Locality Sensitive Hashing in Hyperbolic Space

TL;DR

This paper introduces the first locality sensitive hashing (LSH) method tailored for hyperbolic spaces, enabling efficient approximate nearest neighbor searches in these non-Euclidean geometries.

Contribution

It presents a novel LSH construction for hyperbolic spaces, including explicit bounds on the parameter , and extends Euclidean lower bounds to hyperbolic geometry.

Findings

Achieves   in hyperbolic plane

Uses dimension reduction for higher dimensions

Establishes lower bounds extending Euclidean results

Abstract

For a metric space $(X, d)$, a family $\mathcal{H}$ of locality sensitive hash functions is called $(r, cr, p_1, p_2)$ sensitive if a randomly chosen function $h\in \mathcal{H}$ has probability at least $p_1$ (at most $p_2$) to map any $a, b\in X$ in the same hash bucket if $d(a, b)\leq r$ (or $d(a, b)\geq cr$). Locality Sensitive Hashing (LSH) is one of the most popular techniques for approximate nearest-neighbor search in high-dimensional spaces, and has been studied extensively for Hamming, Euclidean, and spherical geometries. An $(r, cr, p_1, p_2)$-sensitive hash function enables approximate nearest neighbor search (i.e., returning a point within distance $cr$ from a query $q$ if there exists a point within distance $r$ from $q$) with space $O(n^{1+\rho})$ and query time $O(n^{\rho})$ where $\rho=\frac{\log 1/p_1}{\log 1/p_2}$. But LSH for hyperbolic spaces $\mathbb{H}^d$ remains largely unexplored. In this work, we present the first LSH construction native to hyperbolic space. For the hyperbolic plane $(d=2)$, we show a construction achieving $\rho \leq 1/c$, based on the hyperplane rounding scheme. For general hyperbolic spaces $(d \geq 3)$, we use dimension reduction from $\mathbb{H}^d$ to $\mathbb{H}^2$ and the 2D hyperbolic LSH to get $\rho \leq 1.59/c$. On the lower bound side, we show that the lower bound on $\rho$ of Euclidean LSH extends to the hyperbolic setting via local isometry, therefore giving $\rho \geq 1/c^2$.