Efficient and Robust Approximate Nearest Neighbor Search Using Hierarchical Navigable Small World Graphs

We present a new approach for the approximate K-nearest neighbor search based on navigable small world graphs with controllable hierarchy (Hierarchical NSW, HNSW). The proposed solution is fully graph-based, without any need for additional search structures (typically used at the coarse search stage...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:IEEE transactions on pattern analysis and machine intelligence Ročník 42; číslo 4; s. 824 - 836
Hlavní autoři: Malkov, Yu A., Yashunin, D. A.
Médium: Journal Article
Jazyk:angličtina
Vydáno: United States IEEE 01.04.2020
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Témata:
ISSN:0162-8828, 1939-3539, 2160-9292, 1939-3539
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Abstract We present a new approach for the approximate K-nearest neighbor search based on navigable small world graphs with controllable hierarchy (Hierarchical NSW, HNSW). The proposed solution is fully graph-based, without any need for additional search structures (typically used at the coarse search stage of the most proximity graph techniques). Hierarchical NSW incrementally builds a multi-layer structure consisting of a hierarchical set of proximity graphs (layers) for nested subsets of the stored elements. The maximum layer in which an element is present is selected randomly with an exponentially decaying probability distribution. This allows producing graphs similar to the previously studied Navigable Small World (NSW) structures while additionally having the links separated by their characteristic distance scales. Starting the search from the upper layer together with utilizing the scale separation boosts the performance compared to NSW and allows a logarithmic complexity scaling. Additional employment of a heuristic for selecting proximity graph neighbors significantly increases performance at high recall and in case of highly clustered data. Performance evaluation has demonstrated that the proposed general metric space search index is able to strongly outperform previous opensource state-of-the-art vector-only approaches. Similarity of the algorithm to the skip list structure allows straightforward balanced distributed implementation.
AbstractList We present a new approach for the approximate K-nearest neighbor search based on navigable small world graphs with controllable hierarchy (Hierarchical NSW, HNSW). The proposed solution is fully graph-based, without any need for additional search structures (typically used at the coarse search stage of the most proximity graph techniques). Hierarchical NSW incrementally builds a multi-layer structure consisting of a hierarchical set of proximity graphs (layers) for nested subsets of the stored elements. The maximum layer in which an element is present is selected randomly with an exponentially decaying probability distribution. This allows producing graphs similar to the previously studied Navigable Small World (NSW) structures while additionally having the links separated by their characteristic distance scales. Starting the search from the upper layer together with utilizing the scale separation boosts the performance compared to NSW and allows a logarithmic complexity scaling. Additional employment of a heuristic for selecting proximity graph neighbors significantly increases performance at high recall and in case of highly clustered data. Performance evaluation has demonstrated that the proposed general metric space search index is able to strongly outperform previous opensource state-of-the-art vector-only approaches. Similarity of the algorithm to the skip list structure allows straightforward balanced distributed implementation.
We present a new approach for the approximate K-nearest neighbor search based on navigable small world graphs with controllable hierarchy (Hierarchical NSW, HNSW). The proposed solution is fully graph-based, without any need for additional search structures (typically used at the coarse search stage of the most proximity graph techniques). Hierarchical NSW incrementally builds a multi-layer structure consisting of a hierarchical set of proximity graphs (layers) for nested subsets of the stored elements. The maximum layer in which an element is present is selected randomly with an exponentially decaying probability distribution. This allows producing graphs similar to the previously studied Navigable Small World (NSW) structures while additionally having the links separated by their characteristic distance scales. Starting the search from the upper layer together with utilizing the scale separation boosts the performance compared to NSW and allows a logarithmic complexity scaling. Additional employment of a heuristic for selecting proximity graph neighbors significantly increases performance at high recall and in case of highly clustered data. Performance evaluation has demonstrated that the proposed general metric space search index is able to strongly outperform previous opensource state-of-the-art vector-only approaches. Similarity of the algorithm to the skip list structure allows straightforward balanced distributed implementation.We present a new approach for the approximate K-nearest neighbor search based on navigable small world graphs with controllable hierarchy (Hierarchical NSW, HNSW). The proposed solution is fully graph-based, without any need for additional search structures (typically used at the coarse search stage of the most proximity graph techniques). Hierarchical NSW incrementally builds a multi-layer structure consisting of a hierarchical set of proximity graphs (layers) for nested subsets of the stored elements. The maximum layer in which an element is present is selected randomly with an exponentially decaying probability distribution. This allows producing graphs similar to the previously studied Navigable Small World (NSW) structures while additionally having the links separated by their characteristic distance scales. Starting the search from the upper layer together with utilizing the scale separation boosts the performance compared to NSW and allows a logarithmic complexity scaling. Additional employment of a heuristic for selecting proximity graph neighbors significantly increases performance at high recall and in case of highly clustered data. Performance evaluation has demonstrated that the proposed general metric space search index is able to strongly outperform previous opensource state-of-the-art vector-only approaches. Similarity of the algorithm to the skip list structure allows straightforward balanced distributed implementation.
Author Yashunin, D. A.
Malkov, Yu A.
