A survey of people-centric sensing studies utilizing mobile phone sensors

Article type: Research Article

Authors: Bayındır, Levent

Affiliations: Department of Computer Engineering, Ataturk University, 25240 Erzurum, Turkey. E-mail: [email protected]

Abstract: Today’s ubiquitous presence of sensors provides a large amount of data which can be analyzed to study human behavior. The last few years saw the birth and diffusion of a new class of sensing systems: smartphones. With a diverse range of embedded sensors, smartphones have now become a commodity, and their capabilities can be leveraged to collect data to be used in different domains, including study of human behavior. This paper presents a review of past research works where mobile phone sensors are used to detect various aspects characterizing human behavior. Methods for automatic recognition of the placement of a mobile phone are first described as useful tools to improve the accuracy of sensing systems. Activity detection, at different abstraction levels from basic body motions to high-level activities, is then surveyed extensively, including studies focusing on detection of transportation mode and characterization of health-related activities such as physical exercise and sleeping. Other related works reviewed in this paper are continuous sensing systems for lifelogging applications, techniques to identify the environment where a user is located, and behavior modeling methods that allow extracting common patterns from behavioral data, studying psychological profiles and predicting future behaviors.

Keywords: Behavior, sensor, people-centric sensing, smartphone, activity recognition

DOI: 10.3233/AIS-170446

Journal: Journal of Ambient Intelligence and Smart Environments, vol. 9, no. 4, pp. 421-448, 2017