Personalized Recommendation Algorithm Based on Data Mining and Multi-objective Immune Optimization

To improve the accuracy of recommendation systems and user satisfaction, a personalized recommendation method combining data mining technology, convolutional neural network and multi-objective immune optimization algorithm is proposed in this paper. First, Pearson correlation coefficient is used to reduce the subjective bias of user ratings. Then, the convolutional neural network model is used to capture the long-term behavior pattern of users, extract deep interest features, and reduce the complexity of the model through ResNet connection. Finally, a multi-objective immune optimization algorithm is used to strike a balance between recommendation accuracy and diversity. The experiment was carried out on three datasets: MovieLens, Donation Dashboard, and Netflix. Compared with traditional algorithms, the average accuracy of the research algorithm on the three datasets was improved to 95.2%, and the root-mean-square error was less than 0.04. In addition, through multi-objective immune optimization, the algorithm significantly enhanced the recommendation diversity, with a hit rate of 0.3781 on the NetfAix dataset and a normalized discounted cumulative gain of 0.2349. The algorithm achieved stable performance in less iterations, and the recall rate was improved to 85%-95%, which was far better than other algorithms. The research method significantly improves the hit rate and normalized discounted cumulative gain value of recommendation results, providing users with more personalized resources © 2024 Slovene Society Informatika. All rights reserved.