Aggregate k Nearest Neighbor Queries in Metric Spaces

Xin Ding, Yuanliang Zhang, Lu Chen, Keyu Yang, Yunjun Gao

Publikation: Bidrag til bog/antologi/rapport/konference proceedingKonferenceartikel i proceedingForskningpeer review

Abstract

Aggregate k nearest neighbor (AkNN) queries are useful in many areas, such as multimedia retrieval and resource allocation, to name but a few. Most of existing works on AkNN query only focus on Euclidean space or specific metric space, which employ properties of particular data to accelerate the query. However, due to the complex data types involved and the needs for flexible similarity criteria seen in real applications, properties of particular data cannot be used for general case. Hence, in this paper, we investigate AkNN search in metric spaces, termed as metric AkNN (MAkNN) search, as metric spaces can support any type of data and flexible similarity criteria as long as satisfying triangle inequality. To efficiently answer MAkNN queries, we develop several pruning techniques and corresponding algorithms based on SPB-tree. Extensive experiments using three real data sets verify the efficiency of our MAkNN algorithms.
OriginalsprogEngelsk
TitelWeb and Big Data : Second International Joint Conference, APWeb-WAIM 2018, Macau, China, July 23-25, 2018, Proceedings, Part II
RedaktørerYi Cai, Yoshiharu Ishikawa, Jianliang Xu
Antal sider17
Vol/bind2
ForlagSpringer
Publikationsdato2018
Sider317-333
ISBN (Trykt)978-3-319-96892-6
ISBN (Elektronisk)978-3-319-96893-3
DOI
StatusUdgivet - 2018
BegivenhedSecond International Joint Conference, APWeb-WAIM 2018 - Macau, Kina
Varighed: 23 jul. 201825 jul. 2018

Konference

KonferenceSecond International Joint Conference, APWeb-WAIM 2018
Land/OmrådeKina
ByMacau
Periode23/07/201825/07/2018
NavnLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Vol/bind10988 LNCS
ISSN0302-9743

Fingeraftryk

Dyk ned i forskningsemnerne om 'Aggregate k Nearest Neighbor Queries in Metric Spaces'. Sammen danner de et unikt fingeraftryk.

Citationsformater