Conference Paper

Similarity Searches on Non-ordered Discrete Data for Emerging Computer Applications: Why and How

2018 International Conference on Computer Applications in Industry and Engineering
Qiang Zhu

ABSTRACT


There is an increasing demand for processing similarity searches on non-ordered discrete data in

numerous emerging computer applications including bioinformatics, biometrics, cybersecurity, social media,

and image processing. To support ecient similarity searches, robust index techniques are required. In

this talk, we will discuss why such similarity searches are important for contemporary applications, what

the unique challenges are in processing them, and how ecient index schemes can be developed to tackle

these challenges. We will also show how classical index methods developed for ordered continuous data

fail to work for non-ordered discrete data. We will present some recent index methods specially developed

for supporting ecient similarity searches on non-ordered discrete data. We will also discuss other related

research issues including how to bulk-load large index trees, how to process similarity searches on hybrid

discrete and continuous data, and how to support other types of searches such as box queries on non-ordered

discrete data. In the end, we will highlight some future research directions in the area.

CAINE 2018



ISBN:
978-1-943436-04-0
PUBLISHER:
ACEE
CHIEF EDITOR:
Debnath
CONFERENCE VENUE:
San Diego, California, USA
CONTACT DETAILS:
Debnath
Copyright © Search Innovations. All rights reserved