[ACM]
The next-generation smartphones should provide a signification improvement in terms of personalization. A major challenge in smartphone personalization is how to tackle the diversity in smartphone usage. This paper presents iLauncher, an intelligent quick launch that can adapt to individual usage patterns on the fly to change apps dynamically without user intervention. The rationale behind the development process is that we collected traces of real user activity and conducted statistical analysis to derive some observations that provide useful insights into the design of our algorithm. The results of experiments conducted to evaluate the proposed algorithm are encouraging in terms of prediction accuracy. We have implemented iLauncher based on the algorithm and released it in an Android marketplace for free usage.