الفهرس 
Programming model for cloud computingOnly 14 pages are availabe for public view
 |  | from | 126 |  |  |
| | | from | 126 | | |
Abstract
The Internet of Things (IoT) and context-awareness IoT applications are recently involved in various domains such as smart homes, traffic, and healthcare. While developing context-awareness IoT applications, the main challenge is managing the massive amounts of data and events to get relevant context information. Context-awareness services are frequently applied in context-awareness IoT applications{u2019} domains. Therefore, there is a challenge while developing these services in a unified and a flexible way to fit different IoT application domains.In the cloud service model, there is a lack of concern about delivering the context information as a service. This is needed for managing and controlling data and events to deliver only the relevant context information to cloud services{u2019} users. This thesis proposes a domain-independent programming model that facilitates developing context-awareness services for various IoT application domains. The proposed programming model enables an adapted execution of the IoT services while hiding the context management details.The proposed programming model applies an ontology-based approach for context modeling, and it applies a rule-based approach for context reasoning.The proposed programming model aims to strengthen the IoT middleware architectural requirements and service requirements by utilizing the semantic technology.The middleware service requirements would improve both data and event management mechanisms. It would enable middleware architectural requirements: context-awareness, adaptability, abstraction, service-based, and semantic interoperability. We demonstrate these supported features by a detailed comparative study with the other related works. We demonstrate the proposed programming model support of various domains by developing two experiments on two different domains.Therefore, the effectiveness of applying the context-awareness services is evaluated by achieving an average F1-score of 0.82,comparable with the average F1-score of 0.73 of current state-of-the-art methods. Moreover, the incorporation of essential compound activities cannot be handled by previous works.