DUDE, WHERE'S MY HOME? IDENTIFYING PLACES WITH GPS DATA

Purpose: Our research focuses on understanding and influencing health behaviours to promote healthier lifestyles. A key component involves analysing individuals' movement patterns to identify opportunities for personalised interventions. This abstract presents a novel approach that utilises GPS and GIS data to extract valuable insights into participants' behaviours without relying on traditional questionnaires. By analysing participants' location data, we can identify frequently visited places, travel patterns, and potential activity levels. This information can be used to infer the purpose of trips, such as work, leisure, or healthcare appointments, providing a deeper understanding of participants' daily routines. We aim to demonstrate the potential of GPS and GIS data for health behaviour researchers and highlight the valuable information that can be gleaned from this data alone. This contribution is aimed at identifying places. Material & Methods: Our proprietary algorithms, which will be publicly available soon, were employed to process and analyse GPS data collected with the open-source mobile application Traccar. Geographic information was sourced from OpenStreetMap (OSM). Results: The analysis of location detection and trip purposes needs to be conducted. Previous research (yet to be published) on a different dataset (n = 40) evaluated trip and transportation mode detection. Trip detection achieved 79% precision, 84% recall, and 81% F1-score, with an overall accuracy of 93%. Transportation mode classification was most accurate for bicycling (over 90% for all metrics), followed by vehicles (over 85%) and walking/running (over 75%), resulting in an overall accuracy of 90%. Conclusion: Using GPS data, we can gather detailed information about people's movements. This data can help us understand individual behaviours and develop personalised strategies to promote healthier lifestyles by encouraging positive changes in movement patterns.