Author_xml – sequence: 1
  givenname: Yu A.
  orcidid: 0000-0003-4324-6433
  surname: Malkov
  fullname: Malkov, Yu A.
  email: yurymalkov@mail.ru
  organization: Samsung AI Center, Moscow, Russia
– sequence: 2
  givenname: D. A.
  surname: Yashunin
  fullname: Yashunin, D. A.
  email: yashuninda@yandex.ru
  organization: Nizhny Novgorod, Russia
BackLink https://www.ncbi.nlm.nih.gov/pubmed/30602420$$D View this record in MEDLINE/PubMed
BookMark eNp9kU1vEzEQhi1URNPCHwAJWeLCZYO_1msfo6q0ldqAaCuOltc7Tlxt1qm9i-Df45C0hx44WR49z2hm3hN0NMQBEHpPyZxSor_cfV_cXM0ZoWrOlNKi4a_QjFFJKs00O0IzQiWrlGLqGJ3k_EAIFTXhb9AxJ5IwwcgMrc-9Dy7AMGI7dPhHbKc84sV2m-LvsLEj4CXYBKW2hLBatzHh21Jwa3yfw7DClwHS7huc7fHS_gor2_aAbze27_HPmPoOXyS7Xee36LW3fYZ3h_cU3X89vzu7rK6_XVydLa4rJ2gzVgJAkcY2VDGvaM2od16D9YJp6S1lToNUviF17WTnVcdbaonVlvjOiSLzU_R537ds8DiVwc0mZAd9bweIUzblPpxQympd0E8v0Ic4paFMZxhvhNSS6R318UBN7QY6s03lLumPebphAdQecCnmnMAbF0Y7hjiMyYbeUGJ2cZl_cZldXOYQV1HZC_Wp-3-lD3spAMCzoGotJJf8L3cRoGw
CODEN ITPIDJ
CitedBy_id crossref_primary_10_1109_JAS_2022_105635
crossref_primary_10_1007_s00530_023_01065_2
crossref_primary_10_14778_3725688_3725705
crossref_primary_10_1145_3725413
crossref_primary_10_1142_S2196888823500148
crossref_primary_10_14778_3725688_3725709
crossref_primary_10_14778_3748191_3748193
crossref_primary_10_1016_j_ins_2023_120076
crossref_primary_10_1080_01431161_2024_2302353
crossref_primary_10_1109_TKDE_2022_3231781
crossref_primary_10_1016_j_patrec_2022_01_025
crossref_primary_10_1109_TVLSI_2025_3544342
crossref_primary_10_1145_3488377
crossref_primary_10_1093_nargab_lqaf040
crossref_primary_10_1145_3705000
crossref_primary_10_1049_itr2_12583
crossref_primary_10_1145_3677129
crossref_primary_10_1109_TKDE_2019_2953897
crossref_primary_10_1109_LRA_2022_3194310
crossref_primary_10_1145_3631939
crossref_primary_10_1155_2021_2011738
crossref_primary_10_21105_joss_08033
crossref_primary_10_2478_cait_2024_0035
crossref_primary_10_1007_s10115_023_01899_2
crossref_primary_10_2478_acss_2025_0004
crossref_primary_10_17586_2226_1494_2025_25_4_710_717
crossref_primary_10_26599_BDMA_2025_9020022
crossref_primary_10_1007_s10115_024_02272_7
crossref_primary_10_1016_j_ins_2020_08_054
crossref_primary_10_1109_TDSC_2024_3383159
crossref_primary_10_1145_3725403
crossref_primary_10_1109_TPAMI_2022_3157197
crossref_primary_10_1109_TMM_2022_3210376
crossref_primary_10_1186_s13059_022_02820_w
crossref_primary_10_1145_3725401
crossref_primary_10_1145_3654920
crossref_primary_10_3390_info14030149
crossref_primary_10_1016_j_autcon_2025_106103
crossref_primary_10_1007_s10772_024_10147_z
crossref_primary_10_1007_s10278_025_01598_0
crossref_primary_10_1109_TEVC_2021_3106672
crossref_primary_10_3390_signals2020021
crossref_primary_10_1016_j_future_2024_03_002
crossref_primary_10_1016_j_ijmedinf_2024_105680
crossref_primary_10_1080_1206212X_2025_2537892
crossref_primary_10_1109_ACCESS_2023_3289836
crossref_primary_10_1038_s41587_020_0591_3
crossref_primary_10_1016_j_softx_2024_102016
crossref_primary_10_1016_j_eswa_2025_128596
crossref_primary_10_1038_s41598_025_12685_1
crossref_primary_10_1016_j_ress_2024_110582
crossref_primary_10_1016_j_isprsjprs_2024_04_005
crossref_primary_10_1038_s41598_025_15203_5
crossref_primary_10_1016_j_patcog_2025_111805
crossref_primary_10_1038_s41467_022_32097_3
crossref_primary_10_1016_j_isci_2023_106536
crossref_primary_10_7717_peerj_cs_1929
crossref_primary_10_1109_TCC_2022_3169329
crossref_primary_10_1007_s11227_021_03699_3
crossref_primary_10_1109_TBDATA_2023_3327220
crossref_primary_10_1145_3709743
crossref_primary_10_1145_3733610
crossref_primary_10_1007_s10346_021_01761_y
crossref_primary_10_3390_machines12050321
crossref_primary_10_1038_s41467_020_18416_6
crossref_primary_10_1038_s41467_021_25773_3
crossref_primary_10_3847_1538_4357_adb0b7
crossref_primary_10_1007_s41060_024_00709_4
crossref_primary_10_1007_s12539_020_00411_6
crossref_primary_10_1016_j_aei_2022_101680
crossref_primary_10_1016_j_eswa_2024_123191
crossref_primary_10_1109_TDSC_2025_3547243
crossref_primary_10_1109_TKDE_2023_3266495
crossref_primary_10_1016_j_enconman_2024_118391
crossref_primary_10_1109_TSG_2024_3474039
crossref_primary_10_1002_cpe_5970
crossref_primary_10_1049_ipr2_12883
crossref_primary_10_1109_TPAMI_2019_2944597
crossref_primary_10_1007_s41781_024_00117_0
crossref_primary_10_1145_3749179
crossref_primary_10_1007_s00366_021_01304_y
crossref_primary_10_17150_2713_1734_2024_6_2__251_262
