A Hybrid Recommendation System Based on Human Curiosity
Alan Menk dos Santos · 2015
Traditional recommendation systems use multiple computational techniques to perform personalized recommendations, and can consider the interests of users and even the context in which they live. However, they usually ignore each individual's personality factors, and hence, the recommendations generated overwhelmingly consider that all the users are identical psychologically. They ignore, for example, the curiosity level of each user, which may indicate that individuals with a high level of curiosity seek visit exotic locations and/or not yet visited by them, or even individuals with a low curiosity level tend to do the same things they did in the past, uninterested in new or different areas. Our paper presents a complete hybrid recommendation system considering the curiosity level of each individual as a decisive factor to recommend sites of South America. In order to prove the efficiency of our system in contrast to traditional recommendation systems, as well as to measure the satisfaction of users about the recommendations, we performed some preliminary experiments with the participation of 105 Brazilian volunteers. The first results indicate that considering the level of curiosity of a user increases the satisfaction with the recommendations.