Effectively mining time-constrained sequential patterns of smartphone application usage
Kuo-Wei Hsu · 2017
With the rapid development of mobile technology, today users have the freedom to download, install, and use various applications on smartphones. For system and application developers, it is certainly advantageous to better understand how users use the applications on their smartphones. To achieve this, we perform pattern mining on real-world data collected from tens of smartphone users for several years. We aim to mine the sequential patterns each of which satisfies a constraint on the maximum time interval between two adjacent application uses. However, we cannot mine all such patterns by first using the time constraint to filter the data and then using a general sequential pattern mining algorithm, neither can we do so by using the state-of-the-art implementation of the algorithm dedicated to mine sequential patterns satisfying the time constraint. In this paper, we present a solution to the problem of mining time-constrained sequential patterns with technical details, and we present the results that will be beneficial to the related studies.