crossref_primary_10_1109_TNET_2022_3217012
crossref_primary_10_1134_S1064562423701570
crossref_primary_10_1016_j_is_2022_102123
crossref_primary_10_1145_3639324
crossref_primary_10_1016_j_knosys_2023_110907
crossref_primary_10_1016_j_cogsys_2024_101216
crossref_primary_10_1007_s13369_024_08974_y
crossref_primary_10_1109_TPAMI_2022_3198729
crossref_primary_10_1016_j_neucom_2024_128172
crossref_primary_10_14778_3749646_3749704
crossref_primary_10_1109_TIFS_2021_3139219
crossref_primary_10_14778_3750601_3750688
crossref_primary_10_3390_rs15174287
crossref_primary_10_1109_TKDE_2024_3408815
crossref_primary_10_1162_coli_a_00543
crossref_primary_10_1109_TFUZZ_2019_2949769
crossref_primary_10_1093_bib_bbab573
crossref_primary_10_1109_TBDATA_2022_3161156
crossref_primary_10_1145_3654970
crossref_primary_10_1080_20964471_2025_2529639
crossref_primary_10_1016_j_sysarc_2025_103514
crossref_primary_10_3389_fnbot_2023_1301785
crossref_primary_10_7717_peerj_cs_1757
crossref_primary_10_1007_s11063_024_11444_z
crossref_primary_10_1142_S1793351X25420024
crossref_primary_10_1162_tacl_a_00415
crossref_primary_10_1007_s10115_022_01742_0
crossref_primary_10_1145_3469379_3469384
crossref_primary_10_1109_TMM_2021_3073811
crossref_primary_10_1145_3749116_3749122
crossref_primary_10_1109_TPAMI_2022_3187350
crossref_primary_10_1109_TCASAI_2025_3540957
crossref_primary_10_1038_s41592_021_01286_1
crossref_primary_10_1109_TKDE_2022_3150403
crossref_primary_10_3390_ai6050093
crossref_primary_10_1109_TMC_2024_3361876
crossref_primary_10_14778_3750601_3750674
crossref_primary_10_1038_s41598_025_09856_5
crossref_primary_10_14778_3748191_3748200
crossref_primary_10_1007_s00799_023_00354_5
crossref_primary_10_1016_j_eswa_2022_117885
crossref_primary_10_1186_s40537_025_01257_9
crossref_primary_10_1109_TSG_2024_3446859
crossref_primary_10_1109_TCSS_2024_3453226
crossref_primary_10_1109_TIM_2025_3557112
crossref_primary_10_3390_app131810468
crossref_primary_10_1177_00405175241293764
crossref_primary_10_14778_3748191_3748212
crossref_primary_10_1049_ell2_12052
crossref_primary_10_1109_ACCESS_2025_3561594
crossref_primary_10_1109_TPAMI_2019_2952606
crossref_primary_10_1073_pnas_2425048122
crossref_primary_10_1145_3486250
crossref_primary_10_1038_s44385_025_00022_0
crossref_primary_10_1145_3654990
crossref_primary_10_1109_ACCESS_2025_3542036
crossref_primary_10_1109_JSTARS_2023_3298994
crossref_primary_10_1109_TPDS_2023_3298038
crossref_primary_10_1088_1742_6596_2711_1_012014
crossref_primary_10_1145_3711903
crossref_primary_10_1109_TCSVT_2022_3213680
crossref_primary_10_1109_TKDE_2023_3270264
crossref_primary_10_1016_j_nlp_2025_100137
crossref_primary_10_1145_3709701
crossref_primary_10_3390_sym16081084
crossref_primary_10_1007_s00778_024_00864_x
crossref_primary_10_1051_shsconf_202419703002
crossref_primary_10_1093_nargab_lqae172
crossref_primary_10_14778_3725688_3725692
crossref_primary_10_1007_s10489_024_06135_0
crossref_primary_10_14778_3734839_3734860
crossref_primary_10_1007_s00778_024_00886_5
crossref_primary_10_1007_s10791_023_09425_2
crossref_primary_10_1145_3589282
crossref_primary_10_14778_3725688_3725696
crossref_primary_10_14778_3746405_3746419
crossref_primary_10_3390_rs16020214
crossref_primary_10_3390_electronics10121434
crossref_primary_10_1007_s13755_023_00253_8
crossref_primary_10_1109_TSE_2025_3531210
crossref_primary_10_1109_ACCESS_2023_3295776
crossref_primary_10_1145_3656482
crossref_primary_10_3390_bdcc6040136
crossref_primary_10_2113_RGG20244760
crossref_primary_10_3390_app13063941
crossref_primary_10_3390_technologies9010010
crossref_primary_10_1145_3634912
crossref_primary_10_1109_MCAS_2024_3476008
crossref_primary_10_1007_s44336_024_00008_3
crossref_primary_10_1016_j_jisa_2022_103335
crossref_primary_10_1016_j_engappai_2024_108912
crossref_primary_10_1016_j_engappai_2023_105902
crossref_primary_10_1109_TBDATA_2021_3101517
crossref_primary_10_1109_ACCESS_2025_3582825
crossref_primary_10_1145_3736754
crossref_primary_10_14778_3746405_3746427
crossref_primary_10_1109_TBDATA_2024_3460534
crossref_primary_10_1007_s00778_024_00874_9
crossref_primary_10_1016_j_ins_2024_120158
crossref_primary_10_1007_s00778_022_00762_0
crossref_primary_10_1016_j_ins_2021_10_027
crossref_primary_10_1016_j_eswa_2024_126254
crossref_primary_10_3390_s23020629
crossref_primary_10_3390_electronics14183605
crossref_primary_10_1109_TKDE_2022_3220683
