Sistem Rekomendasi Anime Menggunakan Metode Simple Additive Weighting

Oleh Fachrul Rozi dan  Dewi Pramudi Ismi

Abstract:

Animated films (anime) are a product of the creative industry favored by many people. There are many anime sites that make it easy for people to get anime. However, because there are so many anime, it makes the audience confused what anime title to watch. Of the 30 questionnaires that have been distributed, 80% of respondents want a system that is able to recommend anime. In this study, an anime recommendation system was built based on the novelty of the release year, the rating value, as well as the similarity of genres and synopsis of anime currently watched by the audience. After getting the value of each parameter by normalizing the novelty of the release year and rating, dividing the current number of genres by the number of genres that are the same as other anime, and the synopsis similarity value using the Cosine Similarity method, the next step is to calculate all existing parameters using the Simple Additive method. Weighting. After calculating all anime, 5 anime with the highest score were taken to be used as recommendations for the anime that was currently being watched. To test the accuracy of the system, testing was carried out on 30 respondents to choose 5 anime from the 15 anime offered, and an accuracy of 84% was obtained. From the results of this accuracy, it can be concluded that this system is feasible to use.

Key Words: anime recommendation system, anime parameters, simple additive weighting