Deep Neural Network Empowered Movie Recommender System Using Hesitant Fuzzy Bi Objective Clustering

The movie recommender system is a highly influential and practical tool that assists individuals in efficiently choosing films to watch. Although recommender systems have been extensively used in academic research for various objectives, such as suggesting movies and recommending books, there has been a lack of focus on providing personalized movie recommendations for individual users. This research presents a new method for recommending movies that combines the Hesitant Fuzzy Clustering technique with a Convolutional Spiking Neural Network Movie Recommender System. The first phase entails obtaining input data from benchmark datasets such as MovieLens 100 K and MovieLens 1 M. These datasets are analysed using Ternary Pattern and Discrete Wavelet Transforms along with Hesitant Fuzzy Bi-objective Clustering technique to choose clusters depending on the retrieved attributes. After that, recommendation of movies uses a Deep Convolutional Spiking Neural Network to forecast user preferences. The efficiency of the proposed model is specifically compared to recent existing methods, like the Multi-Model Trust Based Movie Recommender Scheme (MT-ML-MRS) and the Graph-Dependent Hybrid Movie Recommendation Scheme (GHRS-MRS), particularly for movie recommendations. The results demonstrate a substantial enhancement, as the suggested model achieves 2.30% and 4.71% greater accuracy for MovieLens 100 k and MovieLens 1 M datasets respectively. The proposed system shows significant improvement over some traditional recommendation models indicating avenues for future research in scalable and intricate systems.