crossref_primary_10_1109_TNET_2024_3473853
crossref_primary_10_37745_ejcsit_2013_vol13n362129
crossref_primary_10_1007_s00170_022_10785_0
crossref_primary_10_1109_TKDE_2023_3341098
crossref_primary_10_1016_j_patcog_2023_109300
crossref_primary_10_1109_TCCN_2025_3528895
crossref_primary_10_3390_app14135520
crossref_primary_10_1145_3637870
crossref_primary_10_1016_j_patcog_2023_109422
crossref_primary_10_1093_bib_bbad157
crossref_primary_10_1021_acs_jcim_5c00394
crossref_primary_10_3390_biomedicines10071491
crossref_primary_10_1109_TIP_2023_3261755
crossref_primary_10_1093_nar_gkae609
crossref_primary_10_1145_3709693
crossref_primary_10_1016_j_neucom_2020_08_084
crossref_primary_10_3390_make7020048
crossref_primary_10_1007_s10489_025_06530_1
crossref_primary_10_1145_3725260
crossref_primary_10_1007_s10586_023_04159_8
crossref_primary_10_1007_s00521_020_05530_1
crossref_primary_10_1007_s10044_025_01462_5
crossref_primary_10_1016_j_datak_2025_102494
crossref_primary_10_1109_TIT_2021_3102177
crossref_primary_10_1007_s42524_022_0225_1
crossref_primary_10_1109_TSMC_2023_3234227
crossref_primary_10_1109_JIOT_2024_3358349
crossref_primary_10_1088_1674_4527_adcdac
crossref_primary_10_14778_3749646_3749650
crossref_primary_10_3390_rs14215619
crossref_primary_10_1016_j_is_2025_102563
crossref_primary_10_1145_3639269
crossref_primary_10_1007_s13042_024_02104_8
crossref_primary_10_1016_j_is_2025_102565
crossref_primary_10_1109_TIE_2020_2978722
crossref_primary_10_1007_s10618_023_00950_8
crossref_primary_10_1109_LRA_2021_3094228
crossref_primary_10_1177_00405175211037186
crossref_primary_10_1038_s41467_023_37969_w
crossref_primary_10_3390_rs16091483
crossref_primary_10_1007_s11042_023_18023_9
crossref_primary_10_1016_j_ins_2025_122414
crossref_primary_10_1155_2022_1154325
crossref_primary_10_1088_2631_8695_ae0003
crossref_primary_10_1093_bib_bbae109
crossref_primary_10_1109_ACCESS_2025_3602730
crossref_primary_10_1145_3609384
crossref_primary_10_1109_TCSVT_2025_3558996
crossref_primary_10_1109_TIP_2022_3163571
crossref_primary_10_1016_j_jss_2021_111008
crossref_primary_10_1093_nar_gkae029
crossref_primary_10_1016_j_patcog_2024_110596
crossref_primary_10_1145_3725399
crossref_primary_10_1109_TCAD_2024_3378600
crossref_primary_10_21105_joss_07489
crossref_primary_10_1016_j_eswa_2022_119054
crossref_primary_10_1145_3725274
crossref_primary_10_7554_eLife_66747
crossref_primary_10_1109_TCAD_2022_3198513
crossref_primary_10_1145_3626710
crossref_primary_10_1007_s13735_024_00342_8
crossref_primary_10_1109_TKDE_2022_3232689
crossref_primary_10_1109_TKDE_2022_3198689
crossref_primary_10_1109_TPAMI_2021_3067706
crossref_primary_10_1007_s00500_022_07081_0
crossref_primary_10_14778_3750601_3750624
crossref_primary_10_3390_math13132199
crossref_primary_10_1016_j_measurement_2024_115245
crossref_primary_10_5715_jnlp_32_886
crossref_primary_10_1109_TITS_2022_3229364
crossref_primary_10_1109_TKDE_2022_3155196
crossref_primary_10_1145_3722552
crossref_primary_10_1016_j_knosys_2022_109597
crossref_primary_10_1109_TIP_2020_3036779
crossref_primary_10_1145_3736716
crossref_primary_10_1109_TMM_2024_3410136
crossref_primary_10_1016_j_knosys_2023_110750
crossref_primary_10_1109_LSP_2022_3148674
crossref_primary_10_14778_3750601_3750619
crossref_primary_10_1016_j_ins_2022_06_086
crossref_primary_10_14778_3717755_3717770
crossref_primary_10_1109_ACCESS_2025_3557609
crossref_primary_10_1145_3603167
crossref_primary_10_1109_JSTARS_2023_3323819
crossref_primary_10_1002_spe_3355
crossref_primary_10_1016_j_knosys_2022_110219
crossref_primary_10_1007_s42514_022_00103_1
crossref_primary_10_1145_3725338
crossref_primary_10_1016_j_future_2024_107586
crossref_primary_10_1093_bib_bbad475
crossref_primary_10_1109_TKDE_2024_3487759
crossref_primary_10_1007_s11760_024_03562_x
crossref_primary_10_1145_3527546_3527552
crossref_primary_10_1016_j_inffus_2023_101979
crossref_primary_10_1016_j_ins_2025_122462
crossref_primary_10_1145_3736589
crossref_primary_10_1016_j_neucom_2022_07_001
crossref_primary_10_1145_3568681
crossref_primary_10_1038_s41598_025_02643_2
crossref_primary_10_1145_3512767
crossref_primary_10_14778_3717755_3717760
crossref_primary_10_1109_TCSVT_2023_3253548
crossref_primary_10_1109_TNET_2022_3187044
crossref_primary_10_1016_j_cose_2024_104249
crossref_primary_10_1016_j_isprsjprs_2025_04_015
crossref_primary_10_1145_3588908
crossref_primary_10_3390_rs15010160
