Socially Assistive Robot for Providing Recommendations: Comparing a Humanoid Robot with a Mobile Application

Socially assistive robotics has gained a valuable role in assisting, influencing and motivating the human behavior in many Human-Machine Interaction contexts. This finding suggested the use of social robots as recommendation interfaces. Despite many other types of recommending technologies exist (e.g., virtual agents, apps on cell phones, etc.), experimental studies convey that human beings result to be more engaged and influenced by the interaction with robots with respect to these other technologies. In particular, this has been shown by comparing social robots with virtual agents; a poor literature copes with the comparison between social robots and applications on mobile phones. To this extent, in this work, we address the comparison between these latest two technologies in the context of movie recommendation, where the two considered interfaces are programmed to provide the same contents, but through different communication channels. We provide the results of two experimental studies with the aim of evaluating the quality of the interaction from both the point of view of the application (by considering the users' acceptance rate of the recommendations) and of the users (by analyzing the users' evaluations) while interacting with the two interfaces. The main result arising from this study is that the social robot is preferred by users although, apparently, it does not change the acceptance rate of the proposed movies.