Maximize spatial influence of facility bundle considering reverse k nearest neighbors

Wang, S; Zhang, Y; Lin, X; Cheema, MA

Maximize spatial influence of facility bundle considering reverse k nearest neighbors

Wang, S Zhang, Y

Lin, X Cheema, MA

Permalink

Publication Type:: Conference Proceeding
Citation:: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2018, 10827 LNCS pp. 684 - 700
Issue Date:: 2018-01-01

Closed Access

	Filename	Description	Size
	DASFAA_2018_Wang2018_Chapter_MaximizeSpatialInfluenceOfFaci.pdf	Published version	1.15 MB	Adobe PDF	View/Open

Copyright Clearance Process

Recently Added
In Progress
Closed Access

This item is closed access and not available.

Full metadata record

Field	Value	Language
dc.contributor.author	Wang, S	en_US
dc.contributor.author	Zhang, Y https://orcid.org/0000-0002-2674-1638	en_US
dc.contributor.author	Lin, X	en_US
dc.contributor.author	Cheema, MA	en_US
dc.date.issued	2018-01-01	en_US
dc.identifier.citation	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2018, 10827 LNCS pp. 684 - 700	en_US
dc.identifier.isbn	9783319914510	en_US
dc.identifier.issn	0302-9743	en_US
dc.identifier.uri	http://hdl.handle.net/10453/133702
dc.description.abstract	© Springer International Publishing AG, part of Springer Nature 2018. Consider a two dimensional Euclidean space, let F be a set of points representing facilities and U be a set of points representing users. Spatial influence of a facility is the number of users who have this facility as one of their k nearest facilities. This is because that users normally prefer to go to their nearby facilities and are naturally influenced the most by their nearest facilities. Given a facility bundle of size t, spatial influence of the bundle is the number of distinct users influenced by any one of them. Existing works on facility selection problem find out top-t facilities with the highest spatial influence. However, the literature lacks study on this problem when the t facilities have the highest spatial influence as a bundle. We are the first to study the problem of Maximizing Bundled Reverse k Nearest Neighbors (MB-RkNN), where the spatial influence of a facility bundle of size t is maximized. We prove its NP-hardness, and propose a branch-and-bound best first search algorithm that greedily select the currently best facility until we get t facilities. We introduce the concept of kNN region such that a group of users have their kNN facilities all belong to the same kNN region. This sharing property of kNN region allows us to avoid redundant calculation with dynamic programming technique. We conduct experiments on real data sets and show that our algorithm is orders of magnitudes better than our baseline algorithm both in terms of CPU time and IO cost.	en_US
dc.relation.ispartof	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)	en_US
dc.relation.isbasedon	10.1007/978-3-319-91452-7_44	en_US
dc.subject.classification	Artificial Intelligence & Image Processing	en_US
dc.title	Maximize spatial influence of facility bundle considering reverse k nearest neighbors	en_US
dc.type	Conference Proceeding
utslib.citation.volume	10827 LNCS	en_US
utslib.for	0805 Distributed Computing	en_US
pubs.embargo.period	Not known	en_US
pubs.organisational-group	/University of Technology Sydney
pubs.organisational-group	/University of Technology Sydney/Faculty of Engineering and Information Technology
pubs.organisational-group	/University of Technology Sydney/Faculty of Engineering and Information Technology/School of Computer Science
pubs.organisational-group	/University of Technology Sydney/Strength - CAI - Centre for Artificial Intelligence
utslib.copyright.status	closed_access
pubs.publication-status	Published	en_US
pubs.volume	10827 LNCS	en_US

Abstract:

© Springer International Publishing AG, part of Springer Nature 2018. Consider a two dimensional Euclidean space, let F be a set of points representing facilities and U be a set of points representing users. Spatial influence of a facility is the number of users who have this facility as one of their k nearest facilities. This is because that users normally prefer to go to their nearby facilities and are naturally influenced the most by their nearest facilities. Given a facility bundle of size t, spatial influence of the bundle is the number of distinct users influenced by any one of them. Existing works on facility selection problem find out top-t facilities with the highest spatial influence. However, the literature lacks study on this problem when the t facilities have the highest spatial influence as a bundle. We are the first to study the problem of Maximizing Bundled Reverse k Nearest Neighbors (MB-RkNN), where the spatial influence of a facility bundle of size t is maximized. We prove its NP-hardness, and propose a branch-and-bound best first search algorithm that greedily select the currently best facility until we get t facilities. We introduce the concept of kNN region such that a group of users have their kNN facilities all belong to the same kNN region. This sharing property of kNN region allows us to avoid redundant calculation with dynamic programming technique. We conduct experiments on real data sets and show that our algorithm is orders of magnitudes better than our baseline algorithm both in terms of CPU time and IO cost.

Please use this identifier to cite or link to this item:

http://hdl.handle.net/10453/133702