Diversifying Recommendations of Collaborative Filtering using Apriori Algorithm for Movies

Many algorithms are popular to generate personalized recommendations based on user preferences for movies. One such algorithm that is very famous nowadays is Collaborative Filtering. However, a common issue with collaborative filtering is the tendency to recommend popular movies, which often leads to a lack of diversity in the recommended items. To solve this problem, we propose a method for collaborative filtering that uses the Apriori algorithm to diversify movie selections. Apriori algorithm is a widely used association rule mining tool. Integrating it with the collaborative filtering framework we can find common itemsets that capture co-occurrence patterns among movies. The algorithm introduces diversity constraints based on the identified frequent itemsets to enhance recommendation diversity. These constraints ensure recommendations reducing the bias towards popular movies. Finally, the proposed method was evaluated on real-world movie datasets. The results demonstrate that incorporating the Apriori algorithm into collaborative filtering significantly enhances recommendation diversity, giving users more diverse and personalized movie recommendations.