Research on App Usage Behavior Analysis and Prediction Algorithms for Mobile Users

This paper systematically solves the problem of Mobile App Usage(MAU) prediction by means of predictability proof, prediction algorithm development, data collection and experimental verification. The MAU sequence is corresponded to the output sequence of the source, and the calculation method of the upper and lower limits of predictability is deduced. Then, an improved algorithm based on Prefix algorithm is put forward for mobile behavior pattern extraction, and a prediction algorithm based on support vector machine(SVM) is also proposed. The experimental results show that the time cost of SVM-based prediction algorithm is less than that of traditional prediction algorithm, so it is more suitable for real-time scenarios.