Solving All-k-Nearest Neighbor Problem without an Index

Autores
Chávez, Edgar; Ludueña, Verónica; Reyes, Nora Susana
Año de publicación
2019
Idioma
inglés
Tipo de recurso
documento de conferencia
Estado
versión publicada
Descripción
Among the similarity queries in metric spaces, there are one that obtains the k-nearest neighbors of all the elements in the database (All-k-NN). One way to solve it is the naïve one: comparing each object in the database with all the other ones and returning the k elements nearest to it (k-NN). Another way to do this is by preprocessing the database to build an index, and then searching on this index for the k-NN of each element of the dataset. Answering to the All-k-NN problem allows to build the k-Nearest Neighbor graph (kNNG). Given an object collection of a metric space, the Nearest Neighbor Graph (NNG) associates each node with its closest neighbor under the given metric. If we link each object to their k nearest neighbors, we obtain the k Nearest Neighbor Graph (kNNG).The kNNG can be considered an index for a database, which is quite efficient and can allow improvements. In this work, we propose a new technique to solve the All-k-NN problem which do not use any index to obtain the k-NN of each element. This approach solves the problem avoiding as many comparisons as possible, only comparing some database elements and taking advantage of the distance function properties. Its total cost is significantly lower than that of the naïve solution.
XVI Workshop Bases de Datos y Minería de Datos.
Red de Universidades con Carreras en Informática
Materia
Ciencias Informáticas
All-k-NN problem
Database
k Nearest Neighbor Graph
Nivel de accesibilidad
acceso abierto
Condiciones de uso
http://creativecommons.org/licenses/by-nc-sa/4.0/
Repositorio
SEDICI (UNLP)
Institución
Universidad Nacional de La Plata
OAI Identificador
oai:sedici.unlp.edu.ar:10915/90536

id SEDICI_268729dda2c1f3409d544e36bde5431b
oai_identifier_str oai:sedici.unlp.edu.ar:10915/90536
network_acronym_str SEDICI
repository_id_str 1329
network_name_str SEDICI (UNLP)
spelling Solving All-k-Nearest Neighbor Problem without an IndexChávez, EdgarLudueña, VerónicaReyes, Nora SusanaCiencias InformáticasAll-k-NN problemDatabasek Nearest Neighbor GraphAmong the similarity queries in metric spaces, there are one that obtains the k-nearest neighbors of all the elements in the database (All-k-NN). One way to solve it is the naïve one: comparing each object in the database with all the other ones and returning the k elements nearest to it (k-NN). Another way to do this is by preprocessing the database to build an index, and then searching on this index for the k-NN of each element of the dataset. Answering to the All-k-NN problem allows to build the k-Nearest Neighbor graph (kNNG). Given an object collection of a metric space, the Nearest Neighbor Graph (NNG) associates each node with its closest neighbor under the given metric. If we link each object to their k nearest neighbors, we obtain the k Nearest Neighbor Graph (kNNG).The kNNG can be considered an index for a database, which is quite efficient and can allow improvements. In this work, we propose a new technique to solve the All-k-NN problem which do not use any index to obtain the k-NN of each element. This approach solves the problem avoiding as many comparisons as possible, only comparing some database elements and taking advantage of the distance function properties. Its total cost is significantly lower than that of the naïve solution.XVI Workshop Bases de Datos y Minería de Datos.Red de Universidades con Carreras en Informática2019-10info:eu-repo/semantics/conferenceObjectinfo:eu-repo/semantics/publishedVersionObjeto de conferenciahttp://purl.org/coar/resource_type/c_5794info:ar-repo/semantics/documentoDeConferenciaapplication/pdf567-576http://sedici.unlp.edu.ar/handle/10915/90536enginfo:eu-repo/semantics/altIdentifier/isbn/978-987-688-377-1info:eu-repo/semantics/reference/hdl/10915/90359info:eu-repo/semantics/openAccesshttp://creativecommons.org/licenses/by-nc-sa/4.0/Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0)reponame:SEDICI (UNLP)instname:Universidad Nacional de La Platainstacron:UNLP2025-09-29T11:18:37Zoai:sedici.unlp.edu.ar:10915/90536Institucionalhttp://sedici.unlp.edu.ar/Universidad públicaNo correspondehttp://sedici.unlp.edu.ar/oai/snrdalira@sedici.unlp.edu.arArgentinaNo correspondeNo correspondeNo correspondeopendoar:13292025-09-29 11:18:37.951SEDICI (UNLP) - Universidad Nacional de La Platafalse
dc.title.none.fl_str_mv Solving All-k-Nearest Neighbor Problem without an Index
title Solving All-k-Nearest Neighbor Problem without an Index
spellingShingle Solving All-k-Nearest Neighbor Problem without an Index
Chávez, Edgar
Ciencias Informáticas
All-k-NN problem
Database
k Nearest Neighbor Graph
