Reflections on KOS based data integration

This paper briefly reviews two contrasting case studies by the authors on semantic data integration within the archaeology field and reflects on some of the key issues encountered, relevant to the NKOS Workshop theme on strategies for alignment of metadata to KOS linked data. Both projects involved diverse datasets with different schema and which employed different terminology. Both projects combined datasets with information extracted from archaeological reports via natural language processing (NLP). In both cases, the semantic framework that afforded data integration was a combination of metadata element sets organised by an ontology with a relevant value vocabulary (eg thesauri and term lists). Ontologies and value vocabularies have been seen as complementary resources for this purpose [1, 2].