With it, we have moved our theories for making recommendations to groups which have been proven in simulated environments, to the instantiation of our model in a real-life scenario: the social network Facebook. There are several reasons for this choice. Firstly, Facebook is used by users to create events and invite their friends to join the activity, so our system can help them in the organization of such events. And secondly, the user activity in the social network can be tracked to obtain information about her. We can obtain a lot of information from Facebook without having to bother our users with a lot of questionnaires and we also obtain a lot of feedback in order to improve our methodology.
To use our application, users only have to start their Facebook account and look for HappyMovie in the applications section. However, before being able to access to the movie recommendation results users fistly have to build their own “individual recommendation profile” which is necessary for our recommendation method. This profile is based on three different aspects:
personality, individual preferences and trust to other users. To obtain the personality and preferences, users must answer two different tests, the personality and the preferences test.
When the application has obtained the three factors that identify each user, the applications is ready to use and users are allowed to create events to go to the movies, invite other friends to those events or erase themselves from them.