Building Ontology Profiles with Contextual Elements for Effective and Efficient Movies Recommendation

In coping with the increasing on-demand movies services provided through the Internet or Cloud platform, on-demand movies providers are competing intensively in providing more varieties and choices of programs. To retain and attract more users, service providers are moving towards recommending more personalized programs to satisfy users’ needs and preferences. While the number of users and items are increasing, the effectiveness and efficiency of the recommendation become the main factors to address. This paper proposes to integrate ontological profiles with contextual elements based on hybrid recommendation approach. In this proposed system, namely HyOC-RS, ontological profile is incorporated with contextual elements to improve the recommendation mechanism. A semantically and hierarchically-linked data model represented the proposed ontological user profile. Performance of HyOC-RS is evaluated in terms of time and space complexity. HyOC-RS with small error measures has proven to be more efficient and accurate as compared to traditional recommendation systems. Additionally, experimental results have shown that HyOC-RS could resolve the problems inherent in many of the traditional content-based and collaborative filtering recommendation systems such as over specialization, data sparsity, new user, and new item problems