Touch and Motion Sensor Data Collection on Smart Devices in Constrained and Unconstrained Environments

Nowadays, there are concerns about the effects of increasing smartphone use among children, which starts as early as age three. Children who watch more than two hours of television per day risk developing cognitive and social problems, according to the National Institutes of Health (NIH). The ability to categorize users based on age enables better control over children's smartphone usage, preventing access to inappropriate or unwanted apps. This precaution helps to create a safer, age-appropriate digital environment for children. This paper introduces a new touch dataset, which was collected from smartphone users and aims to categorize users according to their age. The main objective of this dataset is to estimate user age and implement age-based access controls in smartphone applications. Utilizing this dataset to allow or restrict access to apps based on the user's age. It can improve child safety by ensuring age-appropriate content and features. For this purpose, we created an app called “In_House,” which consists of various tasks that the user must complete, such as tap, swipe, type, and zoom tasks. This app not only records touch data but also sensor data while performing various tasks. Finally, this app collects each user's finger size. We collected data from 103 children aged 3 to 14 years old and 90 adults aged 15 to 70 years old. Furthermore, statistical features include mean, median, variance, and different kinds of machine learning algorithms to perform age classification based on the size of the user's fingers.