The next weave of Mobile Ad hoc Networks and Wireless Sensor Networks stands for the Internet of Things which aims to build a new ecosystem integrating the physical world with the digital one. Such a concept combines smart objects and the Internet infrastructure to ensure the ubiquitous accessibility, monitoring, and management of IoT resources like sensors, actuators, appliances, places, people, etc. However, this new ecosystem is highly dynamic, heterogeneous with an endless number of incorporated entities, with the same or different properties, evolving in a context that changes continuously. These challenging features may considerably compromise the IoT solutions’ data interpretation, processing capacity, scalability, and the reuse and redundancy management of their entities. That is why we propose a lightweight, generic, and modular semantic model for designing IoT applications based on service composition. This model rests on Semantic Web Technologies and Service-Oriented Approaches. It provides a practical semantic description of IoT resources, such as sensors, actuators, services, environments, user requests, etc. It also meets the context-awareness and scalability properties and enhances the reuse and redundancy management of IoT entities. The proposed use-case scenario illustrates the interest, feasibility, and suitability of this model. Further conducted tests also show its performances and its high searching and querying ability.