A Survey on K-Regret Queries

PhD Qualifying Examination


Title: "A Survey on K-Regret Queries"

by

Mr. Weicheng WANG


Abstract:

K-regret queries have attracted more and more attention in the multi-criteria 
decision-making area. Given a dataset, each query tries to select k points so 
as to minimize the maximum regret ratio, i.e., find k representative points in 
the whole dataset using the regret ratio as a measurement. It focuses on how 
much disappointed a user might be if s/he only sees the k points instead of the 
whole dataset. Different from the well-known top-k and skyline queries, 
k-regret queries can be seen as bridging the advantages of these existing 
queries: controllable output and less explicit preference requirement. 
Specifically, compared with top-k queries, they do not need the user to specify 
his/her preference in advance. Superior to skyline, it only returns k points 
instead of the unpredictable output size which may overwhelm the user. Thus, a 
lot of studies on this topic have been done. In this survey, we first introduce 
top-k queries and skyline queries. Then, we focus on the existing studies on 
k-regret queries, showing the point selection strategies and claiming the 
advantages and the disadvantages of each method. We also demonstrate several 
existing studies which are closely related to k-regret queries. We conclude by 
showing the limitation of current studies and provide new directions for future 
work.


Date:			Thursday, 2 July 2020

Time:                  	2:00pm - 4:00pm

Zoom meeting:           https://hkust.zoom.us/j/91634534411

Committee Members:	Dr. Raymond Wong (Supervisor)
 			Prof. Dimitris Papadias (Chairperson)
 			Prof. Dik-Lun Lee
 			Prof. Ke Yi


**** ALL are Welcome ****