A Context-aware application recommendation system for mobile device users

Download
2015
Bayram, Gamze
Development of smartphones and applications has opened a whole new world for mobile device users. Although this new world has many benefits due to a large diversity, regarding specific application domains, it is getting more complex day by day. In this study, a context-aware application recommendation system that recognizes the situation of users, predicts, and recommends the interactions that are likely to happen by the users in their specific context is developed. The proposed system is based on a hybridization of the Case-Based Reasoning and a Rule Based Reasoning approach that is derived from traditional association rule mining algorithms. Evaluation of the proposed model is done by using a real life dataset collected from individual records of four subjects. These four people were kept track for varying durations from approximately eight months to fourteen months. Results are encouraging when compared to that of previous studies in this domain. Therefore combining these two approaches provides an effective solution to the domain of recommendation systems. To the best of our knowledge, this kind of a hybrid approach has not been utilized in this domain.