Home // International Journal On Advances in Life Sciences, volume 4, numbers 1 and 2, 2012 // View article

Human Behaviour Analysis Using Data Collected from Mobile Devices

Authors:
Muhammad Awais Azam
Jonathan Loo
Sardar Khan
Muhammad Adeel
Waleed Ejaz

Keywords: Behaviour, Cell Tower ID, Bluetooth Proximity, Neural Networks, Jaccard Index, Decision Trees

Abstract:
Human behaviours are multifarious and myriad in nature. It is a challenging task to envisage and learn the human behaviour from daily routine activities. The profusion of wireless enabled mobile devices in daily life routine and advancement in pervasive computing has opened new horizons to analyse and model the contextual information. The aim of this research work is to infer the behaviour of low entropy mobile people using contextual data collected from mobile devices such as GSM location patterns (cell tower ID data) and Bluetooth proximity data. Both the GSM and Bluetooth data itself do not reveal much information about the behaviour of the users. Therefore, the challenge is to find out whether such data can infer human behaviour to understand and aid the unusual activities and routines of low entropy people such as elderly people and early stages of dementia patients. In this paper, a framework is created to analyse the contextual data for behaviour detection. There are four different steps in this framework to achieve the objective of the research work. In the first step, the contextual data is first classifies into different locations to obtain the movement patterns of the users. In the second and third step, a probability matrix and training data is obtained respectively, depending upon the user’s movement on daily and hourly basis. In the fourth step, a decision engine i.e. Neural Network (NN) and Decision Trees (DT) is used to detect the behaviour of the low entropy user. Results have shown that cell tower ID data gives behaviour of the user on high level scale for example movement patterns in GSM cells that does not help to identify any lower level activities such as attending the lecture, traveling in a bus. Whereas, Bluetooth data gives us more information about the lower level activities depending on the social relations and close proximity of other users.

Pages: 1 to 10

Copyright: Copyright (c) to authors, 2012. Used with permission.

Publication date: June 30, 2012

Published in: journal

ISSN: 1942-2660