包含非数值型属性的交互式遗憾最小化查询  

Interactive Regret Minimization Query with Categorical Attributes

作　　者：王美静 郑吉平[1,2] WANG Meijing;ZHENG Jiping(College of Computer Science and Technology,Nanjing University of Aeronautics and Astronautics,Nanjing 211106,China;State Key Laboratory for Novel Software Technology,Nanjing University,Nanjing 210093,China)

机构地区：[1]南京航空航天大学计算机科学与技术学院,南京211106 [2]南京大学计算机软件新技术国家重点实验室,南京210093

出　　处：《小型微型计算机系统》2024年第3期513-520,共8页Journal of Chinese Computer Systems

基　　金：国家自然科学基金项目(U1733112,61702260)资助;中央高校基本科研业务费专项资金项目(NS2020068)资助.

摘　　要：近年来,遗憾最小化查询作为多准则决策的重要工具之一,逐渐成为数据查询的研究热点.遗憾最小化查询中,通过与用户进行交互,不断地学习用户的偏好,可以有效帮助降低查询的遗憾率.然而,已有的研究仅针对具有数值型属性的数据定义遗憾率并据此进行遗憾最小化查询,当面对包含非数值型属性的数据时,这些方法不再适用.本文提出的遗憾率定义针对包含非数值型属性的数据,并给出新的交互式遗憾最小化查询问题的定义.在此基础上,采用“支配”的概念提出用于预处理的skyline删减算法,利用偏好矩阵帮助学习用户偏好,并提出用于解决交互式遗憾最小化的算法MECR_QS.最后,实验结果表明MECR_QS算法能有效处理包含非数值型属性的交互式遗憾最小化查询.In recent years,the regret minimization query,as one of the important tools of multi-criteria decision making,has gradually become a research hotspot in data queries.For regret minimization queries,the regret ratios can be effectively decreased by interacting with users and constantly learning the users′preferences.However,existing studies only define regret ratios for the data with numerical attributes and perform regret minimization queries on them.These methods are no longer applicable for the data with categorical attributes.In this paper,a new definition of interactive regret minimization query problem is proposed by defining the regret ratio for the data containing categorical attributes.Further,the concept of"domination"is adopted and a skyline pruning algorithm for preprocessing is proposed.Moreover,the preference matrix is used to help learn the users′preferences,and the algorithm MECR_QS is proposed to answer the interactive regret minimization query.Finally,the experimental results show that the algorithm MECR_QS can effectively deal with the interactive regret minimization query for the data with categorical attributes.

关 键 词：遗憾最小化查询 交互 非数值型属性 skyline筛选 

分 类 号：TP392[自动化与计算机技术—计算机应用技术]