Another virtue of wavelet forests?

TL;DR

This paper demonstrates that wavelet forests offer improved access locality over wavelet trees, making them advantageous even without higher-order compression or when not using BWT, supported by experimental evidence.

Contribution

The paper provides experimental evidence that wavelet forests have better access locality than wavelet trees, expanding their applicability beyond BWT-based compression.

Findings

Wavelet forests have superior access locality compared to wavelet trees.

Wavelet forests are effective even without higher-order compression.

They are beneficial when the text is not a BWT.

Abstract

A wavelet forest for a text $T [1..n]$ over an alphabet $\sigma$ takes $n H_0 (T) + o (n \log \sigma)$ bits of space and supports access and rank on $T$ in $O (\log \sigma)$ time. K\"arkk\"ainen and Puglisi (2011) implicitly introduced wavelet forests and showed that when $T$ is the Burrows-Wheeler Transform (BWT) of a string $S$, then a wavelet forest for $T$ occupies space bounded in terms of higher-order empirical entropies of $S$ even when the forest is implemented with uncompressed bitvectors. In this paper we show experimentally that wavelet forests also have better access locality than wavelet trees and are thus interesting even when higher-order compression is not effective on $S$, or when $T$ is not a BWT at all.