The KIWI project developed a platform called OntoExtractor enabling semi-automatic extraction of ontologies. This platform is composed of different modules. A first module extract typical instances from a data flow. A second module propose some topological relations among the detected classes. A third module assist the knowledge engineer in transforming classes and relations tentatively proposed by the system in a standard knowledge representation format. This paper address the specific problem of detecting is-a and part-of relations among typical instances of a document base.
|Titolo:||Detecting is-a and part-of relations in heterogeneous data flow|
CERAVOLO, PAOLO (Primo)
|Parole Chiave:||Bottom-up design; Fuzzy Logic; Knowledge Representation; Ontology|
|Data di pubblicazione:||2005|
|Digital Object Identifier (DOI):||10.1007/11552413_16|
|Tipologia:||Book Part (author)|
|Appare nelle tipologie:||03 - Contributo in volume|