Analisa dan Penerapan Metode Algoritma K-Means Clustering Untuk Mengidentifikasi Rekomendasi Kategori Baru Pada List Movie IMDb

Abstract

IMDb (Internet Movie Database) is a comprehensive website that offers information about movies from all over the world, as well as various information about director, actor, actress, and writer biographies and award nominations. Visitors to the IMDb website can browse ratings and reviews based on the movies they plan to watch. Top 250 Movies and Most Popular Movies are two categories on IMDb. Because the results of the highest rating and the largest votes are only displayed based on the highest order of votes or ratings, the two existing categories are judged less useful and irrelevant to the suggestions for visitors to choose and decide on a film. This is due to the results of the highest rating and the most numerous votes, as determined by the highest ruling on either the votes or the rating. As a result of this, data mining with the K-means clustering algorithm is used to geolocate data in order to view data and accuracy using Davies-Bouldin Index (DBI) to combine ratings and votes with average approach to determine the centroid. Based on the results of this study, it is concluded that the DBI population with the highest accuracy is Cluster K=2 with population 509, with a score of 0.456, based on the voting and rating information, it can be deduced that a new category of movies called Best Recommended Movie is being recommended to potential moviegoers on the imdb.com website.

Stream Clustering for Selection Recommendations Using K-Means Algorithm: A Case Study in the Informatics Study Program

“…The information resulting from this analysis can be one of the options considered by study programs to provide advice to students regarding the choice of appropriate concentration paths. with their abilities [8].…”