The analysis of movement data can reveal rich information about animal or human behaviour in different environments. Human population is increasingly urbanised1 and in cities, people spend large portions of their time indoors. Human indoor movement patterns and indoor environments display characteristics different to the movement of other animals and natural environments, respectively, hence the need for tailored analytical approaches and methods. The specific characteristics of the movement are further emphasized by differences in data collection techniques in indoor environments, and consequently the data collected.
In this paper, we introduce some of the major challenges in real-life indoor movement data collection impacting on the analysis of patterns mined from such data. Our examples are drawn from a large dataset of Wi-Fi-based tracking of visitors to large retail spaces (shopping malls) in two major Australian cities. The overall focus of our project (currently in its early stages) is the mining of contextualised indoor behavioural patterns that would enable improved product recommendation (Kantor et al. 2011) to visitors.