Journal of Computer Applications ›› 2026, Vol. 46 ›› Issue (6): 1904-1912.DOI: 10.11772/j.issn.1001-9081.2025060736

• Advanced computing • Previous Articles    

Intelligent recommendation model incorporating decision cost constraints and Lagrangian solution algorithm

Jinpeng YE1, Jiubing LIU2(), Zixing CHEN1, Jiaxin LIU3, Dun LIU4, Biao XU5   

  1. 1.College of Mathematics and Computer Science,Shantou University,Shantou Guangdong 515821,China
    2.School of Business,Shantou University,Shantou Guangdong 515821,China
    3.School of Management,Shanghai University,Shanghai 200444,China
    4.School of Economics and Management,Southwest Jiaotong University,Chengdu Sichuan 610031,China
    5.College of Engineering,Shantou University,Shantou Guangdong 515063,China
  • Received:2025-06-07 Revised:2025-10-10 Accepted:2025-10-11 Online:2025-10-30 Published:2026-06-10
  • Contact: Jiubing LIU
  • About author:YE Jinpeng, born in 2003. His research interests include intelligent recommendation.
    CHEN Zixing, born in 2004. His research interests include intelligent recommendation.
    LIU Jiaxin, born in 2002, M. S. candidate. Her research interests include strategic management.
    LIU Dun, born in 1983, Ph. D., professor. His research interests include granular computing and knowledge discovery, three-way decisions.
    XU Biao, born in 1981, Ph. D., lecturer. His research interests include intelligent optimization, dynamic multi-objective optimization.
    First author contact:LIU Jiubing, born in 1988, Ph. D., associate professor. His research interests include three-way decisions and optimization, intelligent recommendation, big data analysis.
  • Supported by:
    National Natural Science Foundation of China(62106135);National Undergraduate Training Program on Innovation and Entrepreneurship(202510560033);General Program of Natural Science Foundation of Guangdong Province(2026A1515011908);Shantou Humanities and Social Sciences Research Project(SK25008)

融入决策代价约束的智能推荐模型及拉格朗日求解算法

叶锦鹏1, 刘久兵2(), 陈子星1, 刘佳欣3, 刘盾4, 徐标5   

  1. 1.汕头大学 数学与计算机学院,广东 汕头 515821
    2.汕头大学 商学院,广东 汕头 515821
    3.上海大学 管理学院,上海 200444
    4.西南交通大学 经济管理学院,成都 610031
    5.汕头大学 工学院,广东 汕头 515063
  • 通讯作者: 刘久兵
  • 作者简介:叶锦鹏(2003—),男,广东江门人,主要研究方向:智能推荐
    陈子星(2004—),男,广东中山人,主要研究方向:智能推荐
    刘佳欣(2002—),女,山东聊城人,硕士研究生,主要研究方向:战略管理
    刘盾(1983—),男,重庆人,教授,博士,CCF会员,主要研究方向:粒计算与知识发现、三支决策
    徐标(1981—),男,安徽淮北人,讲师,博士,主要研究方向:智能优化、动态多目标优化。
    第一联系人:刘久兵(1988—),男,江西瑞金人,副教授,博士,CCF高级会员,主要研究方向:三支决策与优化、智能推荐、大数据分析
  • 基金资助:
    国家自然科学基金资助项目(62106135);国家级大学生创新创业训练项目(202510560033);广东省自然科学基金面上项目(2026A1515011908);广东省自然科学基金面上项目(2025A1515010167);汕头市人文社科研究项目(SK25008)

Abstract:

To address the problem that the existing intelligent recommendation do not consider decision cost constraints, an intelligent recommendation model incorporating decision cost constraints and a Lagrangian solution algorithm were proposed. Firstly, based on the user-item rating matrix, the SVD++ (Singular Value Decomposition Plus Plus) model was adopted to predict unknown ratings of users on items. Secondly, according to the predicted ratings, a single-objective optimization model of intelligent recommendation under decision cost and distribution diversity constraints was constructed. Thirdly, the distribution diversity constraint was relaxed into the objective function, and a Lagrangian relaxation model under decision cost constraint was established. Finally, a dual sub-gradient algorithm based on greedy strategy was designed to solve the constructed Lagrangian relaxation model efficiently. Experimental results on the MovieLens dataset show that compared with the Gurobi solver, the proposed algorithm reduces the solution time by at least 90.317% significantly, with the objective function value decreased by no more than 0.694%; compared with the LightGCN (Light Graph Convolution Network) method, the constructed model achieves higher recommendation accuracy on all test cases, and improves the distribution diversity on 77.8% of cases. The above fully verifies the comprehensive advantages of the proposed model and solution algorithm in terms of efficiency and performance.

Key words: intelligent recommendation system, recommendation accuracy, distribution diversity, decision cost, Lagrangian relaxation

摘要:

针对现有智能推荐未考虑决策代价约束的问题,提出一种融入决策代价约束的智能推荐模型及拉格朗日求解算法。首先,基于用户-项目评分矩阵,采用SVD++(Singular Value Decomposition Plus Plus)模型预测用户对项目的未知评分;其次,依据预测评分,构建在决策代价与分布多样性约束下的智能推荐单目标优化模型;再次,将分布多样性约束松弛至目标函数中,以建立决策代价约束下的拉格朗日松弛模型;最后,设计基于贪心策略的对偶次梯度算法,以实现所构建拉格朗日松弛模型的高效求解。在MovieLens数据集上的实验结果表明:与Gurobi求解器相比,所提算法在目标函数值仅降低不超过0.694%的情况下,求解耗时显著降低了至少90.317%;与LightGCN(Light Graph Convolution Network)方法相比,所建模型在全部测试样例上推荐准确性均获提升,分布多样性在77.8%的样例上得以改善。以上充分验证了所提模型与求解算法在效率与性能方面的综合优势。

关键词: 智能推荐模型, 推荐准确性, 分布多样性, 决策代价, 拉格朗日松弛

CLC Number: