Taxonomy learning is an important task for developing successful applications as well as knowledge obtaining, sharing and classification. The manual construction of the domain taxonomies is a time-consuming task. To reduce the time and human effort will build a new taxonomy learning approach named as TaxoFinder. TaxoFinder takes three steps to automatically build the taxonomy. First, it identifies the concepts from a domain corpus. Second, it builds CGraphs where a node represents each of such concepts and an edge represents an association between nodes. Each edge has a weight indicating the associative strength between two nodes. Lastly TaxoFinder derives the taxonomy from the graph using analytic graph algorithm. The main aim of TaxoFinder is to develop the taxonomy in such a way that it covers the overall maximum associative strengths among the concepts in the graph to build the taxonomy. In this evaluation, compare TaxoFinder with existing subsumption method and show that TaxoFinder is an effective approach and give a better result than subsumption method.

Taxonomy learning, ontology learning, TaxoFinder, concept taxonomy, concept graphs, similarity, associative strength.

[1] M.A.Hearst, “Automatic acquisition of hyponyms from large text corpora,” in Proc.14th Conf. Comput. Linguistics, 1992, vol. 2,pp. 539–545
[2] [2] F.M.Suchanek, G.Ifrim, and G.Weikum, “Combining linguistic and statistical analysis to extract relations from web documents,”in Proc. 12th ACM SIGKDD Int. Conf. Knowl. Discovery Data Mining, 2006, pp. 712–717.
[3] [3] E.-A. Dietz, D. Vandic, and F. Frasincar, “TaxoLearn: A semantic approach to domain taxonomy learning,” in Proc. IEEE/WIC/ACM Int. Joint Conf. Web Intell. Intell. Agent Technol., 2012, pp. 58–65.
[4] [4] W. Wang, P. Mamaani Barnaghi, and A. Bargiela,“Probabilistic topic models for learning terminological ontologies,” IEEE Trans.Knowl. Data Eng., vol. 22, no. 7, pp. 1028–1040, Jul. 2010.
[5] [5] Z. Kozareva and E. Hovy, “A semi-supervised method to learn and construct taxonomies using the web,” in Proc. Conf. Empirical Methods Natural Language Process., 2010, pp. 1110–1118.
[6] [6] P. Velardi, S. Faralli, and R. Navigli, “OntoLearn Reloaded: A graph-based algorithm for taxonomy induction,” Comput. Linguistics,vol. 39, no. 3, pp. 665–707, 2013.
[7] [7] K. Meijer, F. Frasincar, and F. Hogenboom, “A semantic approachfor extracting domain taxonomies from text,” Decision SupportSyst., vol. 62, pp. 78–93, 2014.
[8] [8] Y.-B. Kang, P. D. Haghighi, and F. Burstein, “CFinder: An Intelligent Key Concept Finder from Text for Ontology Development,”Expert Syst. Appl., vol. 41, no. 9, pp. 4494–4504, 2014.
[9] [9] Yong-Bin Kang, Pari Delir Haghigh, and Frada Burstein,”TaxoFinder: A graph-based approach for taxonomy learning.” Vol.28, no 2,2016.