基于MapReduce的个性化推荐算法研究

he personalized recommendation, through which users are free from the vast information retrieval, is an important way to obtain information after search engine. Collaborative filtering, a simple algorithm, is able to handle complicate objects and have good reommendations, becoming the most successful and the most widely used technology in personalized recommendation. Content-based recommendation figures out data sparsity and cold start problems collaborativer filtering facing. On this basis, this paper proposes a hybrid recommendation algorithm Hybrid-TopN, which fuses collaobrative filtering and content-based recommendation and implements parallelly on Hadoop platform. The Algorithm sovles the bottleneck problems of traditional collaborative filtering algorithm, improves recommendation accuracy and recall rate and reduces the time complexity.