Pith. sign in

REVIEW 1 cited by

Unsupervised Space Partitioning for Nearest Neighbor Search

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2206.08091 v2 pith:QJKLE25J submitted 2022-06-16 cs.LG cs.DBcs.IRstat.ML

Unsupervised Space Partitioning for Nearest Neighbor Search

classification cs.LG cs.DBcs.IRstat.ML
keywords annspartitioningclusteringfunctionlossmethodsearchstate-of-the-art
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Approximate Nearest Neighbor Search (ANNS) in high dimensional spaces is crucial for many real-life applications (e.g., e-commerce, web, multimedia, etc.) dealing with an abundance of data. This paper proposes an end-to-end learning framework that couples the partitioning (one critical step of ANNS) and learning-to-search steps using a custom loss function. A key advantage of our proposed solution is that it does not require any expensive pre-processing of the dataset, which is one of the critical limitations of the state-of-the-art approach. We achieve the above edge by formulating a multi-objective custom loss function that does not need ground truth labels to quantify the quality of a given data-space partition, making it entirely unsupervised. We also propose an ensembling technique by adding varying input weights to the loss function to train an ensemble of models to enhance the search quality. On several standard benchmarks for ANNS, we show that our method beats the state-of-the-art space partitioning method and the ubiquitous K-means clustering method while using fewer parameters and shorter offline training times. We also show that incorporating our space-partitioning strategy into state-of-the-art ANNS techniques such as ScaNN can improve their performance significantly. Finally, we present our unsupervised partitioning approach as a promising alternative to many widely used clustering methods, such as K-means clustering and DBSCAN.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Cluster with Auctions for Vector Search

    cs.IR 2026-07 conditional novelty 6.0

    CwA jointly learns a balanced database partition and a query-adapted neural probing function, achieving up to 4.7x higher ANNS throughput at equal recall.