With the growing number of people using SmartTVs and Smartphones, designing a recommender system for on- demand streaming media, such as movie streaming, has been an attractive, yet challenging work. There are many factors that influence people to enjoy a movie. Smart devices provide many kinds of data from its sensors that can help us deduce, for example, whether it is the time, the day, the location, or the combination of those that makes a great experience in watching a particular movie. However, designing the algorithm to consider all these factors can lead into a very complicated decision tree. To address this issue, we propose a simple evolutionary computational approach that can be used to search through those huge numbers of possible combinations of solutions, and find the relevant factors when recommending a movie to particular type of users.