Context-Enriched Regular Human Behavioral Pattern Detection From Body Sensors Data

• Main Authors: Walaa N. Ismail, Mohammad Mehedi Hassan, Hessah A. Alsalamah
• Format: Article
• Language: English
• Published: IEEE 2019-01-01
• Series: IEEE Access
• Subjects: Activity monitoring; Apriori; body sensors; FP growth; productive periodic frequent patterns; smart data
• Online Access: https://ieeexplore.ieee.org/document/8664150/

Extracting indicative characteristics from the sensor data provide diverse avenues for improving the well-being of the elderly people living alone in their homes through understanding and identifying their behavioral patterns while considering any environmental changes. In this paper, we present a new model to explore the challenges associated with mining patterns from the body sensor data and their potential use in discovering regular human routines through mining periodic patterns from a non-uniform temporal database. The non-uniform nature of the temporal database adds more challenges to the mining of periodic patterns as the items may have different periodicity and frequency occurrences. Another challenge is how to discover the correlation between the discovered patterns. In addition, we examine the context-enriched periodic patterns which provide more insights about residents' health. A new algorithm for the contextualized-correlated periodic pattern mining from a non-uniform temporal database is presented along with an extensive evaluation of its performance using a real-life dataset.