RecoHybrid: hybrid ant colony optimization and FP-growth for feature selection in recommender systems

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

Recommender systems play a pivotal role in modern information retrieval, aiding users in discovering relevant content or products in a sea of choices. To enhance their effectiveness, this study introduces RecoHybrid, a novel approach that combines Ant Colony Optimization (ACO) for feature selection with Frequent Pattern Growth (FP Growth) for pattern mining in the context of recommender systems. In RecoHybrid, ACO is employed to identify a subset of relevant features from a high-dimensional dataset. This process efficiently reduces the dimensionality of the data, improving both computational efficiency and recommendation quality. Furthermore, the selected features are then subjected to FP Growth, a powerful pattern mining technique, to uncover intricate item associations and user preferences. The experimental evaluation of RecoHybrid demonstrates its effectiveness in improving recommendation quality across various datasets. Comparative analyses against conventional recommendation algorithms illustrate significant enhancements in recommendation accuracy, diversity, and coverage. Additionally, RecoHybrid exhibits impressive scalability, making it suitable for real-world recommender system applications. Our results suggest that RecoHybrid provides a robust and efficient means of optimizing recommender systems by combining the strengths of ACO and FP Growth Algorithm (ACO-FPGA). It harnesses the power of feature selection to reduce data dimensionality and pattern mining to extract meaningful insights, ultimately leading to more accurate and personalized recommendations. The experiments use the Movielens dataset for evaluation, and the results show the effectiveness of the proposed technique.