With the huge growing amount of information continuously produced, shared and available online, finding relevant and beneficial contents or services at a single or few clicks have become almost impossible. Most of the time, we will be returned with thousands of irrelevant web links. As such, a recommender system which recommends contents or services that likely meet the user’s needs is crucial, especially in the IPTV domain when the choices for program selection has no time and physical boundary restriction. In this research, we propose a system named COHRS, which is a context-aware recommender system based on ontological profiling in IPTV domain. Ontological approach improves user profiling process and thus improving the accuracy of a recommendation system. Experimental evaluations indicate that COHRS is able to overcome the drawbacks such as over specialization, data sparsity and inefficiency issue of most traditional recommender systems.