In today’s digital landscape, social networks crucially shape perceptions and preferences but often facilitate echo chambers, limiting diverse thought and increasing polarization. This research addresses echo chambers on social networks, specifically Facebook, by developing diverse recommender systems using the MovieLens 100k dataset. We implement collaborative filtering, content-based filtering, and hybrid approaches to increase content diversity and reduce echo chamber effects. Our analysis confirms the presence of echo chambers and evaluates various recommendation algorithms’ effectiveness in mitigating these effects. We propose enhancements for recommendation algorithms, including incorporating serendipity and novelty, diversifying data sources, and integrating user feedback to promote diverse viewpoints. Our experiment shows that the clustering distance-based method performs best for both movie and social network datasets with diversity values of 0.84 and 0.56 respectively. This study enriches the literature on online echo chambers and suggests strategies for more inclusive recommendation systems to encourage open dialogue and understanding.