Scholarlink

<p>The rapid growth of online platforms has amplified the need for efficient recommendation systems that personalize user experiences and enhance decision-making. Collaborative filtering (CF) is a widely used approach, but it often struggles with challenges such as data sparsity and the cold-start problem, particularly for new users or items. Singular value decomposition (SVD) effectively mitigates data sparsity, while demographic information helps address the cold-start issue by identifying users with similar behaviors. Meanwhile, quantum computing, with its novel computational paradigms, offers promising solutions to optimize recommendation processes. This paper proposes a classical-quantum hybrid recommendation system that integrates collaborative filtering, demographic data, and quantum-enhanced algorithms. Designed to generate personalized movie recommendations for both existing and new users, the system demonstrates superior performance compared to traditional and state-of-the-art collaborative filtering approaches. Empirical evaluations in simulated quantum environments reveal significant improvements in root mean squared error (RMSE) and mean absolute error (MAE), showcasing the potential of the quantum singular value decomposition (QSVD) based proposed system to address existing limitations and revolutionize recommendation systems.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Classical-quantum hybrid recommendation system based on collaborative filtering and demographics

  • 期刊论文

  • 发表日期：

    2025年7月3日

查看全文
  • Anindita Raychaudhuri,
  • Arijit Dey

摘要

The rapid growth of online platforms has amplified the need for efficient recommendation systems that personalize user experiences and enhance decision-making. Collaborative filtering (CF) is a widely used approach, but it often struggles with challenges such as data sparsity and the cold-start problem, particularly for new users or items. Singular value decomposition (SVD) effectively mitigates data sparsity, while demographic information helps address the cold-start issue by identifying users with similar behaviors. Meanwhile, quantum computing, with its novel computational paradigms, offers promising solutions to optimize recommendation processes. This paper proposes a classical-quantum hybrid recommendation system that integrates collaborative filtering, demographic data, and quantum-enhanced algorithms. Designed to generate personalized movie recommendations for both existing and new users, the system demonstrates superior performance compared to traditional and state-of-the-art collaborative filtering approaches. Empirical evaluations in simulated quantum environments reveal significant improvements in root mean squared error (RMSE) and mean absolute error (MAE), showcasing the potential of the quantum singular value decomposition (QSVD) based proposed system to address existing limitations and revolutionize recommendation systems.