On the LSH Distortion of Ulam and Cayley Similarities

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Chierichetti, Flavio, Giacchini, Mirko, Kumar, Ravi, Tani, Erasmo
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910212018929664
author Chierichetti, Flavio
Giacchini, Mirko
Kumar, Ravi
Tani, Erasmo
author_facet Chierichetti, Flavio
Giacchini, Mirko
Kumar, Ravi
Tani, Erasmo
contents Locality-sensitive hashing (LSH) has found widespread use as a fundamental primitive, particularly to accelerate nearest neighbor search. An LSH scheme for a similarity function $S:\mathcal{X} \times \mathcal{X} \to [0,1]$ is a distribution over hash functions on $\mathcal{X}$ with the property that the probability of collision of any two elements $x,y\in \mathcal{X}$ is exactly equal to $S(x,y)$. However, not all similarity functions admit exact LSH schemes. The notion of LSH distortion measures how multiplicatively close a similarity function is to having an LSH scheme. In this work, we study the LSH distortion of the Ulam and Cayley similarities, which are popular similarity measures on permutations of $n$ elements. We show that the Ulam similarity admits a sublinear LSH distortion of $O(n / \sqrt{\log n})$; we also prove a lower bound of $Ω(n^{0.12})$ on the best LSH distortion achievable. On the other hand, we show that the LSH distortion of the Cayley similarity is $Θ(n)$.
format Preprint
id arxiv_https___arxiv_org_abs_2605_11921
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle On the LSH Distortion of Ulam and Cayley Similarities
Chierichetti, Flavio
Giacchini, Mirko
Kumar, Ravi
Tani, Erasmo
Data Structures and Algorithms
Information Retrieval
Locality-sensitive hashing (LSH) has found widespread use as a fundamental primitive, particularly to accelerate nearest neighbor search. An LSH scheme for a similarity function $S:\mathcal{X} \times \mathcal{X} \to [0,1]$ is a distribution over hash functions on $\mathcal{X}$ with the property that the probability of collision of any two elements $x,y\in \mathcal{X}$ is exactly equal to $S(x,y)$. However, not all similarity functions admit exact LSH schemes. The notion of LSH distortion measures how multiplicatively close a similarity function is to having an LSH scheme. In this work, we study the LSH distortion of the Ulam and Cayley similarities, which are popular similarity measures on permutations of $n$ elements. We show that the Ulam similarity admits a sublinear LSH distortion of $O(n / \sqrt{\log n})$; we also prove a lower bound of $Ω(n^{0.12})$ on the best LSH distortion achievable. On the other hand, we show that the LSH distortion of the Cayley similarity is $Θ(n)$.
title On the LSH Distortion of Ulam and Cayley Similarities
topic Data Structures and Algorithms
Information Retrieval
url https://arxiv.org/abs/2605.11921