模糊环境下消费类电子产品推荐方法研究

Recommendation Method for Consumer Electronic Products under Fuzzy Environment

基于消费类电子产品自身的特点,提出一种新型的推荐方法。首先,考虑专家对备选产品进行评分时存在信息共享性,提出处理专家之间相互关联的直觉梯形模糊Shapley加权几何算子,并研究该算子的相关性质。其次,针对专家权重信息部分已知情况,建立规划模型求解专家的Shapley权重值,并运用所提算子对专家意见进行集成以得到各标准下各备选产品的综合评分。在给定顾客目标需求的基础上,基于投影法求出各备选产品在目标产品上的投影值,并将投影值最大的产品推荐给顾客以完成推荐过程。最后,以戴尔中国官网上的笔记本电脑购买为例,阐述该推荐方法的可行性和实用性。

A new recommendation method is put forward for consumer electronic products. Firstly, considering that there exists information sharing when experts evaluate the alternative products, the intuitionistic trapezoidal fuzzy Shapley weighted geometric operator is proposed to process the interactive characteristics between experts, and its properties are also investigated. Secondly, as information about experts＇ weights is incompletely known, a programming model is established to solve the experts＇ shapley weight values. Furthermore, the proposed oper- ator is used to integrate the experts＇ opinions to obtain the comprehensive score of each alternative product under each criterion. On the basis of similarity degree, a product with the largest projection value can be recommended to customers. Finally, a typical case of buying laptop on Dell China website is provided to further illustrate the feasibility and practicability of the recommending method.