Hybrid Wizard of Oz: Concept Testing a Recommender System

In this paper we report the results from a Hybrid Wizard of Oz study consisting of a critical incident interview and an in-situ simulation. The study was aimed at validating the need for a contextualised and personalised Point-Of-Interest (POI) recommender and understanding the detailed user needs for it. Our key findings include: feeling bored as a key trigger to search for POIs, trust issues with the existing recommendation sources, intent to find free activities, information needs on areas-of-interest beyond points-of-interest, socialising support, and language barriers. With this study we also exemplify a cost and time effective approach for design of intelligent systems.