Ontology alignment for internet of things

Abstract

Semantic ontology development has become a research pursuit due to the diversity of entities and continual developments within and between collaborating entities. Existing ontology matching tools continue to become obsolete more rapidly than ever before. Therefore, the development of newer ontologies is an important aspect in ‘smart’ applications, such as those based on the IoT framework. Since smart applications require collaboration amongst a diverse set of technologies, ontologies need to be developed that help align all facets involved in such applications. The main challenges facing ontology development and matching are a lack of protocols for the interoperability of devices for man-ufacturers and continual development in terms of capabilities of the devices as well as the need and requirement for newer devices. The essence of ontology mapping and alignment lies in the ability of the agents in IoT applications to relate to datasets through semantic correspondences between the devices’ and agents’ ontologies. The OAEI (Ontology Alignment Evaluation Initiative). measures mappings using various measurement frameworks and calculations to enable evalua-tion by comparing the value of Recall, Precision, and F_score. The OAIE has been widely recognised as an efficient approach for evaluation. In this project, emphasis is placed on evaluating the efficacy of a proposed ontology match-ing tool tested by the benchmarks set by the OAEI. This evaluation, using known and acceptable standards, helps rate the matching tool against existing developments and determine the contribution offered by them. I used two approaches to measure the similarities between two ontologies: edit dis-tance and Smith Waterman algorithms. The testing was carried out via semantic comparison as well as simulation to help evaluate the various factors prescribed by the OAEI measurement and benchmarking standards against other tools.