Mapping to high-level ontology. How to map ontology learning results to high-level ontologies or how to construct high-level ontologies from fine-grained learning results is crucial to effectively organizing ontologies. Linking linguistic entities to ontology concepts and relations can be carried out either manually [107] or automatically [108]. For example, using path analysis (See Sect. 2), a concept that has a shorter path to other concepts can be placed higher in an ontology hierarchy. Another alternative is to treat the top-level concepts in an ontology as seeds and then to induce other related concepts in a bootstrapping manner until the desired depth is reached [40, 109]. The mapping problem becomes more complex when multiple ontologies are involved. Ontology merging or integration [110–112] remains a challenging task.