基于熵权的多粒度犹豫模糊语言VIKOR群推荐方法

发布时间：2018-03-19 10:25

本文选题：多粒度犹豫模糊语言术语集　切入点：VIKOR方法　出处：《控制与决策》2018年01期 　论文类型：期刊论文

【摘要】：针对群推荐系统中被推荐项目具有多粒度性、犹豫性、模糊性的特点,首先在多粒度犹豫模糊语言术语集的基础上,提出多粒度犹豫模糊语言熵的概念及计算公式,采用熵的计算公式求解被推荐项目的属性权重;然后,将传统的多准则妥协解排序法(VIKOR)拓展到多粒度犹豫模糊领域,并对其妥协解公式加以改进,将改进的VIKOR方法用于群推荐;最后,从理论分析、数值计算和敏感性分析3个方面将改进的VIKOR方法与逼近理想解排序法(TOPSIS)进行对比,所得结果表明了所提出方法在群推荐应用中的合理性和有效性.
[Abstract]:In view of the multi-granularity, hesitancy and fuzziness of recommended items in group recommendation system, the concept and calculation formula of multi-granularity hesitation fuzzy language entropy are put forward on the basis of multi-granularity hesitating fuzzy language terminology set. The entropy formula is used to calculate the attribute weight of the recommended items, and then the traditional multi-criteria compromise solution ranking method is extended to the multi-granularity hesitation fuzzy field, and the compromise solution formula is improved, and the improved VIKOR method is applied to group recommendation. Finally, the improved VIKOR method is compared with the approximate ideal solution sorting method from three aspects: theoretical analysis, numerical calculation and sensitivity analysis. The results show that the proposed method is reasonable and effective in the application of group recommendation.
【作者单位】：合肥工业大学管理学院;蚌埠学院计算机工程学院;
【基金】：国家自然科学基金重大项目(71490725);国家自然科学基金项目(71371062) 国家973计划项目(2013CB329603) 安徽省教育厅重点自然科学项目(KJ2015A300)
【分类号】：C934

【相似文献】