A recommendation algorithm based on density peak clustering and key users

At present, the research on recommendation algorithm is limited to how to improve the existing algorithm or design new algorithm, blindly pursue the accuracy and diversity of the recommended results, rarely study the importance of the individual user in the recommendation system, but study the importance of the individual user in the recommendation system can improve the recommendation efficiency and enhance the robustness of the recommendation system. Aiming at this problem, the key users determination method based on density peak clustering is proposed, which can effectively distinguish the key users of the recommendation system. Furthermore, the recommendation algorithm based on density peak clustering and key users is proposed. The proposed algorithm is efficient through experimental verification, which not only improves the recommendation efficiency, but also improves the recommendation accuracy and diversity.