title_short Solving All-k-Nearest Neighbor Problem without an Index
title_full Solving All-k-Nearest Neighbor Problem without an Index
title_fullStr Solving All-k-Nearest Neighbor Problem without an Index
title_full_unstemmed Solving All-k-Nearest Neighbor Problem without an Index
title_sort Solving All-k-Nearest Neighbor Problem without an Index
dc.creator.none.fl_str_mv Chávez, Edgar
Ludueña, Verónica
Reyes, Nora Susana
author Chávez, Edgar
author_facet Chávez, Edgar
Ludueña, Verónica
Reyes, Nora Susana
author_role author
author2 Ludueña, Verónica
Reyes, Nora Susana
author2_role author
author
dc.subject.none.fl_str_mv Ciencias Informáticas
All-k-NN problem
Database
k Nearest Neighbor Graph
topic Ciencias Informáticas
All-k-NN problem
Database
k Nearest Neighbor Graph
dc.description.none.fl_txt_mv Among the similarity queries in metric spaces, there are one that obtains the k-nearest neighbors of all the elements in the database (All-k-NN). One way to solve it is the naïve one: comparing each object in the database with all the other ones and returning the k elements nearest to it (k-NN). Another way to do this is by preprocessing the database to build an index, and then searching on this index for the k-NN of each element of the dataset. Answering to the All-k-NN problem allows to build the k-Nearest Neighbor graph (kNNG). Given an object collection of a metric space, the Nearest Neighbor Graph (NNG) associates each node with its closest neighbor under the given metric. If we link each object to their k nearest neighbors, we obtain the k Nearest Neighbor Graph (kNNG).The kNNG can be considered an index for a database, which is quite efficient and can allow improvements. In this work, we propose a new technique to solve the All-k-NN problem which do not use any index to obtain the k-NN of each element. This approach solves the problem avoiding as many comparisons as possible, only comparing some database elements and taking advantage of the distance function properties. Its total cost is significantly lower than that of the naïve solution.
XVI Workshop Bases de Datos y Minería de Datos.
Red de Universidades con Carreras en Informática
description Among the similarity queries in metric spaces, there are one that obtains the k-nearest neighbors of all the elements in the database (All-k-NN). One way to solve it is the naïve one: comparing each object in the database with all the other ones and returning the k elements nearest to it (k-NN). Another way to do this is by preprocessing the database to build an index, and then searching on this index for the k-NN of each element of the dataset. Answering to the All-k-NN problem allows to build the k-Nearest Neighbor graph (kNNG). Given an object collection of a metric space, the Nearest Neighbor Graph (NNG) associates each node with its closest neighbor under the given metric. If we link each object to their k nearest neighbors, we obtain the k Nearest Neighbor Graph (kNNG).The kNNG can be considered an index for a database, which is quite efficient and can allow improvements. In this work, we propose a new technique to solve the All-k-NN problem which do not use any index to obtain the k-NN of each element. This approach solves the problem avoiding as many comparisons as possible, only comparing some database elements and taking advantage of the distance function properties. Its total cost is significantly lower than that of the naïve solution.
publishDate 2019
dc.date.none.fl_str_mv 2019-10
dc.type.none.fl_str_mv info:eu-repo/semantics/conferenceObject
info:eu-repo/semantics/publishedVersion
Objeto de conferencia
http://purl.org/coar/resource_type/c_5794
info:ar-repo/semantics/documentoDeConferencia
format conferenceObject
status_str publishedVersion
dc.identifier.none.fl_str_mv http://sedici.unlp.edu.ar/handle/10915/90536
url http://sedici.unlp.edu.ar/handle/10915/90536
dc.language.none.fl_str_mv eng
language eng
dc.relation.none.fl_str_mv info:eu-repo/semantics/altIdentifier/isbn/978-987-688-377-1
info:eu-repo/semantics/reference/hdl/10915/90359
dc.rights.none.fl_str_mv info:eu-repo/semantics/openAccess
http://creativecommons.org/licenses/by-nc-sa/4.0/
Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0)
eu_rights_str_mv openAccess
rights_invalid_str_mv http://creativecommons.org/licenses/by-nc-sa/4.0/
Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0)
dc.format.none.fl_str_mv application/pdf
567-576
dc.source.none.fl_str_mv reponame:SEDICI (UNLP)
instname:Universidad Nacional de La Plata
instacron:UNLP
reponame_str SEDICI (UNLP)
collection SEDICI (UNLP)
instname_str Universidad Nacional de La Plata
instacron_str UNLP
institution UNLP
repository.name.fl_str_mv SEDICI (UNLP) - Universidad Nacional de La Plata
repository.mail.fl_str_mv alira@sedici.unlp.edu.ar
_version_ 1844616059836956672
score 13.070432