crossref_primary_10_1007_s11263_021_01511_6
crossref_primary_10_1038_s41467_023_39279_7
crossref_primary_10_1145_3725325
crossref_primary_10_1007_s10489_024_05969_y
crossref_primary_10_1038_s43588_021_00029_8
crossref_primary_10_14778_3749646_3749697
crossref_primary_10_1016_j_imavis_2025_105517
crossref_primary_10_1016_j_fsidi_2025_301930
crossref_primary_10_1145_3709661
crossref_primary_10_1145_3735128
crossref_primary_10_1007_s00778_024_00879_4
crossref_primary_10_1016_j_eswa_2023_120832
crossref_primary_10_1002_widm_70021
crossref_primary_10_1038_s41467_021_24991_z
crossref_primary_10_1038_s42256_025_01112_9
crossref_primary_10_1007_s00778_024_00894_5
crossref_primary_10_3390_s25082571
crossref_primary_10_1109_TIFS_2024_3488527
crossref_primary_10_1145_3576922
crossref_primary_10_17586_2226_1494_2025_25_2_339_344
crossref_primary_10_1016_j_ins_2022_05_072
crossref_primary_10_1093_bib_bbae586
crossref_primary_10_1016_j_knosys_2022_108757
crossref_primary_10_1016_j_isprsjprs_2025_04_002
crossref_primary_10_1109_OJCS_2025_3602355
crossref_primary_10_1109_TIE_2022_3222696
crossref_primary_10_17586_2226_1494_2025_25_4_694_702
crossref_primary_10_1007_s11704_025_50480_3
crossref_primary_10_1007_s00778_022_00729_1
crossref_primary_10_1109_TIP_2022_3189803
crossref_primary_10_1016_j_simpa_2025_100772
crossref_primary_10_1038_s41586_024_08411_y
crossref_primary_10_1017_nlp_2024_53
crossref_primary_10_1093_bioinformatics_btaf198
crossref_primary_10_26599_BDMA_2025_9020009
crossref_primary_10_1016_j_autcon_2025_106490
crossref_primary_10_1038_s41592_020_0905_x
crossref_primary_10_1109_TIT_2022_3226479
crossref_primary_10_1109_TKDE_2025_3545176
crossref_primary_10_3233_JIFS_236010
crossref_primary_10_1109_TNSE_2024_3524576
crossref_primary_10_14778_3750601_3750700
crossref_primary_10_1109_TC_2022_3155956
crossref_primary_10_3390_app132413177
crossref_primary_10_1016_j_eswa_2025_126711
crossref_primary_10_1007_s00778_022_00771_z
crossref_primary_10_1016_j_neucom_2025_131284
crossref_primary_10_1038_s41467_023_44290_z
crossref_primary_10_1109_TGRS_2025_3586317
crossref_primary_10_1038_s41586_024_08247_6
crossref_primary_10_1016_j_asoc_2024_112353
crossref_primary_10_1007_s10472_023_09882_x
crossref_primary_10_1371_journal_ppat_1012095
crossref_primary_10_1371_journal_pone_0281275
crossref_primary_10_1109_TKDE_2024_3381111
crossref_primary_10_1145_3709679
crossref_primary_10_1109_TIFS_2024_3488501
crossref_primary_10_1109_TVLSI_2024_3496589
crossref_primary_10_1145_3652609
crossref_primary_10_1007_s00778_024_00893_6
crossref_primary_10_1109_TCYB_2025_3562032
crossref_primary_10_1016_j_neunet_2025_107316
crossref_primary_10_1002_jrs_70047
crossref_primary_10_1109_TII_2021_3120141
Cites_doi 10.1007/978-3-642-41062-8_28
10.2307/2786545
10.1109/CVPR.2016.223
10.1007/978-3-319-25087-8_10
10.1109/TPAMI.2014.2321376
10.1016/j.is.2012.06.005
10.1007/978-3-642-32153-5_10
10.3115/v1/D14-1162
10.1145/1963405.1963487
10.1109/TPAMI.2007.70815
10.1145/2393347.2393378
10.1007/BF02574694
10.1103/PhysRevE.82.036106
10.1145/335305.335325
10.1145/276698.276876
10.1109/TIT.2015.2418284
10.1007/978-3-319-14998-1_18
10.1103/PhysRevLett.102.238703
10.1145/1109557.1109601
10.1109/TPAMI.2013.231
10.1007/978-3-319-46475-6_48
10.1371/journal.pone.0158162
10.1038/30918
10.1002/(SICI)1097-4571(199009)41:6<391::AID-ASI1>3.0.CO;2-9
10.1016/0031-3203(80)90066-7
10.1016/j.is.2016.03.003
10.1007/978-3-642-15992-3_29
10.1016/j.is.2013.10.006
10.1145/78973.78977
10.1109/TPAMI.2014.2343223
10.1007/978-3-540-77096-1_23
10.1007/s007780200060
10.1109/IPDPS.2007.370210
10.1038/nphys1130
10.1007/978-3-319-27671-7_27
10.1109/CVPR.2016.616
10.1109/CVPR.2014.298
10.1023/B:VISI.0000029664.99615.94
10.14778/2824032.2824059
10.1109/5.726791
10.1109/CVPR.2012.6248038
10.1038/35022643
10.1016/j.is.2015.02.001
10.1137/1.9781611973068.36
10.1145/2020408.2020576
10.1038/ncomms8651
10.1109/TPAMI.2010.57
ContentType Journal Article
Copyright Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2020
Copyright_xml – notice: Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2020
DBID 97E
RIA
RIE
AAYXX
CITATION
NPM
7SC
7SP
8FD
JQ2
L7M
L~C
L~D
7X8
DOI 10.1109/TPAMI.2018.2889473
DatabaseName IEEE All-Society Periodicals Package (ASPP) 2005-present
IEEE All-Society Periodicals Package (ASPP) 1998-Present
IEEE Electronic Library (IEL)
CrossRef
PubMed
Computer and Information Systems Abstracts
