Context-aware smartphone application category recommender system with modularized Bayesian networks

The number of applications available since the late 2010's, and the number of smartphone user sharply increasing. However, not all applications are not useful or helpful. In other words, to obtain satisfactory results in the search can be difficult means. Users to find what they want to search for a many times. To solve this problem, previous studies have proposed the use of recommender systems. Most of the system uses age, gender, preference based collaborative filtering. Collaborative filtering has the problem that data sparsity, cold-start or needs lots of users' personal data. In this paper, we propose a smartphone context-aware application category recommendation. We use Bayesian-network to inference context and recommend the category when inference context and we have set the probability of using category from collected data. We tested our proposed system with F1 measure, accuracy of inference context.