Pengenalan Aktivitas Berkendara Berbasis Sensor Accelerometer dengan Fitur Auto Feedback Menggunakan Algoritma K-Nearest Neighbor

Abstract

Nowadays, the use of mobile phones is one of the causes of traffic accidents. It is very alarming considering the use of mobile phones is one of the human activities which nowadays are often performed. This study is trying to give a solution to reduce the causes of traffic accident caused by the use of mobile phone while driving by utilizing the accelerometer sensor contained on smartphones. Data on accelerometer sensor is classified using K-Nearest Neighboar algorithm to determine whether mobile phone users are driving or not driving. This study also make auto feedback application that can automaticaly reply the messages when mobile phone users are driving so that is expected to create safe driving conditions. In order to determine the accuracy of the algorithm to classify the activity of driving, then conducted testing on “auto feedback” application by sending several messages while mobile phone users are driving and when not driving.