DOMAIN INTEROPERABILITY FOR IOT APPLICATION THROUGH KNOWLEDGE EXTRACTION

Abstract

The recent IoT applications belong to various domains like smart city, weather forecasting, healthcare, transportation, etc. Generally, the domain knowledge is represented by a domain ontology which is designed by RDF, RDF-S, and OWL. A machine can understand the domain ontology, but the machine cannot understand the cross-domain knowledge in IoT applications because of interoperability issues. There are no standards available to describe all IoT-related transport domain terms. Due to a lack of interoperability, machines are not capable of processing various domain knowledge. In the proposed system, the ontological terms are converted into vectors using Google word2vector algorithms and the machine can process these vectors efficiently. The vectorized transport terms are clustered using their semantic similarity distance. Moreover, the contribution is made in terms of clustering in the domain of transportation. The transport domain terms are extracted based on 1000 research papers and 90 transportation domain ontologies.