Tourists' emotion and satisfaction estimation : mitigating burden, cost, and privacy concerns

Details

Supervisors

Sadre, Ramin ; Yasumoto, Keiichi ; Matsuda, Yuki

Faculty

Ecole polytechnique de Louvain

Degree label

Master [120] en sciences informatiques, à finalité spécialisée

Abstract

The ever-growing amount of sensors and smart devices has lead to a growing interest in context-aware systems, to leverage the data collection possibilities these ubiquitous sensors can provide. Smart tourism is one of these research directions, where the aim is to recognise the state of tourists to provide them with a personalised sightseeing experience. With wearable sensing devices, it becomes possible to monitor tourist's behavioural cues and facial expressions, amongst other signals. By applying artificial intelligence and machine learning techniques, this data can prove useful to recognise their emotional status and predict their satisfaction level at given points in time. Such data is expected to offer beneficial insights regarding an individual's interests, and can serve to improve their overall sightseeing experience. This master thesis, realised in collaboration with the Ubiquitous Computing Systems Laboratory at the Nara Institute of Science and Technology, focuses on their previously devised EmoTour system, aiming to make it better-suited for real-world applications. EmoTour is a context-aware system, capable of estimating a tourist's emotions and satisfaction throughout sightseeing by combining multimodal sensing with predictive models. Two main directions are explored, the first of which centres on reducing the financial cost and physical burden of the system. The second direction aims to answer the privacy concerns such a system raises, and is examined through a series of different approaches. New experimental data was gathered to evaluate the applicability of the proposed changes to the system, and to further investigate differences between user subgroups.