Electronics & Communications Abstracts
Technology Research Database
ProQuest Computer Science Collection
Advanced Technologies Database with Aerospace
Computer and Information Systems Abstracts – Academic
Computer and Information Systems Abstracts Professional
MEDLINE - Academic
DatabaseTitle CrossRef
PubMed
Technology Research Database
Computer and Information Systems Abstracts – Academic
Electronics & Communications Abstracts
ProQuest Computer Science Collection
Computer and Information Systems Abstracts
Advanced Technologies Database with Aerospace
Computer and Information Systems Abstracts Professional
MEDLINE - Academic
DatabaseTitleList PubMed

Technology Research Database
MEDLINE - Academic
Database_xml – sequence: 1
  dbid: NPM
  name: PubMed
  url: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed
  sourceTypes: Index Database
– sequence: 2
  dbid: RIE
  name: IEEE Electronic Library (IEL)
  url: https://ieeexplore.ieee.org/
  sourceTypes: Publisher
– sequence: 3
  dbid: 7X8
  name: MEDLINE - Academic
  url: https://search.proquest.com/medline
  sourceTypes: Aggregation Database
DeliveryMethod fulltext_linktorsrc
Discipline Engineering
Computer Science
EISSN 2160-9292
1939-3539
EndPage 836
ExternalDocumentID 30602420
10_1109_TPAMI_2018_2889473
8594636
Genre orig-research
Journal Article
GrantInformation_xml – fundername: Russian Foundation for Basic Research; RFBR
  grantid: 16-31-60104 mol_a_dk
  funderid: 10.13039/501100002261
GroupedDBID ---
-DZ
-~X
.DC
0R~
29I
4.4
53G
5GY
6IK
97E
AAJGR
AARMG
AASAJ
AAWTH
ABAZT
ABQJQ
ABVLG
ACGFO
ACGFS
ACIWK
ACNCT
AENEX
AGQYO
AHBIQ
AKJIK
AKQYR
ALMA_UNASSIGNED_HOLDINGS
ASUFR
ATWAV
BEFXN
BFFAM
BGNUA
BKEBE
BPEOZ
CS3
DU5
E.L
EBS
EJD
F5P
HZ~
IEDLZ
IFIPE
IPLJI
JAVBF
LAI
M43
MS~
O9-
OCL
P2P
PQQKQ
RIA
RIE
RNS
RXW
TAE
TN5
UHB
~02
AAYXX
CITATION
5VS
9M8
AAYOK
ABFSI
ADRHT
AETIX
AGSQL
AI.
AIBXA
ALLEH
FA8
H~9
IBMZZ
ICLAB
IFJZH
NPM
PKN
RIC
RIG
RNI
RZB
VH1
XJT
Z5M
7SC
7SP
8FD
JQ2
L7M
L~C
L~D
7X8
ID FETCH-LOGICAL-c417t-4ee807a7182f81521fcf9eaf4296fa12c9e68f7055c6df8d3b1a0a9a0fdc44ee3
IEDL.DBID RIE
ISICitedReferencesCount 863
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000526541100005&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 0162-8828
1939-3539
IngestDate Sun Sep 28 06:22:32 EDT 2025
Sun Jun 29 16:13:20 EDT 2025
Wed Feb 19 02:31:40 EST 2025
Sat Nov 29 05:15:58 EST 2025
Tue Nov 18 21:24:08 EST 2025
Wed Aug 27 02:42:21 EDT 2025
IsPeerReviewed true
IsScholarly true
Issue 4
Language English
License https://ieeexplore.ieee.org/Xplorehelp/downloads/license-information/IEEE.html
https://doi.org/10.15223/policy-029
https://doi.org/10.15223/policy-037
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-c417t-4ee807a7182f81521fcf9eaf4296fa12c9e68f7055c6df8d3b1a0a9a0fdc44ee3
Notes ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
content type line 23
ORCID 0000-0003-4324-6433
PMID 30602420
PQID 2374696299
PQPubID 85458
PageCount 13
ParticipantIDs crossref_citationtrail_10_1109_TPAMI_2018_2889473
crossref_primary_10_1109_TPAMI_2018_2889473
proquest_miscellaneous_2163011259
proquest_journals_2374696299
pubmed_primary_30602420
ieee_primary_8594636
PublicationCentury 2000
PublicationDate 2020-04-01
PublicationDateYYYYMMDD 2020-04-01
PublicationDate_xml – month: 04
  year: 2020
  text: 2020-04-01
  day: 01
PublicationDecade 2020
PublicationPlace United States
PublicationPlace_xml – name: United States
– name: New York
PublicationTitle IEEE transactions on pattern analysis and machine intelligence
PublicationTitleAbbrev TPAMI
PublicationTitleAlternate IEEE Trans Pattern Anal Mach Intell
PublicationYear 2020
Publisher IEEE
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Publisher_xml – name: IEEE
– name: The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
References ref57
ref13
ref12
ref15
ref52
ref55
ref11
ref54
ref10
andoni (ref51) 0
beecks (ref59) 2013
ref17
ref16
ref19
boytsov (ref50) 0
ref46
ref45
ref48
ponomarenko (ref33) 2014
ref47
andoni (ref8) 0
ref42
ref41
rehurek (ref58) 0
ref44
ref43
ref49
ref7
ref9
ref4
ref6
ref5
ref40
ponomarenko (ref30) 0
yianilos (ref3) 0; 93
ref35
ref34
ref37
ref36
ref31
ref32
ref2
ref1
ref39
ref38
babenko (ref14) 0
arya (ref18) 0
bolettieri (ref53) 2009
ref23
ref26
ref25
ref20
ref22
ref21
ref28
ref27
ref29
paredes (ref24) 2008
gonzalez (ref56) 2008; 30
References_xml – ident: ref49
  doi: 10.1007/978-3-642-41062-8_28
– ident: ref42
  doi: 10.2307/2786545
