Discovering self-quantified patterns using multi-time window models

Purpose A new research domain known as the Quantified Self has recently emerged and is described gaining self-knowledge through using wearable technology to acquire information on self-monitoring activities physical health related problems. However, very little about impact of time window models discovering self-quantified patterns that can yield insights. This paper aims discover multi-time models. Design/methodology/approach proposes a analytical workflow developed support streaming k -means clustering algorithm, based an online/offline approach combines both sliding damped An intervention experiment with 15 participants used gather Fitbit data logs implement proposed workflow. Findings The results reveal model exploring evolution micro-clusters labelling macro-clusters accurately explain regular irregular individual behaviour. Originality/value preliminary demonstrate they have finding meaningful patterns.