Abstract—The proliferation of smartphones has become a ubiquitous platform for acquiring and analyzing data. Smartphones’ embedded sensors have become an effective source for human spatial and activity-based analysis. Machine Learning (ML) has made significant progress in learning features from these raw sensor data with high accuracy. However, domain experts, knowing ML, can apply machine learning techniques for various aspects. In this research, we have introduced—a smartphone sensor data collection and analysis platform for people in general who have little or no knowledge of machine learning but can avail the services of machine learning for their purpose. We have built an Android application for collecting sensor data and developed an Automated Machine Learning (AutoML) based web platform for data pre-processing, visualization, and analysis. Spatial analysis has been conducted on our AutoML based web application on GPS sensor data. We evaluated the most visited places of our app users using clustering techniques. The experiment shows that the DBSCAN clustering algorithm gives superior performance over K-means clustering for our spatial analysis on GPS sensor data.
 

Index Terms—smartphone sensor, AutoML, Android application, sensor data, GPS, spatial analysis, DBSCAN, K-means clustering

 

Cite: Kazi Taqi Tahmid, Khandaker Rezwan Ahmed, Moontaha Nishat Chowdhury, Koushik Mallik, Umme Habiba, and H. M. Zabir Haque, "An Integrated Crowdsourcing Application for Embedded Smartphone Sensor Data Acquisition and Mobility Analysis," Journal of Advances in Information Technology, Vol. 13, No. 5, pp. 503-511, October 2022.

Copyright © 2022 by the authors. This is an open access article distributed under the Creative Commons Attribution License (CC BY-NC-ND 4.0), which permits use, distribution and reproduction in any medium, provided that the article is properly cited, the use is non-commercial and no modifications or adaptations are made.