– ident: ref17
  doi: 10.1109/CVPR.2016.223
– ident: ref23
  doi: 10.1007/978-3-319-25087-8_10
– ident: ref6
  doi: 10.1109/TPAMI.2014.2321376
– ident: ref57
  doi: 10.1016/j.is.2012.06.005
– ident: ref25
  doi: 10.1007/978-3-642-32153-5_10
– ident: ref52
  doi: 10.3115/v1/D14-1162
– ident: ref29
  doi: 10.1145/1963405.1963487
– volume: 30
  start-page: 1647
  year: 2008
  ident: ref56
  article-title: Effective proximity retrieval by ordering permutations
  publication-title: IEEE Trans Pattern Anal Mach Intell
  doi: 10.1109/TPAMI.2007.70815
– ident: ref19
  doi: 10.1145/2393347.2393378
– year: 2009
  ident: ref53
  article-title: CoPhIR: A test collection for content-based image retrieval
– ident: ref48
  doi: 10.1007/BF02574694
– ident: ref35
  doi: 10.1103/PhysRevE.82.036106
– volume: 93
  start-page: 311
  year: 0
  ident: ref3
  article-title: Data structures and algorithms for nearest neighbor search in general metric spaces
  publication-title: SODA
– ident: ref41
  doi: 10.1145/335305.335325
– year: 2008
  ident: ref24
  article-title: Graphs for metric space searching
– ident: ref9
  doi: 10.1145/276698.276876
– ident: ref38
  doi: 10.1109/TIT.2015.2418284
– start-page: 793
  year: 0
  ident: ref51
  article-title: Optimal data-dependent hashing for approximate near neighbors
  publication-title: Proc Annual ACM Symp on Theory of Computing
– ident: ref10
  doi: 10.1007/978-3-319-14998-1_18
– year: 0
  ident: ref30
  article-title: Approximate nearest neighbor search small world approach
  publication-title: Proc Int Conf Inf Commun Technol
– start-page: 1574
  year: 0
  ident: ref50
  article-title: Learning to prune in metric and non-metric spaces
  publication-title: Proc Adv Neural Inf Process Syst
– ident: ref28
  doi: 10.1103/PhysRevLett.102.238703
– ident: ref45
  doi: 10.1145/1109557.1109601
– ident: ref11
  doi: 10.1109/TPAMI.2013.231
– ident: ref15
  doi: 10.1007/978-3-319-46475-6_48
– ident: ref44
  doi: 10.1371/journal.pone.0158162
– ident: ref43
  doi: 10.1038/30918
– ident: ref2
  doi: 10.1002/(SICI)1097-4571(199009)41:6<391::AID-ASI1>3.0.CO;2-9
– start-page: 271
  year: 0
  ident: ref18
  article-title: Approximate nearest neighbor queries in fixed dimensions
  publication-title: Proc 9th Annual ACM-SIAM Symp on Discrete Algorithms
– ident: ref46
  doi: 10.1016/0031-3203(80)90066-7
– ident: ref5
  doi: 10.1016/j.is.2016.03.003
– ident: ref21
  doi: 10.1007/978-3-642-15992-3_29
– ident: ref26
  doi: 10.1016/j.is.2013.10.006
– ident: ref27
  doi: 10.1145/78973.78977
– ident: ref7
  doi: 10.1109/TPAMI.2014.2343223
– ident: ref39
  doi: 10.1007/978-3-540-77096-1_23
– ident: ref4
  doi: 10.1007/s007780200060
– ident: ref40
  doi: 10.1109/IPDPS.2007.370210
– ident: ref32
  doi: 10.1038/nphys1130
– ident: ref20
  doi: 10.1007/978-3-319-27671-7_27
– ident: ref47
  doi: 10.1109/CVPR.2016.616
– start-page: 125
  year: 2014
  ident: ref33
  article-title: Comparative analysis of data structures for approximate nearest neighbor search
  publication-title: Data Anal
– ident: ref16
  doi: 10.1109/CVPR.2014.298
– ident: ref1
  doi: 10.1023/B:VISI.0000029664.99615.94
– start-page: 45
  year: 0
  ident: ref58
  article-title: Software framework for topic modelling with large corpora
  publication-title: Proceedings of the LREC 2010 Workshop on New Challenges for NLP Frameworks
– ident: ref34
  doi: 10.14778/2824032.2824059
– start-page: 1225
  year: 0
  ident: ref8
  article-title: Practical and optimal LSH for angular distance
  publication-title: Proc Adv Neural Inf Process Syst
– year: 2013
  ident: ref59
  article-title: Distance-based similarity models for content-based multimedia retrieval
– ident: ref54
  doi: 10.1109/5.726791
– ident: ref12
  doi: 10.1109/CVPR.2012.6248038
– ident: ref31
  doi: 10.1038/35022643
– start-page: 2055
  year: 0
  ident: ref14
  article-title: Efficient indexing of billion-scale datasets of deep descriptors
  publication-title: Proc IEEE Conf Comput Vis Pattern Recognit
– ident: ref55
  doi: 10.1016/j.is.2015.02.001
– ident: ref37
  doi: 10.1137/1.9781611973068.36
– ident: ref22
  doi: 10.1145/2020408.2020576
– ident: ref36
  doi: 10.1038/ncomms8651
– ident: ref13
  doi: 10.1109/TPAMI.2010.57
SSID ssj0014503
Score 2.7377121
Snippet We present a new approach for the approximate K-nearest neighbor search based on navigable small world graphs with controllable hierarchy (Hierarchical NSW,...
