Operations Research and Management Science ›› 2017, Vol. 26 ›› Issue (3): 54-62.DOI: 10.12005/orms.2017.0057

• Theory Analysis and Methodology Study • Previous Articles     Next Articles

Determination of Product Attribute Weight and TOPSIS Method Improvement Based on the User Choice

LI Bao-zhen1, FENG Sheng-jie2   

  1. Department of Economics and Management, Jiangsu University of Science and Technology, Zhenjiang 212000, China
  • Received:2015-05-04 Online:2017-03-25

基于用户选择的产品属性权重确定及对TOPSIS法的改进

李保珍1, 封胜杰2   

  1. 江苏科技大学 经济管理学院,江苏 镇江 212000
  • 作者简介:李保珍(1975-),男,博士,副教授;主要研究方向:网络信息资源管理、社会计算与群体智能;封胜杰(1989-),男,硕士研究生,主要研究方向:信息管理。
  • 基金资助:
    国家自然科学基金面上项目(71273121,71673122);全国统计科学基金重点项目(2015LZ29);国家自然科学基金重点项目(71331003)

Abstract: Traditional TOPSIS(Technique for Order of Preference by Similarity to Ideal Solution)evaluation method is insufficient, for its weight is subjective, and updating the plan makes rank reversal. This paper tries to improve the determination of attribute weights based on objective data of the user participate in the evaluation and product attributes, through improving the standardization of data matrix and determining the absolute ideal points to eliminate the rank reversal in traditional TOPSIS method. The example based on the improved TOPSIS is applied to the different brand mobile phone comprehensive evaluation, which shows that the improved TOPSIS evaluation method has certain validity and feasibility.

Key words: multiple attribute decision making, TOPSIS, attribute weights, range transform, rank reversal

摘要: 传统TOPSIS评价法具有权重主观确定、更新方案产生逆序等不足,本文尝试基于用户参与选择的客观行为数据和产品属性数据以改进属性权重的确定;通过极差变换以改进初始矩阵的标准化及最优点、最差点的确定,以消除TOPSIS法评价结果可能出现的逆序问题。最后,基于改进后的TOPSIS法应用于不同品牌手机综合评价的实例验证,结果显示改进后的TOPSIS评价法更具可行性及有效性。

关键词: 多属性决策, TOPSIS法, 属性权重, 极差变换, 逆序

CLC Number: