Σημασιολογικά κατευθυνόμενη εξόρυξη γνώσης με χρήση οντολογιών
Πετράκης, Ανδρέας Δ.
The evolution of the Web in Web Semantic has motivated the development of standards and technologies for the expression and distribution of Ontologies. This makes them immediately a powerful tool that should be reviewed for use in various scientific fields, and in particular in that of Data Mining. Using a conception of representation through OWL Ontologies, using Reasoning Engines Analysis for operating the expressiveness of Ontologies and metrics for the emergence of semantic measures of distance / similarity, it is possible to achieve the integration of the data mining techniques for mining interestingness patterns. This integration is examined through the Association Rules algorithm A priori in Weka and evaluated in the field of terrorism. This approach is able to support cases in the form of specific rules that have been adopted by the experts using various statistical approaches.