Cross-domain Information Fusion And Personalized Recommendation In Artificial Intelligence Recommendation System Based On Mathematical Matrix Decomposition

This article discusses a cross-domain information fusion matrix decomposition algorithm that aims to improve the accuracy of personalized recommendations in artificial intelligence recommendation systems. The algorithm utilizes various techniques such as Levenshtein distance detection, natural language processing, and graph convolutional networks to optimize potential feature vectors and integrate features from different fields. The results of the experiments show a significant improvement in score accuracy compared to the non-fused approach.