Module extraction is an important reasoning task, aiding in the design, reuse and maintenance
of ontologies. Reasoning services such as subsumption testing and MinA extraction have been
shown to bene t from module extraction methods. Though various syntactic traversal-based
module extraction algorithms exist for extracting modules, many only consider the subsumee
of a subsumption statement as a selection criterion for reducing the axioms in the module.
In this dissertation we extend the bottom-up reachability-based module extraction heuristic
for the inexpressive Description Logic EL, by introducing a top-down version of the heuristic
which utilises the subsumer of a subsumption statement as a selection criterion to minimize
the number of axioms in a module. Then a combined bidirectional heuristic is introduced
which uses both operands of a subsumption statement in order to extract very small modules.
We then investigate the relationship between MinA extraction and bidirectional reachabilitybased
module extraction. We provide empirical evidence that bidirectional reachability-based
module extraction for subsumption entailments in EL provides a signi cant reduction in the
size of modules for almost no additional costs in the running time of the original algorithms. / Computer Science / M. Sc. (Computer Science)
| Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:unisa/oai:umkn-dsp01.int.unisa.ac.za:10500/4876 |
| Date | 08 1900 |
| Creators | Nortje, Riku |
| Contributors | Britz, Katarina, Meyer, Thomas Andreas |
| Source Sets | South African National ETD Portal |
| Language | English |
| Detected Language | English |
| Type | Dissertation |
| Format | 1 online resource (vi, 109 leaves) : ill. |
Page generated in 0.0013 seconds