基于上下文偏好的Web数据库查询结果Top-K排序方法  

A Top-K Query Results Ranking Approach Based on Contextual Preferences for Web Database

作　　者：孟祥福[1] 马宗民[2] 李昕[3] 张霄雁[1] 王星[1] 

机构地区：[1]辽宁工程技术大学电子与信息工程学院,辽宁葫芦岛125105 [2]东北大学信息科学与工程学院,沈阳110819 [3]辽宁工业大学计算中心,辽宁锦州121001

出　　处：《计算机学报》2014年第9期1986-1998,共13页Chinese Journal of Computers

基　　金：国家青年科学基金(61003162);辽宁省高等学校杰出青年学者成长计划(LJQ2013038);辽宁省自然科学基金(2013020028)资助~~

摘　　要：为了解决Web数据库多查询结果的问题,该文提出了一种基于上下文偏好的查询结果top-k排序方法,首先提出了一种带偏好程度的上下文偏好模型:i_1>i_2,d|X,表示在上下文条件X下,项i_1与i_2相比,用户偏好项i_1的程度为d(0.5≤d≤1),带偏好程度的上下文偏好通过在查询历史中使用关联规则挖掘获得.基于上下文偏好,提出了一种查询结果top-k排序方法,给出了相应的元组排列创建、聚类和top-k排序算法.实验结果表明,提出的上下文偏好模型具有较强的偏好表达能力,top-k排序方法能够较好地满足用户需求和偏好并且具有较高的执行效率.To deal with the problem of too many results returned from a Web database, thispaper proposes a contextual preference-based top-k query results ranking approach. A contextualpreference model with an interest degree, which takes the form of il 〉i2, d ｜ X, meaning thatitem il is preferred to item i2 with an interest degree d in the context of X, is proposed. Thecontextual preferences can be obtained from query history by using the association rule miningalgorithm. Based on the contextual preferences, a top-k query results ranking method is proposedand the corresponding tuple＇s order creating, clustering and top-k ranking algorithms are presented.The results of experiments demonstrated that the contextual preference model has a strongexpression ability of preferences, the top-k ranking method can meet the user＇s needs and prefer-ences closely and it has high performance as well.

关 键 词：WEB数据库 上下文偏好 元组聚类 top-k排序 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]