User Behaviour Analysis in a Simulated IoT Augmented Space * David Massimo Free University of Bozen-Bolzano Bolzano, Italy damassimo@inf.unibz.it Elena Not Fondazione Bruno Kessler Trento, Italy not@fbk.eu Francesco Ricci Free University of Bozen-Bolzano Bolzano, Italy fricci@unibz.it ABSTRACT In this paper we present a demo application aimed at support- ing the research in the field of tourism and mobility support in IoT augmented areas. The application collects tourists’ choices while browsing Points of Interest (POIs) descriptions through a map-based interface that simulates user movement between POIs. Collected observations serve two purposes: the computation and testing of recommendation strategies for POIs (both for on-line and off-line studies); the generation of simulated users’ behaviour under alternative scenario and context conditions (e.g., weather, or the presence of a novel POI). Author Keywords Simulation Environment; IoT; Recommender Systems. INTRODUCTION Internet of Things (IoT) enables novel types of interactions with wireless sensor networks. It is now becoming a main- stream approach to the development of tourism and mobility applications [2]. In this demo we present an environment for simulating interactions with an IoT enabled area. By using the system, users can move in a virtual space that represents the real world. The system has been developed to: (1) overcome the lack of available and suitable datasets for studying decision making in such scenarios and (2) perform early testing before the actual deployment of an IoT network, such as that com- posed by a collection of beacons signalling the user proximity to a POI. More precisely, the goals of the designed system are to: • Register context-dependent decisions taken by users in a simulated IoT enabled area. * The research described in this paper is supported by the project Suggesto Market Space, funded by the Autonomous Province of Trento. Permission to make digital or hard copies of part or all of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for third-party components of this work must be honored. For all other uses, contact the owner/author(s). IUI’18 Companion March 7–11, 2018, Tokyo, Japan © 2018 Copyright held by the owner/author(s). ACM ISBN 978-1-4503-5571-1/18/03. DOI: https://doi.org/10.1145/3180308.3180316 • Collect user’s feedback (e.g., like/dislike) on POIs in order to compute and evaluate recommendation strategies in both on-line and off-line settings. • Evaluate the usability and effectiveness of a mobile app that supports the user (tourist) in the navigation of a real physical space augmented with IoT devices. • Simulate users’ decisions in a range of alternative and pos- sibly new contextual settings. The designed system supports the research on user behaviour learning in IoT enabled areas by providing a controlled envi- ronment that allows conducting large scale simulations. Stud- ies on user behaviour learning in IoT augmented spaces have been recently proposed [6, 3]. They have focused on indoor environments with a limited set of IoT augmented POIs. Con- versely, the proposed system makes possible to deal with a larger and outdoor area where several IoT devices are located. In addition, external factors that may influence user behaviour (e.g., weather conditions) have not been previously considered. In [5] we discussed the expected benefits of the exploitation of context in user behaviour learning and simulation as well as in the generation of recommendations. USER BEHAVIOUR OBSERVATIONS AND FEEDBACK The interaction process with the proposed simulation environ- ment unfolds as follows. A user starts a simulated visit to a city by providing demographic information (e.g., his age and gender), his POI preferences (e.g., specific museums) and topic preferences (e.g., culture or relax). Then, the system proposes an itinerary, which is visualised over an interactive map, in the form of a path optimised according to the stated preferences. From now on the user can simulate movements along the proposed itinerary. He can decide to visit the next POI in the itinerary or deviate from it and visit other POIs located in the proximity of his virtual location. To let the user perceive the cost/time required for moving from a place to another place, the system animates the movement of a map placeholder. During the user decision making process (to visit a POI) the system notifies him the current simulated contex- tual conditions (e.g., weather, daytime, crowdedness), so that he can take them into account when deciding to visit a POI. As soon as the user reaches a POI, a pop-up window, which emulates the GUI of a mobile app, illustrates the reached place with information and media as it would happen when the visitor approaches the beacon associated to the POI (Figure