Commonality, i.e., high similarity, identifies the document sets or clusters being most relevant to the
domain of our interest (defined by the ontology). The hypothesis is that individual clusters having high
similarity with neighbouring entities are with high probability of the same domain. This hypothesis is
backed up by observed patterns of collocated terms within a domain, equally different domains have
different collocation pattern of terms. However, the similarity of clusters depends a lot on the quality of
the ontologies, especially the semantic distance between the entities. The result of this step is a domain
relevance score for each cluster of an entity with respect to the ontology.