The research in Recommender Systems has evolved considerably over the past years; however, to date the investigation on how emotions could be used to complement such technologies is sparse. This Master's thesis took the initial steps in the research of the inclusion of affective feedback in a recommender system for entertaining videos. The investigation here presented consisted on two related aspects of affective recommender systems: i) finding the factors that could condition the adoption of this innovation; and ii) designing and implementing a data collection process in the form of a web service, due to the lack of data resources for developing this system which was detected, was, resulting in the creation of a dataset with emotional information detected during the visualization of entertaining videos.

After analysing the dataset, it was found that there is a correlation between the emotions and the opinions (provided explicitly by the watchers) on the videos. The study of the adoption, conducted with an adaption of a well-known adoption of innovations model, concluded that the perceived enjoyment of the use of the system and the social influence are the two factors conditioning the most the intention of adoption. Regarding the trust on such a system, the expectancy that the system would use the detected emotions only for creating recommendations is the most influencing factor. The most relevant contribution of the project is the dataset which could be used in future research on the topic, together with the developed methodology and web application, which can be considered as an embryo of a future affective recommender system.