Decoding user behaviour from Smartphone interaction event streams

Friedrichs, Björn 2023. Decoding user behaviour from Smartphone interaction event streams. PhD Thesis, Cardiff University. Item availability restricted.

Preview	PDF (Bjoern Friedrichs, PhD, Thesis) - Accepted Post-Print Version Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (18MB) | Preview
	PDF (Cardiff University Electronic Publication Form) - Supplemental Material Restricted to Repository staff only Download (217kB)

Abstract

The smartphone has become an everyday device for many people around the world and has led to an evolution in the way we use these devices. This has led to increased research interest in the effects of smartphone use on psychological traits, which could have a positive impact in clinical or self-help settings by identifying positively influencing variables. In this thesis, a new model to extract behaviour information from a stream of usage is presented. The model aligns with previous methods in the research area but focuses on establishing a generalisable three-step process of processing user interaction to extract new user behaviour knowledge. This introduces a structured approach to smartphone usage evaluation and enables the implementation of customisable applications. It also creates a baseline to compare previously defined metrics which describe smartphone usage. Usage derived from metrics which could be considered high-level such as screen-on time is self-evident and therefore are common measure to distinguish usage between users. However, within usage sessions, they suffer from limitations such as a strong skew towards short bursts of usage because of how smartphones are often used. By utilising direct interactions with the user interface (such as taps and scrolls), usage at a lower level can be considered which can carry more elemental characteristics of behaviour. Thus, they can be used to model behaviour more accurately, which can be aligned with the user’s mental state to identify habits which are caused by problematic use patterns. This enables the isolation of user trait classes reflecting smartphone addiction and impulsivity.

Item Type:	Thesis (PhD)
Date Type:	Acceptance
Status:	Unpublished
Schools:	Computer Science & Informatics
Subjects:	Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Funders:	School of Computer Science and Informatics 3 year stipend
Date of First Compliant Deposit:	1 November 2023
Date of Acceptance:	March 2023
Last Modified:	02 Nov 2023 10:07
URI:	https://orca.cardiff.ac.uk/id/eprint/163612

Actions (repository staff only)

Edit Item