A Proposal for Creating Congruent Content of Different Media Types by Interactive Evolutionary Computation

We enjoy multiple media content in our daily lives, and using congruent content composed of different media types is ideal. However, obtaining congruent media content is difficult. Interactive Evolutionary Computation (IEC) is known as an effective approach to obtain good solutions for each user. However, conventional IECs treat single media types. This study proposes new Interactive Evolutionary Computation that searches the congruent media content in terms of a good combination of different types of media content. In the proposed IEC, the solution candidate contains different media types as its variables. A system was constructed by employing a genetic algorithm to investigate the effectiveness of the IEC with experiment. The target of creation was the congruent content composed of music and scent. Twelve participants evaluated a set of musical melodies and scents repeatedly. As results, a trend of increase in the mean fitness and significant increase in maximum fitness were observed in the search experiment. However, enough increase in the fitness was not observed in the evaluation experiment.