KASR: A Keyword-Aware Service Recommendation Method on MapReduce for Big Data Applications

摘要: Service recommender systems have been shown as valuable tools for providing appropriate recommendations to users. In the last decade, amount of customers, services and online information has grown rapidly, yielding big data analysis problem service systems. Consequently, traditional often suffer from scalability inefficiency problems when processing or analysing such large-scale data. Moreover, most existing present same ratings rankings different users without considering diverse users' preferences, therefore fails meet personalized requirements. this paper, we propose a Keyword-Aware Recommendation method, named KASR, address above challenges. It aims at presenting recommendation list recommending effectively. Specifically, keywords are used indicate user-based Collaborative Filtering algorithm is adopted generate recommendations. To improve its efficiency in environment, KASR implemented on Hadoop, widely-adopted distributed computing platform using MapReduce parallel paradigm. Finally, extensive experiments conducted real-world sets, results demonstrate that significantly improves accuracy over approaches.