Hybrid Approaches Based On Collaborative Filtering To Recommending Mobile Apps

With the rapid emergence of mobile devices,smart phones have penetrated into every aspect of people's daily life.The explosive growth of mobile applications makes it difficult for mobile users to find suitable and interesting applications.Many application markets provide keywords search functions and recommend applications with high download counts.However,downloading an application is a vague indicator of the user's preference degree over that application,as the user may probably uninstall that application immediately after it has been installed.Mobile app recommendation has been explored by many researchers and some industry solutions are proposed for mobile users.Some industry solutions involve users' personal data such as social networking,but we only use users' application usage history.In this paper,we propose three hybrid models based on collaborative filtering to make mobile app recommendations.The first model(IUCF)leverages user-factor matrix,which is the intermediate product of latent factor model,instead of sparse user-item-ratings matrix to compute the similarities between users.The second model(IICF)is similar to the IUCF,but it leverage item-factor matrix to increase the accuracy of similarities between items.The third model(HLN)sums the predictions of the latent factor model and item-oriented approach,thereby capturing the advantages of both approaches.Our experiment results over 6,568 applications and 25,302 users clearly show that the hybrid models have better performance than both neighborhood model and latent factor approach.