An Upper-Ontology-Based Approach for Automatic Construction of IOT Ontology

Abstract

Ontology gives us a reliable group of concepts and the relations between concepts in an IOT system. It does not only save words of format but also accurately transfers semantic data between human users and the computers. Hence, the usefulness of resources in IOT system depends on whether the domain ontology can be constructed effectively and correctly. In this paper we propose an automated method to construct the IOT ontology. First, we explain the necessity of introducing ontology automatic construction in IOT system and summarize the major challenges in existing approaches. Secondly, we introduced the existing ontology construction methods and summarize their issues. Thirdly, we give a framework of our ontology construction and research the key algorithms in detail: (1) knowledge-tuple extraction algorithm which contains contextual information; (2) concept semantic similarity algorithm which is based on the structure of tuple; (3) knowledge-tuple extraction model which is based on the structured information. Then we build a prototype and evaluate the ontology. Finally, we make conclusions and suggest directions for future research.

Towards a Semantic Web of Things: A Hybrid Semantic Annotation, Extraction, and Reasoning Framework for Cyber-Physical System

“…Längkvist et al [ 26 ] has surveyed unsupervised feature learning and deep learning models for time series modeling, which is feasible for high-level pattern recognition and finding states from streaming sensor data. Xu et al [ 27 ] propose an upper-ontology-based approach for automatically generating an IoT ontology. The method proposes an end-to-end framework for ontology construction based on calculating semantic similarity with input schemas with existing IoT ontologies, e.g., SSN.…”

Data Lifecycle and Tagging for Internet of Things Applications

“…Another relevant area of inquiry is the field of semantic annotation, which may help make data fusion more efficient [6,7]. Previous works propose manual and semi-automatic knowledge-based construction methods based on a unified semantic representation model for annotation.…”