A Survey on Interactive Queries

PhD Qualifying Examination


Title: "A Survey on Interactive Queries"

by

Mr. Qixu CHEN


Abstract:

The field of multi-criteria decision-making has witnessed a growing interest in 
interactive queries. Given a dataset, interactive queries first learn the 
user's preference by asking the user several rounds of simple questions, and 
then the relevant tuples meeting the user's preference are retrieved. This 
novel query type can be viewed as combining the strengths of the well-known 
top-k and skyline queries. In particular, it eliminates the need for users to 
specify a preference function in advance, which is a demanding and sometimes 
impractical requirement in top-k queries. Moreover, it provides a controlled 
number of tuples as output, addressing the potential issue of overwhelming 
results that can occur in skyline queries.

This survey begins by presenting the conventional top-k query and skyline 
query. Subsequently, we discuss algorithms for two popular types of interactive 
queries: the interactive regret-minimizing query and the interactive best point 
retrieval query. We examine the strengths and limitations associated with each 
approach. We also introduce several relevant fields which are closely related 
to interactive queries. To conclude, we point out the main challenges 
encountered in interactive query research, and provide new directions for 
future work.


Date:                   Friday, 3 May 2024

Time:                   4:00pm - 6:00pm

Venue:                  Room 5506
                        Lifts 25/26

Committee Members:      Prof. Raymond Wong (Supervisor)
                        Prof. Ke Yi (Chairperson)
                        Dr. Wilfred Ng
                        Prof. Xiaofang Zhou