Towards adding Linked Data to Ontology Learning Layers
Meisam Booshehri, Peter Luksch · 2014
Manual creation of ontologies is a time-consuming, costly and complicated process. Consequently, over the past decade a significant number of methods have been proposed for (semi)automatic generation of ontologies from existing data, especially textual ones. However, there are still significant limitations in this area. This study is an early effort towards reusing the semantic knowledge freely available in Web of Linked Data to improve the results of ontology learning from text in terms of multilingual making of ontology terms, classification of dangling instances, recommending appropriate intensions for ontology concepts, and concept hierarchy enrichment. Actually, these are the tasks associated with the second, third and fourth layer of Ontology Learning Stack. In our first stage experimental efforts, Factforge was used to implement the research objectives. Then, the results gained by an expert were compared against those obtained automatically. Finally, the experimental results verified the importance of the proposed approach through the achieved improvements in most of the objectives mentioned.