ContextOntology''. It defines a subontology by taking a single class of another ontology, called anchor, and adding all related classes within a certain distance to it. The relation currently used is the subclass realtion, but any other class to class relation could be used just as well. The distance used is to be called context ''depth''. The resulting set of classes is then interpreted as a new (sub)ontology. Every interface method of the original classes could be used, but return setshave to be truncated to the classes that are actually contained in the new ontology.11 A basic concept, that is used within this approach, is the ContextOntology. It defines a subontology by taking a single class of another ontology, called anchor, and adding all related classes within a certain distance to it. The relation currently used is the subclass relation, but any other class to class relation could be used just as well. The distance is to be called context ''depth''. The resulting set of classes is then interpreted as a new (sub)ontology. Every relation/property of the original classes could be used, but have to be truncated to the classes that are actually contained in the new ontology. 12 12 13 13 In this context, a ''Hotspot'' consists of the following attributes: 14 * a source, defining a small part of the original ontology to be aligned. Since this expects the ontology data as a PhaseLib:Ontology, special effort has to be taken 14 * a source ContextOntology 15 * a target ContextOntology 16 * an alignment between those two 15 17 16 The algorithm requires two exchangeable modules, defined by the following interfaces. 18 The general idea of the algorithm is now, to identify a Hotspot using single ''sure matches'' as anchors. Then, this Hotspot is optimised in size with respect to a hotspot quality measure. Finally, the resulting hotspot is joined with every other hotspot it overlaps (in both, source and target). 19 20 To implement this approach, the algorithm requires two exchangeable modules, which are specified by the interfaces HotspotIdentifier and HotspotMeasure. 21 22 17 23 18 24 == Characteristics ==