A scalable semantic data fusion framework for heterogeneous sensors data

Data fusion is a fundamental research topic especially in the Internet of Things (IoT). A massive quantity of data is increasingly being generated by heterogeneous sensors which make data integration more difficult. A noticeable body of research has attempted to mitigate the incompatibility between the collected data to facilitate meaningful data integration between machines by using the semantic web technologies. However, there are still some critical issues including scalability and measurement unit conflicts. Therefore, this paper proposes a scalable semantic data fusion framework that aims at improving the scalability of data fusion and detecting and reconciling measurement unit conflicts. This framework is fully implemented to demonstrate its scalability during the process of data fusion, and its ability to handle measurement unit conflicts. Two experiments were conducted to evaluate the scalability and effectiveness of the proposed framework using real dataset that was collected from different sensors. To evaluate the scalability of the proposed framework, a set of queries was adapted and the average response time was calculated from the execution of every query. Whereas, the total number of the conflicts detected and resolved by the proposed framework were used to evaluate the effectiveness. Experimental results show that the proposed framework improves the scalability of data fusion among heterogeneous sensors’ data, and effective in detecting and resolving data unit conflicts.