Purpose: This study seeks to deliver accurate, customized movie recommendations using the Sort Filter Skyline (SFS) algorithm. The approach considers factors like budget, box office earnings, popularity, runtime, and audience ratings to align closely with each user's specific preferences.
Methods: The Sort Filter Skyline (SFS) algorithm is employed, designed to identify and recommend items different from others within the dataset. Initially, the data undergoes preparation through pre-processing before being analyzed to compute entropy using the entropy formula. Before carrying out the dominance test, the SFS algorithm organizes the data based on entropy values.
Result: In this research, 176 skyline objects were identified from a dataset containing 4,803 movies, including well-known titles like "Avatar" and "Titanic." The Skyline Filter Sort (SFS) algorithm pinpointed these objects within 4 seconds. Additionally, evaluation results using synthetic data, as depicted in the data visualization, revealed that the number of attributes increased from 1 to 7. The dataset size grew, and the execution time also rose—from 18 seconds to 170 minutes. Despite this increase, the algorithm demonstrated efficient performance with optimized processing times.
Novelty: This study showcases the successful application of the SFS algorithm for generating personalized movie recommendations while tackling the difficulty of aligning viewer preferences with the extensive selection of films available. The findings offer important insights into enhancing recommendation systems by implementing algorithms efficiently and managing execution time complexity, contributing fresh perspectives to the field.