Analysis of Actual Smartphone Logs for Predicting the User's Routine Settings of Application Volume

Although the volume settings of smartphones are important for users, they still need to push the hardware sound button manually. The purpose of our study is to improve the usability of volume settings. Our proposed method predicts the user's routine volume settings by learning the actual daily smartphone logs. Related works used suitable volume settings input by the experimental participants to learn the volume setting pattern for each user. In contrast, this study uses actual smartphone logs. This paper describes three results of the analyses of many actual smartphone logs. First, we investigate the rate at which the users change the application volume. Second, we examine the accuracy of the results predicted by our method. Third, we classify the test users as users for whom our method effectively works or others. Finally, we discuss the appropriateness of the predicted results for users' routine settings.