Social Data Mining through Distributed Mobile Sensing

TL;DR

This paper introduces a distributed mobile sensing framework utilizing existing smartphone sensors to collect environmental and location data for data mining and activity classification, demonstrating initial promising results.

Contribution

It presents a novel distributed sensing framework leveraging existing mobile phone hardware for environmental data collection and activity analysis.

Findings

Preliminary data analysis shows potential for environmental and activity classification.

The framework successfully integrates with users' daily routines using embedded mobile sensors.

Future work includes expanding data collection and refining classification methods.

Abstract

In this article, we present a distributed framework for collecting and analyzing environmental and location data recorded by human users (carriers) with the use of portable sensors. We demonstrate the data mining analysis potential among the recorded environmental and location variables, as well as the potential for classification analysis of human activities. We recognize that the success of such an experimental framework relies on the adoption rate by its candidate user network; thus, we have built our experimental prototype on top of hardware equipment already embedded within the potential users' everyday routine - i.e. hardware sensors installed on modern mobile phones. Finally, we present preliminary analysis results on our collected data sample, as well as potential further work directions and proposed use case scenarios.