SourceID proquest
pubmed
crossref
ieee
SourceType Aggregation Database
Index Database
Enrichment Source
Publisher
StartPage 824
SubjectTerms Algorithms
approximate search
Approximation algorithms
artificial intelligence
big data
Biological system modeling
Brain modeling
Complexity theory
Data models
data structures
Graph and tree search strategies
Graphs
graphs and networks
information search and retrieval
information storage and retrieval
information technology and systems
Metric space
Multilayers
nearest neighbor search
Performance evaluation
Proximity
Routing
Search problems
search process
Searching
similarity search
Structural hierarchy
Title Efficient and Robust Approximate Nearest Neighbor Search Using Hierarchical Navigable Small World Graphs
URI https://ieeexplore.ieee.org/document/8594636
https://www.ncbi.nlm.nih.gov/pubmed/30602420
https://www.proquest.com/docview/2374696299
https://www.proquest.com/docview/2163011259
Volume 42
WOSCitedRecordID wos000526541100005&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
journalDatabaseRights – providerCode: PRVIEE
  databaseName: IEEE Electronic Library (IEL)
  customDbUrl:
  eissn: 2160-9292
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0014503
  issn: 0162-8828
  databaseCode: RIE
  dateStart: 19790101
  isFulltext: true
  titleUrlDefault: https://ieeexplore.ieee.org/
  providerName: IEEE
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1La9wwEB42IYf2kGcfzgsVemud2F6tJR2XkHQLrQltCnszsh7pwtYbsruhP78z8oMU0kBvxpZkwzfjmZFmvgF4b9BmWl7lMbcjERMBXaxHsoqlcWmiE--tD-z6X0RRyOlUXQ_gY18L45wLyWfujC7DWb5dmDVtlZ3LkSJ-qw3YEEI0tVr9iQEfhS7I6MGghmMY0RXIJOr85nr89TNlccmzTErFBTXPQVeZzFPylz0KDVb-7WsGm3O1839fuwvbrW_Jxo0w7MHA1fuw0_VtYK0a78PLRySEB_DzMrBI4FpM15Z9W1Tr5YqNiWv89wz9WccK4rnFewVto6LMsCZHmYV0AzaZUQ1zaKkyZ4V-mN1SNRb7_kvP5yzk6rBPxIq9fAU_ri5vLiZx238hNjwVq5g7JxOh0XplXpKd98Yrpz2asNzrNDPK5dITHY_JrZd2WKU60QoRtobj5OFr2KwXtXsLTPHUVLaqjOFDnmupK66sVQqdOW2y3EaQdiiUpiUnpx4Z8zIEKYkqA4glgVi2IEbwoZ9z11BzPDv6gCDqR7boRHDcgV222rsss6HgucrRUkfwrn-MekeHKbp2izWOQUcW_40YPUbwphGSfu1Otg6ffucRvMgoag_5P8ewubpfuxPYMg-r2fL-FIV7Kk-DcP8B74j0TQ
linkProvider IEEE
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1LT9wwEB5RilRxKAX6CI_WSL21gSTrJPZxhaCLukSo3UrcIsePdqVttmJ3UX9-Z5yHQCpI3KLEdiJ943xje-YbgI8aOdPwKgu5SfOQBOhClYoqFNrGkYqcM86r64_zohDX1_JqDT73uTDWWh98Zo_p0p_lm7le0VbZiUgl6Vs9g-cp50ncZGv1ZwY89XWQ0YfBOY4LiS5FJpInk6vh5QXFcYnjRAjJcyqfg84yEVR0j5F8iZWHvU3POudbT_veV_Cy9S7ZsDGHbViz9Q5sdZUbWDuRd2DzjgzhLvw68zoSOBZTtWHf5tVqsWRDUhv_O0WP1rKClG7xXkEbqWg1rIlSZj7ggI2mlMXsi6rMWKFupz8pH4t9_61mM-ajddgX0sVevIYf52eT01HYVmAINY_zZcitFVGukL8SJ4jpnXbSKockljkVJ1raTDgS5NGZccIMqlhFSiLGRnPsPHgD6_W8tu-ASR7rylSV1nzAMyVUxaUxUqI7p3SSmQDiDoVSt_LkVCVjVvplSiRLD2JJIJYtiAF86vv8acQ5Hm29SxD1LVt0AjjowC7b-bsok0HOM5khVwdw1D_GmUfHKaq28xW2QVcW_464fgzgbWMk_didbe39_50f4MVocjkuxxfF133YjJtQcAoHOoD15c3KHsKGvl1OFzfvvY3_A_2u9vM
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Efficient+and+Robust+Approximate+Nearest+Neighbor+Search+Using+Hierarchical+Navigable+Small+World+Graphs&rft.jtitle=IEEE+transactions+on+pattern+analysis+and+machine+intelligence&rft.au=Malkov%2C+Yu+A&rft.au=Yashunin%2C+D+A&rft.date=2020-04-01&rft.eissn=1939-3539&rft.volume=42&rft.issue=4&rft.spage=824&rft_id=info:doi/10.1109%2FTPAMI.2018.2889473&rft_id=info%3Apmid%2F30602420&rft.externalDocID=30602420
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0162-8828&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0162-8828&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0162-8828&client=summon