Enrichment of ontological taxonomies using a neural network approach[go to overview]
The taxonomy is a fundamental component of an ontology. In a taxonomy, classes are arranged hierarchically linked by a subclass-of relation. Complete taxonomies have exactly one most common class, called the root class. In Wikidata, the root class is the class "entity". The root class is unique, as it is the only class, which has no superclasses in the taxonomy. However, Wikidata’s taxonomy is incomplete in regard to this property. Orphan classes are classes, which are not the root class, but still do not have superclasses. Thereby, orphan classes violate the uniqueness of the root class.
This thesis consists of two parts. First, we extract and analyze Wikidata's taxonomy in regards to the orphan class problem. Secondly, we develop a hybrid algorithm combining neural word embeddings and classification methods to find an appropriate superclass for a given orphan class. The best-performing approach using a kNN classifier has an accuracy of 23.83%.
28.09.17 - 10:15