International Congress on Recent Innovations in Engineering, Science and Technology - Istanbul - Turkey (2019-10-31)

Movie Recommendation System Based on Users’ Feature
In these generations, there is too much information that's easy for us misunderstanding because of different social media reports. In view of this, we want to establish a recommender system (RS) which could be based on users’ characteristic and directly provide the correct information for users. In order to collect the data-base, we focus on Movie topics. We choose the IMDb website which provide information related to films, television programs, home vedios, vedio games, and streaming content that also collecting many ratings and reviews from users. Hence, we use that information in our database. To build our Movie Recommender Sys-tem (MRS) based on users’ characteristic, we also analyze users’ individual data to extract users’ feature. Based on users’ feature, movie ranking/score, and movie revenue, we build our MRS model. We called this MRS model as MRS_F model. In the experiment, we use 5000 IMDb movie dataset and recommender 5 top movie for users. We also asked users’ to fill the questionnaires to make sure that our MRS_F can correct provide them the movie which they really want to see. The results show that 80% users are satisfy the system results, but 20% users think the results are not what they want. We conclude the results that the MRS_F should be included more features from users that will be closer to what the user wants.
Assist. Prof. Pei-Chun Lin, Mr. Shih-Feng Lin, Mr. Nureize Arbaiy