OGIA: Ontology Integration and Generation Using Archaeology as a Domain

The need for generating ontologies for a highly specialized domain like archaeology is rare and is the need of the hour. It has vital applications in the Web 3.0-based search engines and indexes. This paper proposes the OGIA framework for ontology generation for archaeology as a domain of choice which works in the paradigm of outgrowing the dataset by incorporating auxiliary knowledge to TF-IDF. The framework also employs topic modelling frameworks, harvesting entities from Wikidata and DBpedia knowledge stores and generation of resource description framework (RDF) retaining its subject and object and further generating metadata. The metadata is automatized by using Bi-LSTM as a classifier. Datasets are also classified using the Bi-LSTM classifier. Semantic similarity and semantic relevance for semantic reasoning is evaluated using Horn’s index. Twitter semantic similarity, Jaccard index, and evolutionary algorithm help in optimizing initial solution sets to a much more optimal solution set for generating ontologies. Overall, a highest average precision percentage of 94.09%, highest average accuracy percentage of 95.09%, highest average recall percentage of 96.09%, and highest average F-measure percentage of 95.07% with the lowest FDR of 0.04 is achieved by the proposed framework.