Abstract
Reasonable Ontology Templates (OTTR) is a language designed to improve the efficiency and quality of building, using, and maintaining knowledge bases. OTTR introduces templates to create patterns of RDF triples. Templates are instantiated by template instances, which can be expanded to an RDF graph, and then stored in a triplestore. It is desirable to use the template instances to dictate the content of the triplestore. Currently, OTTR has no way of updating the triplestore when a change to the template instances occurs, besides rebuilding all triples. In this thesis, we have created several algorithms to more efficiently update the triplestore based on changes to template instances. Our algorithms aim to update only the parts affected by the change. First, we created a simple solution with excellent performance, but strict assumptions about input data. Then, we created other solutions that remove one or several assumptions. All solutions significantly outperform OTTR`s current solution in a typical use case. We investigate the performance of the different solutions and compare them to each other. The result of this thesis lays the groundwork for how updates can be part of OTTR in the future.