Learning Concepts, Taxonomic and Non-
Mehrnoush Shamsfard · 2006
This paper discusses theknowledge extraction process in an ontology learning systemcalled Hasti. Itexploits an automatic, hybrid,symbolicapproachto acquire conceptual knowledge andconstruct flexible and dynamicontologies fromscratch. This approach starts froma smallkernel andlearns concepts,taxonomicand non-taxonomic relations andaxiomsfromnatural language texts. Thefocus ofthis paper ison extraction of concepts andconceptual (taxonomic andnon- taxonomic) relations usinglinguistic and template-driven methods Inthis paper Iwill first present abrief overview on ontologylearning systemsand then describing thelifecyclefortheontology learning andbuilding process inHasti, the knowledge extraction processwillbediscussed inmore details. AtlastI willpresent some experimental results ofimplementation and testing theproposed model.