Predicting Human Psychological Factors from Movie Ratings and Genres Using Genetic Algorithms

Personality computing is emerging research in artificial intelligence and psychology. The genetic algorithms are efficient in optimization problems. The selection of features must be based on the problem in hand and can be achieved effectively using feature selection and genetic algorithms. This research focuses on feature selection using genetic algorithms to select the best subset of features and to psychological factors of human rather than his predictive behaviors in movies preferences. In this work, first the prediction of user's movie genres preferences from movie ratings. The feature selection genetic algorithm on movie preferences results high fitness features of movie genres. Then the genre preferences are mapped to big five factor model using Pearson correlations. The result concludes that the mapped preferences-traits are near optimum.