Application of Machine Learning in Movie Recommendation using Harris Hawks Optimization and K-means (HHO-k-means) Clustering

Abstract

In this study, a novel movie recommender system with Harris Hawks Optimization— k-means (HHO-k-means) clustering is proposed. The paper presents an empirical comparison of several clustering algorithms-k-means, PCA-k-means, SOM-Cluster, PCA-SOM, and HHO-k-means-across varying numbers of clusters. The performance metrics employed are Precision, Recall, Mean Absolute Error (MAE) and Root Mean Square Error (RMSE). The results show that the HHO-k-means algorithm consistently outperforms the other methods in terms of these metrics across all cluster sizes. It demonstrates higher precision, higher recall, lower MAE, and lower RMSE. Conversely, the PCA-k-means method generally exhibits less favorable results as the number of clusters increases. These findings suggest that the HHO-k-means algorithm may provide a more accurate clustering approach. © 2023, Ismail Saritas. All rights reserved.