RECITE: A framework for user trajectory analysis in cultural sites

Abstract

The Internet of Things (IoT) has recently been applied in the domain of cultural exhibition enabling the cultural sites to provide more personal and proactive experiences to their visitors. To come up with valuable services, several solutions to analyze the spatio-temporal trajectories of visitors have been put forward. However, they neither consider the inherent uncertainty of the underlying indoor positioning technologies - Bluetooth Low Energy (BLE), RFID, etc. - nor other visitors' features apart from the spatio-temporal ones (e.g. the level of interaction with the museum displays). For that reason, the present work introduces RECITE, a framework to classify trajectories representing visitors' actions that copes with the aforementioned limitations of existing solutions. Firstly, RECITE states a novel mapping process for a BLE-based indoor positioning system to accurately detect the visitors' locations. On top of this mechanism, RECITE includes an ensemble of fuzzy rule classifiers able to tag the visitors' ongoing trajectories in real time considering both spatio-temporal and other behavioural factors. Finally, the framework has been evaluated in a case of use scenario showing quite promising results.